web page contains state-level time-series data on individual and family
income inequality in the U.S. from 1963 to 2003. We have computed
measures of income inequality at the state and national level using
the following datasets:
Population Survey (CPS) from 1963 to 2003. These data were supplied
The CPS Merged Outgoing
Rotation Group (MORG) survey from 1979 to 2000. These data were
supplied by the National Bureau of
The U.S. Census in 1970, 1980, 1990 and 2000 based on 1% samples.
These data were downloaded from the University of Minnesota Population
Public Use Microdata Series website.
datasets available here include two measures of earnings inequality:
1) the ratio of the 90th income percentile to the 10th percentile;
and 2) the gini coefficient. They also contain information on income
deciles, mean earnings, maximum and minimum values, standard deviations
and sample sizes. Each dataset contains this information for all respondents
and for men and women separately.
For the March CPS survey, data on four measures of income are
available: total annual family income, total annual individual income,
annual individual wage and salary income and weekly individual wage
and salary income. For each of these measures, we provide two
datasets: one contains smoothed data and the other contains the raw,
unsmoothed data. We recommend use of the smoothed data. Similar measures of income are available for the Census data. For
the CPS MORG survey, data on weekly wage & salary earnings as
well as hourly wages are available. The Census and MORG datasets are
not smoothed. All datasets can be downloaded as either space-delimited
ASCII text files or STATA (release 7) data files. Reading the accompanying
documentation files before using the data is strongly recommended.
These data were prepared by
Jake Rosenfeld for the
Sage program on the Social Dimensions of Inequality. Direct
all inquiries to Josh Guetzkow at