to solve all of the problems. But an unruly tool in the hands of overconfident people is a recipe for disaster. It would be far more productive to assume that the Internet is highly unstable; that trying to rebuild one’s policies around a tool that is so complex and capricious is not going to work; and that instead of trying to solve what may essentially be unsolvable global problems, one would be well-advised to start on a somewhat smaller scale at which one could still grasp, if not fully master, the connections between the tool and its environment.
But such caution may suit only the intellectuals. Despite the inevitable uncertainty surrounding technology, policymakers need to make decisions, and technology plays a growing role in all of them. Predictions about how technology might work are thus inevitable, or paralysis would ensue. The best policymakers can do is to understand why so many people get them wrong so often and then try to create mechanisms and procedures that could effectively weed out excessive hype from the decision-making process.
The biggest problem with most predictions about technology is that they are invariably made based on how the world works today rather than on how it will work tomorrow. But the world, as we know, doesn’t stand still: Politics, economics, and culture constantly reshape the environment that technologies were supposed to transform, preferably in accordance with our predictions. Politics, economics, and culture also profoundly reshape technologies themselves. Some, like radio, become cheap and ubiquitous; others, like the airplane, become expensive and available only to a select few. Furthermore, as new technologies come along, some older ones become obsolete (fax machines) or find new uses (TVs as props for playing games on your Wii).
Paradoxically, technologies meant to alleviate a particular problem may actually make it worse. As Ruth Schwartz Cowan, a historian of science at the University of Pennsylvania, shows in her book More Work for Mother, after 1870 homemakers ended up working longer hours even though more and more household activities were mechanized. (Cowan notes that in 1950 the American housewife produced singlehandedly what her counterpart needed a staff of three or four to produce just a century earlier.) Who could have predicted that the development of “labor-saving devices” had the effect of increasing the burden of housework for most women?
Similarly, the introduction of computers into the workforce failed to produce expected productivity gains (Tetris was, perhaps, part of some secret Soviet plot to halt the capitalist economy). The Nobel Prize-winning economist Robert Solow quipped that “one can see the computer age everywhere but not in the productivity statistics!” Part of the problem in predicting the exact economic and social effects of a technology lies in the uncertainty associated with the scale on which such a technology would be used. The first automobiles were heralded as technologies that could make cities cleaner by liberating them of horse manure. The by-products of the internal combustion engine may be more palatable than manure, but given the ubiquity of automobiles in today’s world, they have solved one problem only by making another one—pollution—much worse. In other words, the future uses of a particular technology can often be described by that old adage “It’s the economy, stupid.”
William Galston, a former adviser to President Clinton and a scholar of public policy at the Brookings Institution, has offered a powerful example of how we tend to underestimate the power of economic forces in conditioning the social impact of technologies. Imagine, he says, a hypothetical academic conference about the social effects of television convened in the early 1950s. The consensus at the conference would almost certainly be that television was poised to strengthen community ties and multiply social capital. Television sets were sparse and expensive, and neighbors had to share and visit each other’s houses. Enter today’s academic conferences about television, and participants are likely to deplore the pervasive “bedroom culture,” whereby the availability of multiple televisions in just one home is perceived as eroding ties within families, not just ties within neighborhoods.
Another reason why the future of a given technology is so hard to predict is that the disappearance of one set of intermediaries is often accompanied by the emergence of other intermediaries. As James Carey, a media scholar at Columbia University, observed, “as one set of borders, one set of social structures is taken down, another set of borders is erected. It is easier for us to see the borders going down.” We rarely notice the new ones being created. In