Since the early nineties, microsimulation techniques have been frequently used on research in Public Economics in Spain, typically for measuring the effects of tax changes. For this purpose, one of the most frequently used data sources has been the European Community Household Panel (ECHP). Despite providing very rich information, this source supplies undervalued capital income data with little disaggregation. This problem must be solved if we want to give certainty to microsimulation exercises. Our proposal consists of assigning values of capital income that come from a sample of Spanish taxpayers (Panel de Declarantes del IRPF) to ECHP by means of statistical matching techniques. The result is a more complete and robust database and, therefore, more reliable when evaluating the effects of tax reforms on revenue and redistribution. These conclusions are illustrated by performing several tax simulations.

Key words: personal income tax, microsimulation, European Community Household Panel, statistical matching.
JEL classification: H24 C81.