Comparing Wealth: Data Quality of the HFCS

Abstract

The Household Finance and Consumption Survey (HFCS) provides information about household wealth (real and financial assets as well as liabilities) from 15 Euro‐countries after the financial crisis of 2007/8. The survey will be the central dataset in this topic in the future. However, several aspects point to potential methodological constraints regarding crosscountry comparability. Therefore the aim of this paper is to get a better insight in the data quality of this important data source. We will first present a synopsis of cross‐country differences, which is the core of the paper. We will compare the sampling processes, the interview modes, the oversampling techniques, the unit and item non‐response rates and how itis dealt with them via weighing and imputation as well as further points which might restrict country comparability. In addition we give a first insight in the selectivity of item nonresponse in a cross‐national setting. We make use of logit models as well as apply a decomposition method suggested by Fairlie (1999, 2005) to identify differences in characteristics as well as structural (cultural) differences in the item non‐response missing process.