Monopsony and Measurement Error in the South African Labour Market

In this paper, I analyse the extent to which monopsony power is present in the South African labour market by estimating the wage elasticity of labour supply to the firm, following Manning's (2003) method. I also consider the extent to which the method of identifying jobto-job separations and the use of poorly imputed earnings data by StatsSA changes results. I use panel data from the Labour Force Survey and Quarterly Labour Force Survey in South Africa, although only for waves in the QLFS in which I have earnings data with no imputations by StatsSA. I find a low elasticity of labour supply to the firm in South Africa of between 0,68 and 0,83, which suggests that there is substantial monopsony power in South Africa. These estimates are far off infinity, which suggests that using perfect competition to model the South African labour market is not a realistic assumption. I find little difference in this elasticity to the firm across gender and race overall, but some differences across gender within race and across race within gender. I find that the labour supply of more educated individuals is more elastic to the firm than those with less education and that within higher educated groups, men are supplied more elastically than women. This suggests that education lessens vulnerability to monopsony power, and more so for men. Lastly, I find that results are very sensitive to the method of identifying separations, as well as to the use of the imputed StatsSA earnings data. In both cases, the elasticities estimated are less than half the above estimates. Despite measurement error being a concern, the low estimates of the elasticity to the firm are unlikely to be biased downwards to the extent that the South African labour market more closely resembles perfect competition. Thus, those designing policy for the South African labour market should do so from an assumption that the market is imperfectly competitive. Furthermore, these estimates are not far off estimates using administrative data, which suggests that survey data can inform on monopsony power in a labour market when carefully analysed.

Grantor dc:publisher.institution

School of Economics

Year dc:date.issued

2024

Author dc:creator

• Forster, Nicholas

Advisor dc:contributor.advisor

• Kerr, Andrew

• Economics

Handle dc:identifier.uri

http://hdl.handle.net/11427/40914

OAI identifier oai:identifier

oai:open.uct.ac.za:11427/40914