Much work has been devoted to deriving and extending decomposable inequality and poverty measures. For example, the use of decomposition techniques to partition inequality into within-race group and between-race group components. This paper espouses this by using a decomposition technique based on the Gini coefficient to discern the relative importance of the major income components in determining overall income inequality. Such analysis shows which income components are more important than others in causing income inequality within any racial group or, more generally, a discussion of the sensitivity of the Gini coefficient to marginal changes in income sources. The use of sensitivity analysis in conjunction with the decomposition analysis helps identify key labor market, state welfare and asset ownership patterns integral to the generation of inequality in South Africa. Finally, a welfare index is derived as a weighted combination of average income and distributional changes. This paper uses 1993 Living Standards Measurement Study survey data. First, the decomposition analysis is illustrated using all African households. The bulk of the paper then analyzes rural African households in the former homeland areas of South Africa as an example of how such work can make more specific inputs into policy debates. The rural sample is first looked at as a whole and then the sample is divided into households above and below a Household Subsistence Level.