Optimal sampling design for household finance surveys using administrative income data

Household finance surveys, which collect detailed information on household income and wealth, are increasingly used for policy-making. They should provide an accurate picture of the economic situation of all households. Unfortunately, the upper parts of the wealth distribution are often missing in household surveys. Since rich households concentrate a large share of total income and wealth, survey-based estimators may be biased. The ideal situation would be to have access to auxiliary information on household finances at the design stage. This is rarely the case. In this paper we present an application that uses tax records in the design of a major survey on household finances. We discuss the methodological challenges of using administrative information for designing the sample. We propose a method for an optimal stratification and sample allocation