Technological concentrations

John Sutton's Sunk Costs and Market Structure had a profound effect on industrial economics. In that work, he divided "sunk costs" into exogenous and endogenous, arguing that high sunk costs, particularly endogenous sunk costs, would lead to a relatively concentrated industry. Sutton, a game theorist, departed from the norm of theorists by proceeding to empirical studies of the European food-processing industry, in which he focused on advertising expenses as the main endogenous sunk cost.

Technology and Market Structure extends his approach into the more complex area of technology. The book is a joy to read for its penetrating lucidity and is a model for integrating theoretical work of the highest quality with applied studies of particular technologies. But it creates some problems as the author tries to match neat theories with messy empirical evidence.

Sutton defines technological determinants of industrial concentration using a parameter "alpha". Alpha is the power to leverage research and development - the extent to which a firm can benefit from its R&D expenditures. Alpha is, however, only partially related to R&D expenditures, so high spenders such as pharmaceuticals get classified as low-alpha industries. The chief problem is that alpha is not directly measurable, and this troubles Sutton more than many theorists because of his commitment to empirical testing.

The difficulty in distinguishing between products and technologies is the main problem. On the surface, this is just the age-old economist's distinction between the supply side (technologies) and the demand side (products). However, one can argue that industry boundaries become hard to define because the "breaks" between technologies and those between products do not match. That is, a firm producing products in closely related market segments may deploy technologies from a diverse range of technical fields.

Re-examined in this light, some of Sutton's "industries" turn out to be specific products embodying alternative technologies. One of his favourite examples is that of industrial flowmeters, which produce a single product but may use nine different sorts of technologies (electromagnetic, ultrasonic, etc). The firms that Sutton considers are concerned about rival producers of flow-meters, but in practice firms may be more concerned about rivals using the same sort of technology (eg ultrasonic) for other products.

In our view, it is the relative ability to shift between technologies or between products that is a key determinant of concentration. A classic case is the machine tools industry, which has been dominated since its inception two centuries ago by small firms. The reason is that machine tools perform only a limited range of functions, but these functions are appropriate for the machinery built for a wide range of manufacturing industries. It may be easier for the firm producing a turret lathe for the sewing-machine industry to start producing turret lathes for the bicycle industry than, say, to start to produce grinding tools for the sewing-machine industry. The firm's core competence (and comparative threat) is the technology rather than the product.

Finally, the technology-product relationships lead us to the role of the firm. The existence of the firm derives from its unique role in mapping knowledge about technologies into knowledge about products. Sutton has

produced a theory bound to

concentration of firms in specific products that will be a landmark in the subject. Further work is, however, required to understand the concentration of technologies, and concentration of firms at large.

Qing Wang is lecturer in technology and innovation management and Nick von Tunzelmann is professor of the economics of science and technology, SPRU, University of Sussex.