The Survey of Income and Program Participation (SIPP) is one of the few household surveys that provide data on the distribution of net worth. This study describes SIPP as a source of data on net worth and presents models based on SIPP that identify the factors that are associated with net worth.

In a structural or causal sense, net worth is a function of inheritance, past levels of disposable income, the propensity to save, and the return on investments. These factors are difficult or impossible to measure in a household survey, so no attempt is made to present a structural model that would estimate causal relationships.

In a recent Census Bureau report, we used the SIPP asset and liability module to analyze and describe the characteristics of household wealth holdings in the U.S. [Bureau of the Census, 1986]. The report presented a descriptive analysis of asset ownership and wealth holdings based on tabulations for various subgroups of the population. The purpose of this paper is to further analyze the factors correlated with wealth holdings by using a multivariate regression model. In order to estimate the model, data from other SIPP supplements were matched to the wealth information. A description of the SIPP data set is presented in the next section. A model is developed in the third section, and the empirical results are presented in the final section.