Using income data to predict wealth

Article type: Research Article

Authors: Kennickell, Arthur B.

Affiliations: Board of Governors of the Federal Reserve System (Ret.), Washington, DC 20551, USA. E-mail: [email protected]

Abstract: This paper considers the relationship between income and wealth using data from three sources: the Survey of Consumer Finances (SCF), the IRS and Forbes Magazine. Because of the legal and ethical constraints on the use of some of these data, the focus of the paper is limited to a few issues that are relevant to improving the design of the SCF and understanding better the quality of the information collected. The most important goal of the paper is estimation of functional relationships between income and wealth. Work to this end results in model predictions that are more effective than using income alone as a proxy for wealth. However, it is clear that the relationship is not strong, and more work is needed, particularly to address the noise introduced into the relationship by year-to-year income variability.

Keywords: Wealth modeling, sampling, survey of consumer finances

DOI: 10.3233/SJI-160312

Journal: Statistical Journal of the IAOS, vol. 33, no. 1, pp. 41-49, 2017