Thomas W. Malone
Are you stifling innovation and creativity by trying to micromanage? Or are you operating your organization as many autonomous fiefdoms and missing the benefits of being one company? Should you give more autonomy to the people who work for you? Or perhaps you feel you should take more control and show "real" leadership?
Nagging questions like these indicate that some of the most difficult problems for managers are those of exercising control. A central issue for organizations in the twenty-first century will be how to balance top-down control with bottom-up empowerment. For example, recent business rhetoric has focused so much on the importance of "empowering" workers that the term has become an almost meaningless clich . Is the talk of empowerment just a fad? Or are fundamental changes making decentralized control increasingly desirable?
Our research suggests that the dramatically decreasing costs of information technology (IT) are changing the economics of organizational decision making, with the result that decentralized control is becoming more desirable in many situations. Moreover, our very notions of centralization and decentralization may be incomplete. When most people talk about decentralized organizations and empowerment, they mean relatively timid shifts of power within a fairly conventional, hierarchical structure. But these forms of empowerment go only halfway toward what is possible. To fully exploit the possibilities of new information technologies, we need to expand our thinking and see radically decentralized organizations—the Internet, all kinds of markets, and scientific communities, for example—as new models for organizing work in the twenty-first century.
Our research also suggests that a simple pattern underlies many future changes. As improvements in technology reduce communication and coordination costs, the most desirable way to make decisions moves through three stages. In the first stage, when communication costs are high, the best way to make decisions is via independent, decentralized decision makers. In the course of history, most economic decisions have been made this way—by people in largely independent tribes, villages, and towns.
As communication costs fall, however, it becomes desirable in many situations to bring remote information together, where centralized decision makers can have a broad perspective and therefore make better decisions than isolated, local decision makers can. The economic history of the twentieth century has been largely the story of this centralizing of decision making in large, global corporations. And, for many kinds of decisions, companies can still derive substantial benefits from centralization.
As communication costs continue to fall, however, there comes a point in many decision-making situations at which connected, decentralized decision makers are more effective. These decision makers can combine the best information from anywhere in the world with their own local knowledge, energy, and creativity. As the economy becomes increasingly based on knowledge work and creative innovation, and as new technologies make it possible to connect decentralized decision makers on a bigger scale than ever before, exploiting such opportunities for empowerment will surely be an important theme in the economic history of the next century.
Of course, many factors other than communication costs affect centralization and decentralization in organizations. Patterns of interpersonal trust, locations of decision-relevant information, personal motivations, prior distributions of power within the organization, government regulations, national cultures, organizational traditions, and individual personalities are all important. In fact, in any given situation at any given point in time, combinations of these other factors can be much more important than communication costs in determining where decisions are made. My goal in this article is not to analyze the complex question of how all these factors interact in particular situations. Instead I will focus on a simpler question: What is the relationship between reducing communication costs over time and the economics of different decision-making structures?
Understanding this relationship is important for three reasons. First, it helps us understand conceptually the economic effects of reduced communication costs if all other factors remain constant. Second, it provides a possible explanation for a variety of well-known facts, such as broad historical trends in organizational structures during the past century. Finally, to the degree you believe that reduced communication costs enabled by IT are likely to be important in the future, this work suggests an effect those changes are likely to have. Whether this factor actually turns out to be important is uncertain, of course. But if relentless improvements in IT continue to reduce communication costs by a factor of ten every few years—as they have been doing—shifts toward more decentralized empowerment are likely to continue.
See, for example, Johansen, Saveri, and Schmid 1995.
See, for example, DiMaggio and Powell 1983; Galbraith 1991; Huber and McDaniel 1986; Markus 1983; Schein 1985; Scott 1992; Thompson 1967.
Our model is particularly intriguing in this regard because, unlike previous models (for example, Gurbaxani and Whang 1991), ours shows how a simple model can explain changes in both directions while nevertheless predicting a broad change in one direction in the long run.