Back on the Rails -- Competition and Productivity in State-owned Industry

The importance of Total Factor Productivity (TFP) in explaining output changes is widely accepted, yet its sources are not well understood. We use a proprietary data set on the oor-level operations at the Bhilai Rail and Structural Mill (RSM) in India to understand the determinants of changes in plant productivity between January 2000 and March 2003. During this period there was a 35% increase in output with minimal changes in the stock of physical capital or the number of employees, but sizable reductions in the number and duration of various types of production delays. We model interruptions to the production process as a function of worker characteristics and find that a large part of the avoidable delay reductions are attributable to training. Overall, changes in all delays account for over half the changes in productivity. Our results provide some explanation for the large within-industry di erences in productivity observed in developing countries and also suggest that specic knowledge-enhancing investments can have very high returns. Our approach also provides an example of how detailed data on production processes can be fruitfully used to better understand TFP changes, which have typically been treated as residuals in growth-accounting exercises.