Estimating Post-tax Social Insurance Benefits: Validity Problems in Comparative Analyses of Net Income Components from Household Income Data

Comparative household micro income databases do not report the level of social transfers after taxation. Consequently, disaggregated redistributive analyses of the welfare state are based on gross income components. In most countries, however, social insurance benefits are subject to taxation. In such instances, the level and equalising effect of social insurance to income inequality are overestimated, both in absolute terms and in relation to nontaxable benefits. One way to avoid this problem is to estimate the level of net social insurance by the use of a so-called proportional tax estimation technique. This technique, however, causes a misspecification of the level of net social insurance in cases where taxation is established at the individual level. In this paper we therefore apply the proportional tax estimation technique for validity analyses on household income data. The question is to what extent this estimation of taxes misspecifies the level of net social insurance. It is found that the proportional tax estimation is viable when separating social and fiscal policies in comparative analyses on household micro income data. The underestimation of the level of net social insurance which is due to the application of the proportional tax estimation technique is negligible compared with the overestimation occurring from not taking taxes into account.