How much does management matter for economic performance? Despite a large industry of business schools, consulting firms, and airport books purporting to teach you the secrets of good management over the course of your next flight, the answer until very recently has been “we don’t know”. In a recent review, Chad Syverson goes as far as to say “no potential driving factor of productivity has seen a higher ratio of speculation to empirical study”.
Together with colleagues from Stanford and Berkeley, I have been working for the last couple of years to try and understand how much management matters by means of a randomized experiment among textile factories in India. In common with most firms in developing countries, the firms (with 300 workers on average) we were working with did not collect and analyze data systematically in their factories, had few systems for regular maintenance and quality control, had weak human resource systems for promoting and rewarding good performers, and had little control over inventory levels. The result was a high level of quality defects, large stockpiles of unorganized inventories, and frequent breakdowns of machines. 20 percent of the labor force was occupied solely in checking and repairing defective fabric (see picture).