The paper examines sensitivity influences on the German personal income distribution in a time-series perspective as well as in a methodically broad manner. The author discusses the following issues: (1) For the first time, (reference) income-dependent, so-called variable equivalence scales are explicitly and extensively applied in a distributional analysis of German data which causes significant increases of income inequality compared with income-independent, constant equivalence scales. (2) Concerning different demarcations of income areas the pattern of income inequality in Germany 1995-2009 is not distinctively changed in the several variants considered. (3) For three alternative inequality indicators out of the class of Generalized Entropy indicators (mean logarithmic deviation, one of Theil’s measures of entropy, and normalized coefficient of variation), the patterns of income inequality over time are nearly the same. (4) Regarding current monthly household net income versus yearly household net income of the previous year, different patterns with respect to income inequality occur during the ob-served period of time. Especially in the first decade of the 21st century the corresponding pat-terns differ from each other. In a further step the new approach related to income distribution, which incorporates variable equivalence scales, is applied to socio-demographic stratification to exemplarily demonstrate the power of this new approach. All in all, the analyses of the paper refer to the necessity of a rigorous methodological foundation of distributional studies, especially concerning the selection of a set of (preferably variable) equivalence scales, the choice of the inequality indicator, and – not least – of the income variable.
Read full abstract