Income Inequality Data in Growth Empirics: From Cross-Sections to Time Series

As in any other field of applied macro-economic or econometric research, researchers who study income inequality have to look for suitable data. Although most researchers just draw on some ready-made dataset, finding reliable data is not that straightforward and can even be very troublesome. This paper highlights some of the pitfalls in the use of inequality data. We deal with sampling problems, the choice of equivalence scale, one dimensional inequality measures, etc. We also introduce and describe a new secondary dataset on inequality data for a number of OECD countries. The main innovation of the dataset is that it creates the possibility to perform a time series analysis on inequality data. The data bring to the fore an additional problem: non-stationary behaviour of inequality series.