The question, then, is not whether we should forecast but rather whether we should articulate, discuss, analyze, and try to improve our forecasts. The premise of explicit forecasting is that by moving beyond our ordinarily unarticulated assumptions about the future, we can better guide our current decisions to achieve a more desirable future state of affairs. It is a matter of whether we will go into the future with our eyes and our minds open or stumble into it with them closed.

The process of forecasting and developing information about the future raises several fundamental problems. We know that we can know nothing with absolute certainty about the future. But we can know, in a weaker sense, a great many useful things about the future: when a contract is scheduled to expire, when an election is expected to be held, when a machine is likely to be replaced, when a new technology is likely to come on line, when the life of everyone who is over 40 today is likely to have ended, how purchasing patterns are likely to change if present social and economic trends continue, and so on. Knowledge in this weaker sense implies a possibility--sometimes high, sometimes low--that something may come along to upset our otherwise secure understanding of the future. When we grant the presence of uncertainty, however slight, then we are in the realm of forecasting. Forecasting is valuable and important even when we have less confidence. Indeed, it is often far more important and/or valuable to focus on those areas where our confidence is low and where uncertainty--and the likelihood of upset--are high.

Highly likely and highly unlikely events are not usually very interesting, unless, of course, they fail to turn out as expected. The "interesting future" includes such possibilities, but it also (and more importantly) includes developments of middling probability whose occurrence or nonoccurrence will surely be decided within the period of interest as a result of decisions and policies implemented between now and then. In a society and a culture based on the scientific process of experimenting to develop and prove what is or can be known about nature, the process of thinking about the future and making forecasts stands alone in its vulnerability. Yet in spite of these risks and the unpredictability of the future, we all constantly make assumptions about the future in guiding our everyday decisions. Occasionally, of course, our assumptions are wrong, and we are surprised by sudden opportunities or developments that create both pain and loss. Nevertheless, as long as the future remains unpredictable, we have no choice but to go on making the best, most reliable assumptions and forecasts about the future we can.

Forecasting and the study of the future raise another major problem. Individual reputations, especially in the academic world, are built on research that follows established rules and procedures, research that should lead to the same results, regardless of the individual conducting the research. Information about the future, however, is based on assumptions about which reasonable persons can and do differ. The information we generate about the future is fundamentally linked to our personal values, concepts, ideas, experience, outlook, and makeup. While forecasting and futures research have to some extent borrowed detailed research methods and scientific concepts from other disciplines, these procedures cannot change the fundamental nature of information about the future. In the end, this information is based upon subjective judgment. As a result, for people whose lives, reputations, and careers rest on successful adherence to the traditional established rules and procedures of research, making forecasts is uncomfortable, even threatening. To speculate about the unknown and the unknowable is to challenge one of the keys to their success-careful use of "proper" research methods. By recognized standards of every professional discipline, even the best information about the future is unacceptable, because its foundation is subjective judgment and thus it cannot meet traditional scientific standards of objectivity, experimental verification, reproducibility, and so on. All information about the future may be judged inadequate in this light, but information about the future-any information about the future-is better than no information about the future.

What is good information about the future? Simply put, it is information that helps us to improve our current performance so that we can achieve a better future than would otherwise occur. Thus, total accuracy in the forecast cannot be the goal. By the time we know that the information about the future is correct, it must be too late to do anything about it. For example, if an air traffic controller watching two planes on a radar screen develops a forecast that the planes are likely to collide, we must ask, "What should we do with this forecast?" We can wait, watching the radar screen to see whether the planes do in fact collide and thus confirm the accuracy of our forecast. But by the time we know that our forecast was correct, we have a catastrophe. The forecast has value only if we use that information to avoid the undesirable future of the forecast catastrophe by directing the aircraft to safer courses. This principle is just as valid for large complex social systems: What should we do now to avoid the catastrophe of bankruptcy for the social security system 25 or 40 years hence? What should we do now to avoid the catastrophe of certain present trends in elementary and secondary education?

A forecast can be a failure even if it turns out to be accurate. The Paley Commission appointed by President Eisenhower in 1952 to study the long-term energy circumstances of the United States generated such a forecast. The forecast that it produced-that the country faced an energy crisis in the mid-1970s-turned out to be quite accurate. The forecast was a failure because it was not used to avoid that crisis. Thus, the key criterion must be that the forecast is used-used to create a better future. To be used, the forecast must be communicated to the relevant decision makers, and they must believe the forecast and have the resources to act on that information. Naturally, the accu racy of previous forecasts derived from similar methods or produced by similar forecasters or forecasting groups would enhance credibility; that is, decision makers will be more likely to assume a method (or forecast) is credible if it has been accurate in the past.

Most of the forecasting we do is implicit--unarticulated--and can appropriately stay that way. Some of the forecasting, however, should be articulated, discussed, debated, evaluated, challenged, changed, modified, and used as we make decisions in an effort to achieve more desirable futures. Forecasting may be both inadequate and better than ignorance of possible futures. The curse of Cassandra was to know the future but be powerless to change it. Forecasting gives us our best information about the future, but we will never know the future nor the curse of Cassandra--for we have the power to change the future.