Bridging Techniques in the Redesign of the Italian Survey on Household Income and Wealth

The design of the Bank of Italy’s Survey on Household Income and Wealth was revised in 2020 to reduce non-sampling error in households’ income and wealth and improve data quality. The new sample allocation resulted in greater participation in the upper parts of the income distribution, determining a reduction in standard errors and in the bias of income and wealth estimators. However, the revision of the sample makes it difficult to compare the results with those obtained in previous survey waves. This paper discusses different weighting systems for taking these differences into account, obtained following three main methodological approaches: cell weighting, raking and inverse probability weighting. Comparing results across different dimensions, the method that produces the most reliable results is based on the use of the raking technique and, therefore, it is the one recommended for time series analysis