As I’ve mentioned in previous posts, technology foresight started out as a method for facilitating long-term planning and policy primarily in relation to research & development (R&D) and innovation policy. In the last decade there has been and increasing focus on the usefulness of the methods used in R&D planning for other policy areas, especially social policy, i.e. education, labor, welfare, etc. The qualifier “technology” is often dropped from the term, referring to the exercises only as “foresight” because technology is not always a defining characteristic of the exercise although it could be argued that technology is an unavoidable factor in any long-term planning. This can cause some conceptual confusion since the generic term “foresight” has also often been used as a catch-all term for any sort of future-oriented policy or decision-making activity. Yet, I am going to venture into the conceptual quagmire and use the term “foresight” rather than “technology foresight” because I want to discuss the use of foresight activities in areas of policy-making that may not be directly related to technology, especially educational policy, and I think the use of the qualifier “technology” may cause confusion. Thus, what I mean by “foresight” is a deliberate and well-defined exercise intended to apply futures methodologies to social policy issues in order to inform long-term policy-making. The specific purpose of foresight activities is to identify and address elements of significant uncertainty that affect social policy and institutions. Foresight activities differ from specific futures methods, such as scenario construction, delphi studies and projections, in that they use a mix of the aforementioned methods, involve a broad range of stakeholder groups, including the general public, subject experts, policymakers, etc., and are specifically intended to facilitate long-term policy making in a range of areas by identifying possible future outcomes based on significant uncertainty factors.

As foresight exercises have transitioned from the somewhat confined realm of R&D policy, which, in the past, was often limited to specific industries or even specific organizations, to the broad realm of social policy, a gap has emerged in the foresight literature concerning the interface between foresight and policy-making. Most of the literature that I have discussed or mentioned in previous posts builds on Irvine & Martin’s seminal work focusing primarily on the structure and implementation of foresight exercises themselves with little attention to how the outcomes from such activities are translated into policy. Some recent research has sought to address this gap by examining closer how foresight functions as a policy-making instrument. Although this recent research has produced valuable results, what is still lacking is framing foresight as a policy-making instrument in terms of existing theoretical frameworks on policy-making processes. In this article I want to briefly address that gap and consider foresight from the perspective of a couple of existing theoretical frameworks, primarily Kingdon’s “multiple streams” model and Mazzoni’s “arenas” model.

Continue reading →

In previous posts I said that I would, at some point, start posting some things here regarding methods used in technology foresight. Here is the first of these.
It might make more sense to start with delphi methods, which really are the most commonly used in large-scale foresight programs, but delphi methods are complex and difficult to get right. In fact, many of the criticisms that were directed at early attempts at technology foresight that I have written about in previous posts really had to do with the lack of proper rigor and objectivity in delphi-type studies. Nevertheless, I’m going to save the delphi methods for later. I will start with scenario planning, which is a very commonly used method, and rigorous when done right, but can be equally suitable for small-scale projects as well as larger comprehensive foresight programs.
The aims of scenario planning are, basically, to construct plausible case descriptions for the future, based on certain variables that are expected to produce a significant amount of uncertainty for future planning. Thus scenario planning seeks to minimize uncertainty by extending developments identified in the present to the future. Most importantly, the purpose of scenario planning is not to present visions of preferred or less preferred futures, but merely to provide decision makers with reasonable expectations of what might transpire given certain types of development over the long-term. In most rigorous scenario planning exercises, there is, in fact, no attempt to evaluate the scenarios generated as that would affect the integrity of the exercise and increase the potential for bias.
This article provides a brief overview of the history of scenario planning, its current state, and methodologies commonly used in scenario planning exercises. I will follow-up in the near future with some scenarios that I have generated focusing on the development of information and communication technologies and its impact on education.
Scenario planning
Modern scenario planning emerged from Herman Kahn’s military planning work for the RAND corporation in the 1950s (Bradfield, Wright, Burt, Cairns & Van Der Heijden, 2005; Lindgren & Bandhold, 2003; Schwartz, 1996). Kahn used scenario planning to help the military prepare for the uncertainties of the cold war period, when high tensions between the U.S. and the Soviet Union fueled global political instability. Kahn’s scenario planning allowed the U.S. government to prepare for a range of possible situations at a time when appropriate immediate responses to unexpected situations could make the difference between cautious restraint and global nuclear war. In the 1970s, scenario planning spread beyond the military sector, to other public policy areas and private businesses. The Royal Dutch/Shell company was especially influential in adapting scenario planning to private sector needs and created a market for consultancy firms that facilitated the further spread of scenario planning in private and public sectors (Van Der Heijden, Bradfield, Burt, Cairns & Wright, 2002; Lindgren & Bandhold).
There is some disagreement as to how, precisely, scenarios should be defined. Lindgren and Bandhold (2003) list several scholars’ definitions of scenarios that, despite some subtle differences, reflect a general agreement that scenarios are not forecasts or projections of desired futures (Lindgren & Bandhold). Scenarios make use of trends and developments in the present to provide subjective descriptions of how these may play out in the medium and long-term future. Snoek (2003), citing Dammers, describes how scenarios differ from other future-oriented planning tools. Dammers (Snoek, 2003) describes a typology of future-oriented planning strategies based on the level of uncertainty inherent in the situation being analyzed as demonstrated by the number of available facts and theories pertaining to the situation. Scenarios are appropriate when there is a relatively high level of uncertainty, where facts pertaining to the situation are few and applicable theories are many. In this, they differ from more analytically oriented strategies, where analysts and decision makers can draw from numerous relevant facts to provide justified projections of futures, and from situations where few available facts and few applicable theories result in highly speculative visions of possible or desired futures.
Dammers’ (Snoek, 2003) typology provides a useful gauge for evaluating when scenarios are appropriate and determining their purpose. Firstly, scenarios are appropriate when a situation is directly or indirectly influenced by rapid or unforeseeable changes, and where conflicting approaches are brought to bear on the situation. Secondly, scenarios are intended to provide decision makers with viable descriptions of future situations, rising from this uncertainty, to highlight possible problems that can be addressed in the present. When scenarios are deemed appropriate, they are presented as a number of plausible futures, each dependent on possible courses of action, rather than a single anticipated or desired vision. Scenario planning and generation does not include evaluation of the scenarios. Rather, they are presented as equally viable consequences of decisions and developments over the medium and long term. As such, scenarios are intended to generate discussion rather than promoting particular courses of action (Van Der Heijden, Bradfield, Burt, Cairns & Wright, 2002).
Methodologies
Modern scenario planning has developed rapidly over the past few decades. Along with refinements regarding the definition of scenario planning, several methodologies pertaining to problem identification and scenario generation have been introduced and refined. Because scenarios have been applied to diverse areas of planning and decision-making, some methodologies have become highly specialized and may not be applicable to all situations. Bradfield et al. (2005) list the following areas in which scenario planning has been applied to illustrate the current diversity of the field (p. 796-797):

• crisis management: ex. civil defense exercises;
• the scientific community: ex. effectively communicating scientific models and theories pertaining to environmental change;
• public policy makers: ex. to involve multiple agencies and stakeholders in policy decisions;
• professional futurist institutes: ex. to communicate critical trends;
• educational institutions: ex. to create future learning environments;
• businesses: ex. for long range planning.

The diversity illustrated by the list above demonstrates the importance of choosing a methodology appropriate to the needs and unique circumstances surrounding the situation that it will be applied to. In this section we consider some commonly used scenario planning methodologies and how they apply to the situation under consideration.
Given the Rand Corporation’s historical role in the development of scenario planning, its influence on the scenario planning methodologies is not surprising. Early scenario planning within Rand was exclusively applied to military planning. At that time, Rand worked on a contract basis for U.S. defense branches. The contracting parties generally presented Rand analysts with potential crises, ex. wars or conflicts in specific areas or of a specific nature, and the role of the Rand analysts was to work backward to derive an explanation, or possible explanations, for the crisis as ordered (DeWeerd, 1973). It was not until the early 1970s that this backward approach was challenged by DeWeerd, working at Rand at the time. In a Rand Paper, DeWeerd described an alternative method that considered current contexts and evolving trends to describe potential future crises rather than backtracking from a hypothetical crisis to determine its causes. In the decades since then, Rand has developed highly quantitative methods using calculated probabilities to generate scenarios based on situations and emerging trends in the present (Camm & Hammitt, 1986). While this methodology provides a high level of reliability, the resource requirements are beyond the reach of many organizations that can benefit from scenario planning.
Modern scenario building is more oriented toward the forward-looking contextual approach described by DeWeerd (1973) than Rand’s earlier backtracking approach. It was precisely along these lines that Schwartz (1996) developed a qualitative methodology for scenario building for Royal Dutch/Shell in the early and mid 1970s. Schwartz especially emphasizes the identification of “driving forces” that are likely to produce change conditions. Relevant driving forces depend on the situation or issue being addressed, but are likely to include social, economic, political, environmental, and technological forces. The goal of identifying relevant driving forces is to observe and expand on emerging trends that are likely to affect the situation or issue being addressed. Finally, the most relevant driving forces and emerging trends are applied to the context of the issue to generate scenarios.
Schwartz’s (1996) qualitative approach, focusing on driving forces, forms the basis for most contemporary scenario building methodologies. However, some currently used methodologies differ depending on the perceived goal of the scenario building activity. For example, Lindgren and Bandhold’s (2003) and Van Der Heijden, et al.’s (2002) methodologies differ from Schwartz’s methodology in several respects, primarily because of the business-orientation of the authors’ scenario building activities. Lindgren and Bandhold’s strategic TAIDA framework, for example, goes beyond what is usually expected of scenario building exercises to include the generation of desired scenarios and an action component to consider short term goal setting in response to generated scenarios. Van Der Heijden, et al.’s STEEP framework also reflects a strong business orientation in the driving forces that are emphasized. These include economic factors such as, fiscal policies and taxation, and social factors such as, demographics and taste. Although the STEEP factors can certainly be applied to most situations in modern Western market-driven societies, they are likely to be beyond the scope of targeted scenario building exercises in many fields.
Of more general interest in Van Der Heijden et al.’s STEEP framework is the process of identifying and categorizing significant driving forces and “polar outcomes” (2002, n.p.). Polar outcomes, in this sense are not necessarily direct opposites, but should reflect the consequences of developments relevant to the scenario context. For example, the proliferation of wireless information technology may result in considerable costs for developed countries due to the need to replace existing infrastructure while it presents a cost-effective means for implementing IT in developing countries where no traditional infrastructure existed before. Thus, the outcome of widespread adoption of wireless IT may have very different consequences depending on the context in which it is adopted.
The goal of determining driving forces and polar outcomes is to identify areas of critical uncertainty. By categorizing forces and outcomes based on their predictability and potential impact on the situation of concern, we identify the potential developmental combinations that are likely to result in unexpected outcomes. It is these unexpected outcomes that are of primary interest for the scenario building exercise since they are least likely to be anticipated. Snoek (2003) employs this method in determining scenarios for the future development of teacher education in Europe. Snoek uses a two-dimensional matrix to evaluate possible scenarios based on two continuum variables derived from an analysis of relevant driving forces. The specific scenarios to be developed are based on points in the matrix. In Snoek’s case, these points simply represent the four quadrants of the two-dimensional matrix. In some cases, further analyses of specific points in the matrix may reveal variable combinations that are more interesting for the scenario building activity than others.
Example of a scenario decision matrix using two dimensions:

In deciding which scenarios are of interest, we would plot points on the above matrix corresponding to various compositions of the A & B dimensions. Usually, scenario planning exercises will generate a number of alternatives. For example, we might decide that realistic and interesting scenarios would be based on the extremes in each matrix compartment, i.e. high A/high B, low A/high B, high A/low B, and low A/low B. This would result in four distinct scenarios that describe circumstances given these compositions. Realistically, extremes are usually not likely, so the scenarios will be based on varying levels of each component. What points to designate for scenario generation is based on in-depth analysis of likely developments over the long-term.
Most scenario experts suggest that an even number of scenarios be constructed and that they correspond to varied levels of each dimension. This is to avoid perceptions that scenarios depict more or less preferred futures. For example, if we were to construct three scenarios based on high A/high B, moderate A/moderate B, and low A/low B, it is very easy to for the human mind to categorize these as representing a scale of preference, ex. highs are good, moderate is an acceptable compromise, and lows are bad. The scenario planner wants to avoid creating these kinds of perceptions because the intent is not to evaluate the scenarios. Rather, the intent is to provide fuel for constructive dialogue about how decisions in the present may affect the future.
References
Bradfield, R., Wright, G., Burt, G., Cairns, G., & Van Der Heijden, K. (2005). The origins and evolution of scenario techniques in long range business planning. Futures, 37, 795-812.
Camm, F., & Hammitt, J. K. (1986). An analytic method for constructing scenarios from a subjective joint probability distribution. Santa Monica, CA: The Rand Corporation.
DeWeerd, H. A. (1973). A contextual approach to scenario construction. Santa Monica, CA: The Rand Corporation.
Lindgren, M., & Bandhold, H. (2003). Scenario planning: The link between future and strategy. New York City: Palgrave Macmillan.
Schwartz, P. (1996). The art of the long view: Planning for the future in an uncertain world. New York City: Doubleday.
Snoek, M. (2003). The use and methodology of scenario making. European Journal of Teacher Education, 26(1), 9-19.
Van Der Heijden, K., Bradfield, R., Burt, G., Cairns, G., & Wright, G. (2002). The sixth sense: Accelerating organizational learning with scenarios. (Kindle version ed.). New York City: John Wiley & Sons.