Courant Research Centre: Poverty, Equity and Growth - Discussion Papers 237
This paper introduces a newly constructed Theil index of between-sectoral manufacturing wage inequality and empirically tests whether the measure can serve as a basis for more general statements about the evolution of broader concepts of inequality, as argued by the authors of the University of Texas Inequality Project (UTIP) for their very similar index. Building on prior work of the UTIP, several concerns regarding the treatment of the raw data as well as important questions of internal and external validity are addressed. The index is based on sector-level data from the UNIDO Industrial Statistics for manufacturing, and I provide a detailed account of how the unbalanced raw data were treated. The newly computed index has the advantage of being consistently measured across countries and years, which makes it a valuable resource for empirical studies focusing on changes in the manufacturing structure within countries over long periods of time. However, its narrow scope also restricts the applicability of the index for other, broader uses. I argue that the latter point is one of the main drawbacks of the index and present evidence that the generalizability from between-sectoral manufacturing wage inequality to overall income inequality is severely limited. This applies not only to the extent to which the index allows conjectures about the overall level of income inequality in a society. There is reason to also question the 'internal' capability of the index to accurately reflect developments in manufacturing wage inequality. I therefore do not recommend using the index as a basis for inference about the development of broader concepts of wageor income equality.
Inequality measurement Wage inequality Income inequality Manufacturing wage inequality