Researchers often need to work with categorical income data. While the typical nonparametric (including midpoint) and parametric estimation methods used to estimate summary statistics both have advantages, they all carry assumptions which cause them to deviate in important ways from real world distributions of income. The method introduced here, Random Empirical Distribution Imputation (REDI), imputes discrete observations using binned income data, while also calcu- lating summary statistics. REDI achieves this through random cold-deck imputation from a real world reference dataset (here, the Current Population Survey ASEC). This imputation method reconciles bins between datasets or across years and handles top incomes. REDI has other ad- vantages of computing an income distribution that is nonparametric, bin consistent, area- and variance-preserving, continuous, and computationally fast. I provide proof of concept using two years of the American Community Survey.
REDI for Binned Data: A Random Empirical Distribution Imputation method for estimating continuous incomes
