Estimating the distribution of household wealth: methods for adjusting survey data estimates using national accounts and rich list data

Income and wealth surveys are usually affected by unit non-response and reporting errors, which con-tribute to a mismatch with macroeconomic figures from national accounts. In this paper, we develop anovel allocation method to address these two issues simultaneously, when only limited external informa-tion is available. The proposed approach combines information from a power law (Pareto) model withimputation procedures based on calibration. We apply the proposed simultaneous approach to producedistributional indicators for four Euro-Area countries that are consistent with their national accounts.In particular, we use data from the 2014 Finnish, French, German, and Italian Household Finance andConsumption Survey and rich list data from Forbes or national press sources, along with householdsector aggregates from national accounts, as auxiliary sources of information. A bootstrap procedure isalso applied to evaluate the precision of the final estimates.