The paper reviews the most significant methods used to reveal the contributions of different income sources in achieving total income inequality. These methods are classified into natural decompositions, marginal decompositions, quasi-marginal approaches, and Shapley decompositions. A comprehensive catalogue of the methods is presented, but the paper goes beyond that, through comparing the results obtained by different methods and discussing their interpretation. Particular attention is given to the issue of the uniformly distributed income source, which is assigned a zero contribution by some leading methods, such as the natural decompositions of the Gini index and the variance. This result is counterintuitive from the relative inequality perspective, which says that a positive uniform transfer should decrease inequality. The paper argues that these decompositions are not incorrect but simply reflect the absolute inequality view, according to which uniform changes do not affect inequality. Furthermore, the methods are scrutinised utilizing hypothetical and real-world sample data on household incomes. A comparative analysis for fifteen EU countries using EUROMOD shows that different methods provide relatively similar rankings of countries regarding the sources’ proportionate contributions.
