Improving Estimates on Housing Wealth With Survey and Administrative Data

ABSTRACT This study estimates housing wealth by combining survey data from the Survey on Household Income and Wealth with administrative data from the Italian Real Estate Market Observatory. Both sources have limitations: survey data are subject to reporting bias, while administrative data may not capture property‐specific characteristics. To address these issues, we employ a finite mixture model combined with factor score prediction to account for measurement errors and exploit the complementarity of the two sources. The results reveal systematic differences between survey and administrative estimates, and show that their integration leads to more reliable and accurate estimates of housing wealth, improving our understanding of its distribution across households.