The main goal of this paper is to complement the Argentine mean income series from National Accounts with inequality estimates in order to obtain aggregate welfare series, which are a better measure of economic performance than the commonly used per capita income statistics. Inequality indices are computed from household survey data adjusted for non-response and income underreporting. The statistical significance of changes in inequality and welfare measures is checked using bootstrapping techniques. One of the main conclusions of the paper is that while welfare assessments coincide among different value judgments in some periods (e.g. 1991-1994), they widely vary in some others, particularly between 1994 and 1998, when the economy experienced moderate growth and large increases in inequality. It is argued that the period 1994-1998 provides an unprecedented laboratory for distinguishing the social preferences of different analysts, according to their evaluation of the performance of the Argentine economy