Measuring inequality and dependences between income sources with administrative data and survey data

Abstract

This paper aims at analyzing the effects of changing from survey to administrative data on inequality and its structure. Taking advantage of the Spanish Survey on Income and Living Conditions (ECV) that continued asking households for their income despite assigning them the income data provided by the Tax Agency and the Social Security administration, different analyses are carried out. By using copula functions we pay special attention to the effect on the dependences between income sources. We find a significant growth in the disposable income of households when using administrative data. The incomes of both tails of the distribution increase considerably more than middle incomes, and administrative data produce significantly lower levels of inequality. Using administrative instead of survey data also gives rise to changes in the structure of inequality by income sources, rising the contribution of capital income. Both methods of data collection also produce significant differences in the observed dependences between income sources.