Estimating Small Area Income Distributions and Income Statistics via the Inequality Process (IP)

John Angle, Inequality Process Institute
Kenneth C. Land, Duke University

The distribution of earned income in a public use microdata area (PUMA) can be estimated from the American Community Survey’s (ACS) public use microdata sample (PUMS). The ACS also publishes this distribution for some sub-PUMA areas along with several scalar statistics of earned income. This paper uses conventional techniques to estimate the distribution of earned income in each sub-PUMA area in the vicinity of Philadelphia, Pennsylvania and Raleigh-Durham, North Carolina using the PUMA level distribution and the scalar statistics of earned income in each sub-PUMA area as the source of “strength”, pretending that its earned income distribution is unknown. The paper then compares the estimates of the conventional small area estimator to those of an unconventional parametric estimator of income distributions, the Macro Model of the Inequality Process.