Association for Public Policy Analysis & Management

The literature has identified several pitfalls in measuring the total distribution of income via household surveys, while substantially smaller problems are found for employee compensation. Similarly, a key challenge in measuring the labour income share is to estimate the labour income of the self-employed, in contrast to employee compensation. Both the labour income share and the labour income distribution play a key role in income inequality. I estimate them, accounting for self-employment, based on microdata. This approach has often been considered appealing but not practical for a large panel of countries. Meanwhile, distribution measurement efforts have seldom focused on labour income.

This analysis includes household surveys for 96 countries, spanning 2004-2017, mainly from the ILO Harmonized Microdata collection. I estimate the labour income of the self-employed based on employees of similar characteristics. Afterwards, the labour income share and distribution are produced. I analyse existing evidence concerning the data quality of household surveys and find that they can provide reasonable estimates of labour income distribution. The main results concerning the labour income share and distribution are: (i) the global labour income share is declining and countercyclical; (ii) the effects of self-employment on the labour income share are highly heterogeneous –highlighting the limitations of widely used rules of thumb; (iii) labour income inequality decreases strongly with national income level, hence average cross-country differences are greatly exacerbated; (iv) within countries, relative increases of labour income at the upper end of the distribution are associated, on average, with relative losses for the rest.