Automation and inequality within firms Evidence from South Africa

Dr Tim Köhler (DPRU, UCT) and co-authors estimate the causal effects of automation on employment and wages within South Africa’s formal manufacturing sector, using matched employer-employee tax administrative microdata (2014 to 2021) and customs import data to identify firms adopting automation capital goods. Applying a staggered difference-in-differences design, the study examines how automation adoption reshapes employment and within-firm wage inequality, a sector-focused contribution to a literature dominated by high-income countries, in a nation where the labour market is the primary driver of inequality. Presented at the 2026 Public Economics Conference.