Finally, Braudel discovered a third level of history-one derived and focused on individual attitudes, values, and beliefs. He argued that these forces bring institutional and structural changes, which in turn bring the changes in the surface indicators. These three levels-superficial, structural/historical, and attitudinal-can be recursive; that is, surface changes "cause" structural changes, which cause changes in values, which then cause changes on the other levels. Consider, for example, how reliable, inexpensive birth control affected society's sexual values and how these value changes affected school enrollments.

These three levels of history can also serve as a means for viewing the future. Early forecasting and futures research methods focused on indicators and measures of surface change, such as unemployment, inflation, economic growth, sales of automobiles, and housing starts. Most of our discussions of the future still focus on these surface indicators, usually using implicit forecasts. For example, on the first page of the Wall Street Journal every day is a historical chart of some surface indicator from which readers can derive their own forecast. This serves to establish the unarticulated, implicit forecast.

Given this view of the future, it is not surprising futures research and forecasting developed and used a host of extrapolative methods in support of long-range planning focused on surface indicators. These methods included, among others, regression analysis, mathematical and judgmental trend extrapolation, Box-Jenkins, and rolling averages. The focus of this traditional long-range planning was internal, based on tracking and forecasting an organization's or a system's internal development.

As surprise developments continued to upset the forecasts and long-range plans produced by these methods, forecasters modified them to take explicit account of unprecedented, new developments and external "surprises." While many of them were surprises on the surface, futures researchers began to increasingly include changing structural and institutional developments. Thus, trend extrapolation was modified to become trend-impact analysis, a probabilistic forecasting technique in which an extrapolative trend is modified by the occurrence of hypothesized external surprise events.

At the same time, models of large systems created through the method known as systems dynamics were modified to include similar external surprises; with this modification, it became known as probabilistic systems dynamics. Relationships among external surprise events were also explored through such methods as cross-impact analysis. Most recently, in the mid-1970s, a method for guiding the development of responses to information about the future on the first two levels was developed--policy impact analysis. While a small scan of the external environment usually was made to identify candidate surprise events, the focus remained on the trends and issues developed from the traditional, internal perspective.

Today, methods are used that will take account of information from the third level of the future--changing values and attitudes. But these methods are in their infancy in this sort of application, and their users frequently have no clear idea of what balance should be struck among potential developments on all three levels. Coping seriously with this question is one of the frontiers of futures research. While we could of course approach this question through opinion research, opinion research has been notoriously weak in forecasting for the simple reason that individual opinion about the future is much more unreliable than opinion about the present. But the real difficulty with polling is not that individuals are unreliable; it is that pollsters who have tried forecasting have aimed at predictive accuracy with "one-shot" estimates. That is, they failed to provide respondents feedback on the original judgment, thereby prohibiting their using that information to reflect on the initial judgments or the arguments behind them. Consequently, the polling technique is limited in obtaining that information useful for managing change and increasing our options.

As futures research and forecasting methods became less quantitative and more qualitative, the process of exploring and responding to changing social structure and values (the second and third levels of the future) was becoming more and more important. The traditional "fire fighting" mode of public relations was modified to include a planning or forecasting role in what became known as public affairs. This role was extended to include policy planning. Thus, it is not unusual to find on a contemporary corporate organizational chart a director of public policy planning, a position stationed philosophically somewhere between the traditional public relations/government relations/public affairs function and the corporate planning function. In some organizations, the policy planning function has been extended to include a director of public issues, who is responsible for anticipating and helping the organization to respond to emerging issues in the external environment. The new field of issues management emerged as public relations and public affairs officers recognized the need to expand the use of forecasting and futures research in their planning and analysis of policy. The merger of futures research concepts and techniques into public policy planning holds the prospect that administrators will eventually pay much closer attention to detailed information about the future on all three levels discussed earlier.

Currently the most common technique used for including more information about the external world from both the second and third levels (structural/institutional change and attitudinal change) is environmental scanning. As with earlier forecasting, environmental scanning focuses first on the surface indicators--derived from newspapers, literature, and periodicals--as signals of underlying change in the more difficult-to-understand structural and attitudinal levels. Although refinements in environmental scanning will have to be developed, this approach to long-range planning moves us into the strategic planning process.

FIGURE 2 ENVIRONMENTAL SCANNING

As noted earlier, one of the major limitations of the traditional long-range planning model is that information about the changing external environment is usually not taken into account systematically or comprehensively. When this omission occurs because of an assumption that "we cannot predict external changes," long-range planning destines itself to surprise and failure, if only because it locks itself to the information known from direct experience in the past and immediate present.

Information from the external environment adds important components to long-range planning, however. First, it identifies new and potentially crucial subjects that should be added to those identified and tracked during monitoring. Second, it identifies possible developments that must be used to adjust the forecasts of the internal issues derived from forecasting--specifically, the surprise events that are used in policy impact analysis or techniques like probabilistic systems dynamics and in other rigorous forecasting methods used in traditional long-range planning.

FIGURE 3 THE STRATEGIC PLANNING PROCESS

Source: Renfro and Morrison 1984.

The argument for combining these two models becomes apparent when the future that happens to the institution and the future that happens for the institution are contrasted. In the future that happens to the institution (the typical "planned" future), new developments are not anticipated before they force their way to the top of the agenda, demanding crisis management and the latest fire-fighting techniques. In this future, issues are usually defined by others whose interests do not necessarily include those of the institution or its purpose. Not only are threats from the external environment not anticipated as early as possible; key opportunities will be missed or diminished in value.

In the future that happens for the institution, in contrast (the "strategic" future), administrative leadership is focused more on fire prevention and less on fire fighting. Hence, it is able to exercise more careful judgment in the orderly and efficient allocation of resources. Certainly management will still have to deal with unforeseen developments, but they will probably be fewer and less traumatic. Thus, institutions will be able to pursue their mission with greater confidence and consistency because they will be interrupted by fewer and smaller fire-fighting exercises.