Nick Carr's essay blog
March 16, 2004
A short article by W. Brian Arthur, together with a rather longer interview with him, in the December 2003 issue of Optimize magazine has me wondering afresh at the tendency of many writers on business technology to ignore the little matter of competition. They look at IT's power to automate and coordinate business processes, often with considerable reductions in labor costs, and leap to the conclusion that the resulting productivity gains will be captured by companies as profits. Here's how Arthur, in describing IT's business impact, conflates productivity and profit: "The statistics show that most of the productivity growth in the U.S. economy is attributable to technology. Ordinary companies are connecting their business processes and seeing these changes affect their bottom lines."
In reality, of course, cost savings do not fall automatically (and certainly not permanently) to the bottom line. Competition gets in the way. Through the commonplace magic of best-practice replication, most discrete process improvements are made at more or less the same time by many different companies in an industry. Unprotected, the gains are then rapidly competed away - through price cuts, for instance - and they end up in the pockets of customers rather than falling to the bottom lines of individual companies. If productivity gains are shared rather than proprietary, in other words, they won't translate into sustainable profit increases.
The replication of process improvements happens particularly quickly when they are built on the application of maturing technologies like IT. IT's rapid performance advancement and equally rapid price deflation, combined with cut-throat competition among vendors and swift best-practice dissemination through consultants and practitioners, ensures that innovations diffuse quickly throughout industries and the entire economy. Industrywide productivity can increase smartly without any individual company reaping additional profits.
This is no mere academic distinction. To the extent that companies overlook the dynamic of competitive replication and assume that cost reductions will necessarily provide a sustained profit boost, they'll tend to make overoptimistic assumptions about the returns from IT investments. Often, they'll end up spending much more aggressively than they should.
The routine confusion of productivity growth and profit gains in the IT literature was one of the original spurs for the research that led to my HBR article "IT Doesn't Matter" (and to my forthcoming book Does IT Matter? Information Technology and the Corrosion of Competitive Advantage). What I set out to do in the article is show that technology can have a profound, even transformative, effect on economies, industries, and business processes without necessarily influencing the competitive fortunes of individual firms. Much of the critical response to the article, interestingly enough, has continued to reflect an inability to (or a conscious decision not to) distinguish the broad, macro effects of IT innovations from their more micro effects on individual companies. Arthur, who takes a potshot at my article in the Optimize interview, again provides a good case in point.
An economist who currently hangs his no doubt capacious hat at the Santa Fe Institute, Arthur is best known for his central, if controversial, role in promoting the idea of increasing returns. More recently, he has been studying the progress of technological revolutions and, in particular, their economic effects. He laid out his findings in an insightful article, "Is the Information Revolution Dead?," in Business 2.0 back in March 2002. Drawing on fascinating historical research by Carlota Perez (among others), Arthur argues that technological revolutions tend to occur in two phases. First comes a turbulent period of exploration when companies experiment with the new technology, and technical and usage standards begin to emerge. This period ends, typically, in a frenzy of speculation, as an investment bubble inflates, then pops. In the second, calmer phase, the new technological infrastructure becomes a fundamental and increasingly ordinary part of business operations, driving broad productivity gains and often forming the foundation for new consumer goods and services. Perez, in her book Technological Revolutions and Financial Capital, terms these two phases the "installation phase" and the "deployment phase" and notes that the important innovations in the deployment phase tend to be less technological than institutional.
I believe this schema is essentially correct. Moreover, I would suggest it dovetails neatly with my own analysis. It is in the installation phase, I would argue, that individual companies have the best chance of achieving competitive advantage through technological innovation. During this time, the technology is not yet part of the shared business infrastructure, standards are still embryonic, and distinctive applications of the technology can set one firm apart from the pack. In contrast, the essence of the deployment phase - the phase we are now in with IT - is that the technology has become part of the shared infrastructure. It can drive broad productivity gains and it can be a platform for product and service innovation, but it is no longer itself a basis for competitive distinction or superior profitability. One could say, in fact, that it is only when a technology ceases to be a source of strategic distinctiveness for individual firms that it begins to deliver its greatest benefits to economies and societies.
Or, to put it in terms that Arthur uses, the golden age begins when the revolution ends.
Copyright 2004 by Nicholas G. Carr. Digital Renderings is a trademark of the author.