Futures research: Did it meet its promise? Can it meet its promise?

Futures research: Did it meet its promise? Can it meet its promise?

TECHNOLOGICAL FORECASTING AND SOCIAL CHANGE 36, 21-26 (1989) Futures Research: Did It Meet Its Promise? Can It Meet Its Promise? THEODORE J. GO...

496KB Sizes 3 Downloads 22 Views





36, 21-26


Futures Research: Did It Meet Its Promise? Can It Meet Its Promise? THEODORE


Introduction Go back with me for a moment to 1963, the year in which Dr. Olaf Helmer and I were working on the first large-scale Delphi inquiry at Rand. If you had asked me then, “What is the name of this field?“, I probably would have answered, “Futures research.” But if you had also asked, “What do you think it will accomplish?“, I would have been less articulate. I probably would have provided a list of half-formed ideas, most of these naive in retrospect but nevertheless indicative of the hope that thinking about the future would be seen as a responsible and necessary enterprise. Most of us tend to have selective memories of the past, but I think that, if pressed, I might have said that this field should yield: 0 A new profession, not necessarily a scientific discipline, but at least a systematic social science with standards of excellence and means of accreting knowledge. 0 New tools for improved forecasting of future developments, not only in “hard” domains such as technology and economics, but also in “soft” areas such as social behavior. These tools would be used in decision making everywhere, particularly in business and government planning. 0 Greater understanding of the limits of knowledge about the future, and the creation of means for dealing with uncertainty. 0 A sense of awareness about the future; a recognition that our current actions or inactions help-directly or indirectly-shape the future world that will contain us all. So how have we done in the past two and one half decades in the context of this retrospectively recognized set of goals? To a greater or lesser extent all have been achieved; to a greater or lesser extent all have been frustrated.

Is It a Rofession? Thinking about the future is certainly more respectable than it used to be. I can remember the time 17 years ago when my youngest daughter, freshly enrolled in the sixth

THEODORE J. GORDON is chairman of The Futures Group, Glastonbury, CT 06033. Address reprint requests to Theodore J. Gordon, The Futures Group, 76 Eastern Boulevard, CT 06033-1264. 0 1989 by Elsevier Science Publishing

Co., Inc.





grade, asked me what 1 did for a living. It was traumatic, but I answered, “Futurist.” Today, the trauma would be considerably reduced; it is less odd than it used to be. But “futurist” is not the name of a profession, nor is “futurism” or “futurology” a discipline. To be a profession, the field would need standards, tools and skills shared by all in the profession. To be a discipline, the field would require a foundation of generally agreedto “truths” which further research could test and to which new discoveries and data could be added. But there are yet few “truths” about the future-few places where understanding exists about what makes change occur. To use Kuhn’s framework, the field has remained largely in a pre-paradigm state. Have any paradigms been proposed‘? Yes, a few, but whether these are sufficient foundation still remains to be seen. The following are some paradigms that may provide a setting for the held. 1. The future is not preordained and can be shaped by the action of individuals, institutions and natural forces. This first paradigm suggests that actions change the future. There is a future without action, and a different one with it. Thus, futures research and predestination are, at least on the surface, antithetical. 2. Any decision or action has a number of possible outcomes; these outcomes may be affected by forces beyond the control of the decision maker or actor. This second paradigm, it seems to me, is already well rooted in the policy and management sciences. Control has its limits; any given action may have many different outcomes-some unpredictably affected by outside forces. 3. Almost all actions initiate chains of consequences; high-order consequences may be more significant, ultimately, than the primarily intended consequence. This third paradigm is the basis for “technology assessment” (TA), the study of higherorder consequences of action. When it was introduced, TA was a new way of looking at things that required analysis of layers of impacts: the people and institutions affected by initially unexpected interactions along the way to an objective. This arena has lost some of its gloss, but it is now important to policy analysis. It is common today to search a priori for unintended consequences. Suggest a policy and you are likely to hear: “Sure. it’s supposed to do great things-but what else will it bring’?” 4. It is possible to distinguish likely futures from desirable futures. This fourth paradigm recognizes that there is a difference between the probability of a future event and its intrinsic desirability. We have found that people can provide separate and distinct judgments about each of the attributes of a future event; inevitably these judgments interact. 5. Experts are more correct in their judgment about what is likely to occur than are nonexperts. There is only weak evidence that supports this fifth paradigm, yet it seems that some people can be more correct in judgments about the future than others; that there are methods for encouraging more accurate judgments of individuals and groups; that groups of experts are usually more accurate than randomly chosen experts; that a person who has demonstrated accuracy in the past is likely to be accurate in the future, at least on a similar topic. 6. Future developments, both physical and societal, often interact systemically. By understanding the system that contains them, the initial system conditions, and system perturbations, system outcomes can be forecast. This sixth paradigm is now being elaborated in very important ways through techniques of chaos analysis. When nonlinear systems are in the range of chaotic behavior, very small changes in initial conditions produce essentially unpredictable changes in out-



come. By adding nonlinear systems analysis to more conventional systems analysis, the universe of problems amenable to solution has been expanded greatly. Taken together, these paradigms have not proved sufficient to create a profession or a discipline. As the field emerged there were many trappings of a discipline-at least a discipline in search of itself. There were journals, some of them juried; futures curricula were established at respected universities; and global institutions devoted to the study of the future were created. But futures research is-perhaps will always be-less than a discipline. The body of knowledge on which a profession can be based is still absent. Nevertheless, professional behavior is still possible and is practiced by many. This behavior includes: l Working ethically in the interest of the client and guarding proprietary findings. 0 Stating and testing assumptions in any policy analysis or forecasting activity. 0 Making processes clear, i.e., avoiding “black boxes” for analysis. 0 Validating results whenever practical, including the search for explanations of past errors. 0 Utilizing appropriate analysis techniques, not rediscovering the wheel. 0 Recognizing that there is no “correct” or “best” future or universal truth with respect to “what must be,” unless the values underlying the judgments are also stated.

Do the New Tools Work? In 1964, the tools of the trade included Delphi-method, simple modeling borrowed from operations research, and simulation gaming. These three “seeds” gave rise to the main branches of futures research techniques: 0 Forecasting Forecasting l Forecasting l

through judgment through modeling through simulation.

While there have been a number of interesting attempts to improve forecasting through judgment, progress has been slow. The central issue is still selecting the individuals whose judgment will be admitted. How can experts, or persons with special knowledge about what developments might take place, be recognized a priori? The approaches to this question have been varied, but recognizing who is likely to be right or wrong in judgments about the future is still a primitive art. Nevertheless, some techniques have evolved that add richness to the process and stimulate imaginations in the pursuit of future-oriented judgments. On-line Delphis, dynamic panel membership and qualifications of panelists by objective means serve as examples. In addition, at The Futures Group, we have found that in-depth, person-to-person interviews with experts are very productive. These interviews utilize the old Delphi principles of anonymity and feedback, but they have several advantages over the old questionnaires. In interviews, the interviewee can tell the interviewer that the wrong question is being asked; this is difficult when questionnaires are used as the medium. The interview process is open ended and thus entirely new avenues can be opened by the interviewee. In addition, in face-to-face contact, interviewees often produce data to support their views, and bring knowledgeable colleagues into the conversation. Modeling activities have blossomed, both as a result of the ubiquity of personal




computers and the evolution of improved modeling techniques. Without cheap computer power, Monte Carlo techniques would have remained very expensive; computer-based experimentation would have been very limited; and nonlinear techniques such as chaos theory would probably not have evolved at all. Perhaps the most important change that has occurred in the domain of modeling is the shift in perception-both among modelers and their audiences--that allow us now to view models as no more than stereotypes or sketches of reality. In the old days, it was not unusual for a modeler to believe naively that the model somehow captured all of reality; today, most people in the field would have a much more modest expectation. Econometric models, despite their complexity, let us down with fair regularity; systems dynamics does not lead to truth-through-feedback; Kondratieb waves damp or resonate depending on one’s viewpoint. Yet modeling has proved to be an excellent means for illuminating interactions and testing assumptions and hypotheses against reality. Models have also proved to be an efficient and provocative way to quantify judgments that would otherwise be vague. (As current work in chaos theory indicates, modeling has a long way to go but the promise is still as bright as ever.) It is difficult to overemphasize the potential importance of simulation. At one time simulation was primarily gaming, each individual taking a role within some stylized set of rules of behavior. There are a number of convergent developments and trends which suggest that simulation can become one of the most important tools of futures research. Computers are, of course, becoming more powerful and less expensive. As noted earlier. modeling is improving. Computer-generated graphics are becoming so good that it is increasingly difficult to distinguish between true and artificially constructed images. Add artificial intelligence and it is not difficult to imagine very advanced simulation systems that permit the creation of accurate environments that stress the student and promote learning. With such software. the emphasis in education may switch from teaching to learning. Pilot training serves ah a current example. In the future. similar simulations will probably be used in business for training of production workers, managers, salespeople, repairmen, and anyone likely to benefit from the stress of practice. In this context. simulation also means that the user can determine the plot of an unfolding story through his or her decisions. Such simulation is. in effect. a means for experimenting with the future and learning in the process. Do We Understand More about the Limits of Knowledge about the Future? I propose that there are two kinds of unknown futures. One is the unknown but discoverable future, available for analysis through appropriate research. The other is the intrinsically unknowable future which is not accessible by any means. We have not. in any real sense, come to grips with this issue over the last 20 years. We know that trend extrapolation can be misleading when discontinuous developments impact those trends; we know that even long-established systems can change their structural elements as a result of an external shock. We have anecdotes about incorrect forecasts (e.g., a nuclear bomb will never be small enough to carry in a missile), and abundant examples of overconservatism and “blind spots” in forecasting future technological, social or political developments. In short, when it comes to discontinuities we cannot. with certainty, distinguish between a ridiculous forecast and a likely forecast. and this inability grows as we move farther out in time. Consider the following current examples which are not forecasts but examples of statements about future technologies that may be ridiculous or likely within the time frame of the next 50 years.



0 The field of nemotransmitters leads to fundamental understanding of the functioning of the mind (as opposed to brain). 0 Deeper understanding of the brain/mind interface leads to creation of specific and effective memory recall drugs-essentially anything heard or read within the last year can be retrieved by an individual. 0 Memory is found to be chemical in nature. 0 Memory can be transmitted person to person by chemical or other means. l Deeper understanding of the mind leads to a plausible explanation for telepathy or other ESP phenomena.

I would guess that 50 years from now, none of these statements will be seen as having been true, i.e., having occurred. Yet, I would not like to bet very much on it. The principal problem is that while I can frame such statements, I have no way of judging their accuracy. We can be relatively certain, however, that the field of neuropsychology will contain staggering surprises about which we cannot even guess. At a further level of abstraction, there are almost certainly fields like neuropsychology that have not even been named yet today. While not much progress has been made in defining the boundaries of the knowable future, great progress has been made in dealing with the uncertainty intrinsic in forecasting. Risk analysis, scenarios, and portfolio theory are all examples of analytic tools designed to facilitate decision making in the presence of uncertainty. Risk analysis examines the upside and downside potential of a given decision, usually probabilistically. Scenarios create images of a future “space”; the effectiveness of potential policies can be tested within this space. Portfolios can be constructed in such a way as to reduce the impact of uncertainties. As an example of methods for dealing with uncertainty, consider a recent study performed by The Futures Group for a client. In this instance, the client intended to introduce a large number of products two years hence. These products were, for the most part, to be bought from Far Eastern sources. There was a probability distribution associated with the level of sales of each product; the distribution was not normal in a statistical sense. Given the fact that excess inventory has costs that change with time, that ordering too few of a given product has costs as well (customer alienation), and that the costs associated with changing the order in midstream are also statistically definable, we created a model that made the uncertainties explicit and showed, product by product, the optimum order size required to achieve maximum profits. At The Futures Group, we find that decision makers are, in general, much more comfortable in dealing with uncertainty than they used to be. We commonly produce ranges of forecasts rather than single values whether we are dealing with forecasts of specific technologies or corporate return on investment. Furthermore, we seek to identify those future developments that will, if they occur, cause long-established trends to deviate or systems to change in structure. Since these future events are usually stated in terms of probability over time, their use implies non-deterministic forecasting that quantifies levels of uncertainty and can feed directly into risk analysis. Is There an Increased Sense of Awareness of the Future? Maybe it is because we are approaching the year 2000, that watershed of forecasts, but it seems to me that people today are more inclined to think about the future than they used to be. For example, ex-President Nixon writes a new book entitled, 1994. Also,



presidential candidates describe their images of the country to which their policies may lead. When we talk about environmental threats, we see them in terms of their future consequences: greenhouse, ozone, and solid waste disposal serve as examples. The public health threat of AIDS is seen not only in terms of its current situation, but also its future human and societal consequences. Consumer demography and economics are now “given” elements of marketing research. Issues management, global, corporate or personal, involves understanding how current trends and future developments may affect us. The Christian Science Monitor writes on goals to 2000. Ask any bright fifth grader about the future of energy or space travel and they are apt to have an answer. As to the future of futures research, I think it must lie in one of two directions. It may yet emerge to stand on its own as a semi-discipline, or it may blend in seamlessly with kindred fields, its lessons and concepts flowing into marketing research, the social, management and policy sciences, economics, and operations research. Which direction futures research takes is ultimately unimportant. What is important and hopefully lasting is its legacy: 1) it is worthwhile thinking about the future, 2) it is possible to think systematically about the future, and 3) without thinking about the future, we abandon our destinies to chance and the decisions of others.