Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
We use a proprietary dataset on the floor-level operations at the largest rail mill in India to study the response of productivity to the threat of entry. Output per active shift increased by 28 percent over 3 years with minimal changes in physical capital and employment. By combining data on the timing of various training programs in the mill with shift-level variation in worker composition, we are able to attribute over half of the higher productivity to training specifically targeted toward improving rail output. Our work suggests high returns to knowledge-enhancing investment in emerging economies. (JEL D22, D24, J24, L23, L32, L61, O14)