Markets are quite good at allocating products and services to those who want them the most (and can pay to fulfil those desires). The insight of Ronald Coase, and after him, Oliver Williamson, was to view business firms as solutions to the problem of resource allocation when markets do not work well in some ways. The boundary between firms and markets changes with the times, as technologies and capabilities evolve. Some firms may swallow their suppliers, others may outsource. Firms use hierarchies to organize decision making, and hierarchies need managers. In my last column (19 October), I noted that markets are also not abstractions, and may require rules, monitors and enforcers to function well. EBay manages its vast electronic marketplace, and various managers in eBay oversee the different components of that market management. In a more traditional manufacturing firm, workers make physical products, and managers organize and oversee those production activities.

The idea of management as a science has been around for over a century, and found expression in time-and-motion studies, process certifications and more broadly in the MBA degree and management consulting profession. Recently, academics such as Nick Bloom have been quantifying management quality through indices of adherence to good practices, and measuring management quality in many firms in several countries. Management quality turns out to vary a lot within and across countries, and is positively correlated with productivity and profits (“Does management matter?" Mint, 11 August 2008). Can one do better than measuring a correlation, though?

Bloom and a set of co-authors have begun tackling that issue as well. The team of academics selected 20 textile plants near Mumbai. These were units in large multi-plant firms with around 300 employees, so not insignificant operations. The plants were randomly assigned to a treatment group (14 of them) and a control group (the other six). They all received free diagnostic consulting from a multinational consulting firm, but the treatment group also received intensive advice on implementation to fix shortcomings in operations. The focus was mostly on the basics of operations, such as the organization of the factory floor, how parts were stored or moved around, how inventories were logged and stored, how machinery was maintained, and so on. Thus, not all aspects of management were addressed, but certainly ones that are important yet basic, and perhaps too obvious to need an MBA, though some improvements required more than just common sense.

Unsurprisingly, the treatment plants were quick to adopt the recommended improvements in operational practices. For example, the system of recording weaving defects was improved to allow analysis at newly instituted daily production meetings. The number of defects in the treatment plants fell by close to half, while there was no change in the control group. Better organization and logging of inventory led to inventories falling in the treatment group, while they stayed the same in the control group. Efficiency as measured by percentage of running time seemed to go up in the treatment group after routine maintenance was introduced. Productivity overall went up by 5%, and profits by $130,000 per firm on average, for very small investments.

So management matters, very emphatically, with a clear causality established from improved basic management practices to improved operational and financial performance. This striking experiment raises another question, though, one that is puzzling. Why did these firms not make the improvements before? Financial constraints were not an issue, nor were the poor external infrastructure or regulatory environment, since those did not change during the experiment.

The study’s authors suggest some alternative answers: firms might not be aware of best practices, they are wary of affordable local consultants who might expose regulatory violations, and international consulting firms are too costly. But the improvements did not seem as if they really needed expensive specialist advice. That leaves incentives: Managers in the firms did not want to bother.

The way I would put it is that the management failure can be traced to two poorly functioning markets. One is obviously the product market—there are enough constraints on competition that inefficiencies can survive for a long time. The second deficiency is in the market for managers. A well-functioning market for managerial talent would allocate good managers to plants that are functioning below potential. The two market inefficiencies are connected, in that lack of competitive pressure in the output market reduces the need for a better functioning management market as well.

The inefficiency of these textile firms thus starkly illustrates how far India has yet to go in microeconomic reforms that create competitive markets for products and for talent. Expensive degrees or consultants may be of value in other cases, but so much of Indian industry is like these textile firms, getting by without needing to make obvious and inexpensive improvements in operations. Firms and managers need the discipline of competitive markets to do their best.

Nirvikar Singh is professor of economics at the University of California, Santa Cruz. Your comments are welcome at