Technologies Always Have Side Effects

in addition to its intended benefits, every design is likely to have unintended side effects in its production and application. On the one hand, there may be unexpected benefits. For example, working conditions may become safer when materials are molded rather than stamped, and materials designed for space satellites may prove useful in consumer products. On the other hand, substances or processes involved in production may harm production workers or the public in general; for example, sitting in front of a computer may strain the user's eyes and lead to isolation from other workers. And jobs may be affected—by increasing employment for people involved in the new technology, decreasing employment for others involved in the old technology, and changing the nature of the work people must do in their jobs.

It is not only large technologies—nuclear reactors or agriculture—that are prone to side effects, but also the small, everyday ones. The effects of ordinary technologies may be individually small but collectively significant. Refrigerators, for example, have had a predictably favorable impact on diet and on food distribution systems. Because there are so many refrigerators, however, the tiny leakage of a gas used in their cooling systems may have substantial adverse effects on the earth's atmosphere.

Some side effects are unexpected because of a lack of interest or resources to predict them. But many are not predictable even in principle because of the sheer complexity of technological systems and the inventiveness of people in finding new applications. Some unexpected side effects may turn out to be ethically, aesthetically, or economically unacceptable to a substantial fraction of the population, resulting in conflict between groups in the community. To minimize such side effects, planners are turning to systematic risk analysis. For example, many communities require by law that environmental impact studies be made before they will consider giving approval for the introduction of a new hospital, factory, highway, waste-disposal system, shopping mall, or other structure.

Risk analysis, however, can be complicated. Because the risk associated with a particular course of action can never be reduced to zero, acceptability may have to be determined by comparison to the risks of alternative courses of action, or to other, more familiar risks. People's psychological reactions to risk do not necessarily match straightforward mathematical models of benefits and costs. People tend to perceive a risk as higher if they have no control over it (smog versus smoking) or if the bad events tend to come in dreadful peaks (many deaths at once in an airplane crash versus only a few at a time in car crashes). Personal interpretation of risks can be strongly influenced by how the risk is stated—for example, comparing the probability of dying versus the probability of surviving, the dreaded risks versus the readily acceptable risks, the total costs versus the costs per person per day, or the actual number of people affected versus the proportion of affected people.