Analytic Approaches to Measuring the Black-White Wealth Gap

Abstract

While wealth inequality is large and growing, racial wealth inequality is even larger and persistent. Yet, the scope of the discussion on racial wealth inequality is hindered, in part, due to data and measurement issues. By comparing data from the Survey of Consumer Finances (SCF) and the Panel Study of Income Dynamics (PSID), we contrast the traditional mean Oaxaca-Blinder decomposition to the distributional Recentered Influence Function (RIF) methods developed by Firpo et al. (2018), while also exploring estimates for varied transformations of the wealth distribution. We notice a remarkable similarity among the untransformed, logarithm-transformed, and inverse hyperbolic sine-transformed versions in both the SCF and PSID data sets. The Oaxaca-Blinder (mean) model shows that disparities in receiving an inheritance explain a larger portion of the racial wealth gap than variations in educational attainment. Conversely, the RIF model at the median indicates that disparities in educational attainment account for more of the wealth gap than differences in inheritance receipt. We show how measurement of the racial wealth gap shifts depending on the regression decomposition method and wealth transformation measure used.

Keywords: race, wealth, inequality, measurement, Oaxaca-Blinder, RIF decomposition

JEL Classification: C13, C18, I14, I30, J31