The “Missing Rich” in Household Surveys: Causes and Correction Approaches

Inequality measures based on household surveys may be biased because they fail to capture the upper tail of the income distribution properly. The "missing rich" problem stems from sampling errors, item and unit nonresponse, underreporting of income, and data preprocessing techniques like top coding. This paper reviews salient approaches to address the underrepresentation of the rich in household surveys. Approaches are classified based on information sources and method. In terms of information sources, the distinction is between within-survey data and survey data combined with external sources (e.g., tax records). In terms of methods, we identify three categories: replacing, reweighting, and combined reweighting and replacing. We show that income inequality levels and trends are sensitive to the correction approach. This paper is a companion piece to the chapter of the same name and includes all the appendices that could not be incorporated into the chapter due to space limitations. (Stone Center on Socio-Economic Inequality Working Paper)