Developing Scenarios: Linking
Environmental Scanning and Strategic Planning
(reprinted from Whiteley, M. A., Porter, J. D., Morrison, J. L., & Moore, N. (1990). Developing scenarios: Linking environmental scanning and strategic planning. Planning for Higher Education , 18(4), 47-60.)
21st century has been the frequent setting for fantastic tales painted by
science fiction writers, imaginative comic books, and the movies.
By now, however, planners for the nation's businesses, government,
colleges and universities need little reminder that the 21st century is within
our strategic-planning horizons. The
nineties are a critical transition decade, frequently characterized by an
acceleration of technological, economic, and social change, increasing
uncertainty, and growing international interdependency.
in this age of change, transformation, and uncertainty takes on a number of new
dimensions, calling for methods that are not mere extrapolations of past trends
but that take into consideration how the future win be different, or
"discontinuous," from the past. A
central objective of this planning is to place the organization in a
strategically flexible position in order to alter its strategies as the
contextual conditions of the organization change (Becker and van Doom, 1987;
For some time, planners in large businesses, marketing, and the defense and energy industries have recognized the need for non-extrapolative planning methods better suited for a time of rapid change and discontinuity. Many of the new methods they devised are based on group or expert judgment techniques and result in the development of alternative scenarios of the future. A recent "Manager's Guide to Forecasting" notes that, of the many non-extrapolative methods now in use, the scenario method is a popular and strong choice in developing strategies for the future. A 1983 survey of Fortune 1000 companies, for example, found that 35 percent of these large and mostly multinational firms use the Multiple Scenario Analysis technique. This figure more than doubled in the four years between 1977 and 1981 (Georgoff and Murdick, 1986; Godet, 1987; Linneman and Klein, 1983).
critical goal of the scenario method is to guide managers and executives in the
organization in their thinking about the future and the implications of the
future for the organization. The
impetus for much of the move away from planning based on extrapolative methods
has been the errors of past extrapolative forecasting.
A notable example was the failure of the energy industry to conceive of
and plan for the energy crisis of the 1970s.
The principal error of extrapolative planning of the type conducted in
the energy industry was the assumption that the future will be a continuation of
the past (Hankinson, 1986; Millett and Randles, 1986).
this article, we discuss a method for developing and writing scenarios for a
college or university. We begin by
reviewing the general literature on scenarios; we then detail a scenario
development project at Arizona State University. This project, conducted in 1988-89, was Arizona State
University's first institution-wide, futures-based planning and scenario
development effort. The focus of
the project for Arizona State University was planning and programming for
affirmative action. An outside
consultant facilitated the group-process portion of the project and instructed
university staff in scenario development. Staff
in the university's Office of Institutional Analysis then developed and wrote a
set of three scenarios to guide the university's affirmative action programming
and planning during the decade of the nineties.
of scenarios vary. The person most
often attributed with the founding of the scenario method in the early 1950s,
Herman Kahn, defined the scenario as "a hypothetical sequence of events
constructed for the purpose of focusing attention on causal processes or
decision points" (Kahn and Weiner, 1967, p. 6).
recently, Georgoff and Murdick (1986), in their comparison of forecasting
techniques, defined scenarios "as smoothly unfolding narratives that
describe an assumed future expressed through a sequence of time frames or
written as a group, provide a range of possible futures.
The optimum number of scenarios and whether the set should include a
"most likely" (i.e., steady state) scenario are issues of debate. The focus of scenarios is on conditions and changes resulting
in a future state, identifying threats and opportunities along the way.
While scenarios may include elements of an organization's internal
environment, most deal primarily with describing the external context within which
the organization operates. The use
of scenarios allows for the integration of both quantifiable economic and
demographic data, along with a wider variety of information on technological
environments, social attitudes, and political trends not usually incorporated in
more traditional forecasting processes. A
critical element of scenario development is specification and analysis of the
interrelationships of factors often viewed as discrete entities in other
planning methods (Godet, 1987; Hankinson, 1986; Linneman and Klein, 1983;
Mendell, 1985; Schnaars, 1987).
approaches can be taken in writing scenarios, ranging from a single person
writing a description of a future situation to the use of an interactive
computer model that uses cross-impact analysis to generate outlines of the
alternatives (e.g., Becker, 1983; Boucher, 1985; Enzer, 1980; Goldfarb and Huss,
1988; Martino, 1983; Mecca and Adams, 1985; Morrison and Mecca, 1989).
process of constructing and writing scenarios is often described as something of
an art form in which the "artist" (i.e., unsuspecting strategic
planner) must weave a number of interrelated events and trends into a coherent
picture of the future, complete with strategic options.
However, over the past thirty years of scenario use, both constructing
and writing scenarios have evolved into a fairly standardized process with
several defined steps:
the scenario development process involves a Delphi or mini-Delphi process to
identify systematically the factors expected to affect the organizational
planning issue and to establish future values for these factors (Godet, 1987;
Morrison, Renfro and Boucher, 1984; Schnaars, 1987). As Morrison, Renfro, and Boucher note in their Futures
Research and the Strategic Planning Process: Implications for Higher Education
(1984), "Delphi was designed to obtain consensus forecasts from a group
of i experts' on the assumption that many heads are indeed often better than
one, an assumption supported by the argument that a group estimate is at least
as reliable as that of a randomly chosen expert' (p. 47).
as Michael Godet (1987) recently warned, the use of Delphi is limited by the
fact that, "the history of forecasting errors shows rather, that we should
distrust predominant ideas; the correct viewpoint is often to be found in the
minority view" (p. 66).
product of group identification and estimation of factors thought to affect the
context of an organization's planning is a large set of interrelated
variables, each with a number of future data points.
These variables usually include trends with a number of future values as
well as events judged on the basis of their probabilities of occurring by a
certain future date.
analysis was developed during the 1950s by Olaf Helmer of the Rand Corporation
to aid in the organization and assessment of large sets of variables into
scenarios. Cross-impact analysis
builds on the convergence of opinions obtained through a Delphi process by
considering interactions between trends and events.
For example, consider how substantial curtailment of federally funded
student financial aid might affect enrollments or minority student access at
public colleges and universities across the country. Or, in the larger environment, consider how the defaulting of
Mexico on loans made by American banks might affect key economic factors in the
United States. Cross-impact
analysis considers the interdependence of factors, netting an interpretation
grid, or "matrix of discovery," from which alternative scenarios can
be developed (Godet, 1987; Minett and Randles, 1986; Schnaars, 1987).
difficulty in using cross-impact analysis is its complexity.
Even a small group of ten trends and ten events has many potential cross
impacts. Further, the specification
of the cross impact can range from a consideration of the direction (+ or -) and
degree of impact to include an estimation of the duration of impact or even
estimations of the nature/intensity of impact in each year of the planning
period. How much interaction occurs
in cross-impact analysis depends on which trends and events are included and the
assumptions made concerning the direction and order of the interactions.
Quite simply, getting people in the organization to commit time and
effort to specifying cross impacts is extremely difficult.
advantage of including cross impacts in developing scenarios is that the process
of specifying cross impacts illuminates relationships among factors that may not
be considered otherwise. It also
enables the planner to visualize a more complete future because reality is a
complicated blending of trends and events.
of the complications inherent in cross-impact analysis, assumptions must be made
to facilitate the process. Over the
past twenty years, several computer algorithms have been developed to simulate
cross impacts based on differing methodologies for determining whether or not an
event occurs (e.g., Monte Carlo or closed-form cross-impact models).
However, these models tend to be complex and subject to severe
because of computer-programming exigencies.
Development at Arizona State University
spring of 1988, Arizona State University launched a futures-based planning
project based on the Delphi judgments of approximately eighty faculty,
administrators, and students. We
then used cross-impact analysis to produce data from which we developed three
alternative futures: the Stable Future, the Turbulent Future, and the Chaotic
Future. This paper focuses on the
development of these three scenarios, using the output generated by cross-impact
State University is in a dynamic environment not unlike many other colleges and
universities, particularly in the growing West. The next ten years are expected to be as increasingly
dynamic. Coupled with this growth
are significant transformations of the metro-Phoenix economy and lifestyle.
Phoenix is, and will continue to be, influenced by changes affecting the
United States and world society, including the following:
focus of the Arizona State University project was the university's affirmative
action program and policies for students, faculty, and staff.
We chose the scenario method to achieve several objectives:
In addition, we hoped to
provide a vehicle to institutionalize, within the organizational culture of the
university, futures-oriented planning.
developed a questionnaire including fifty-four trends and forty-three events.
This set included international trends and events such as the possibility
that "a major regional conflict involving U.S. military troops
erupts," local trends and events such as "the resale value of homes in
the metro Phoenix area," and internal university trends and events such as
the "enrollment caps are implemented." We asked participants to do
Using a Delphi approach, we
provided participants with group medians and interquartile ranges and asked
individuals to re-vote. This
process included a half-day workshop and another round of the Delphi
estimated the cross impacts of events on events and events on trends.
The objective of cross-impact analysis is to examine systematically the
interrelationships in the trends and events set, a process that involves asking
how the prior occurrence of a particular event might have an impact on other
events or trends in the set. For
example, for the trend "Enrollment at Arizona State University,"
suppose that the event "AU non-military federal government student
financial aid funds are cut by 50 percent" occurs.
We then ask the question, How would the occurrence of this event affect
this trend? Would the enrollments decrease by 10 percent?
Or would enrollments remain unchanged?
respect to the relationship of events to events, suppose the following event
occurs: "A major regional conflict involving U.S. military troops
erupts." We ask the question, How will the occurrence of this event affect
another event, such as, "A major depression occurs." Would such a
conflict decrease the probability of a depression? If so, by how much?
cross-impact judgments were elaborated in terms of the direction (+ or -) of the
impact in the trend value or event probability, the magnitude of the impact, and
the duration of the impact. Figure
1 illustrates how the trend "level of federal student financial aid"
is impacted by three events: major
depression, balanced budget amendment, and regional conflict . For example, if a
major depression occurs, the estimated impact on student
financial aid would be a 5 percent decrease in the year following the
year in which the depression occurs. By
the beginning of the third year, the effect of a depression on the level of
student financial aid, however, was estimated to be zero.
On the other hand, a regional conflict was estimated to increase the
level of student financial aid by 2 percent in the year after this event occurs,
increasing the trend level by 10 percent five years after occurrence.
By the seventh year, the influence of this event on the trend return was
estimated to return to zero.
values of the cross-impact judgments were then used to adjust the Delphi values
for interacting events and interacting events and trends. Figure 2 illustrates
the change in trend level showing the effects on level of student financial aid
if the three events occur.
Delphi values, with the cross-impact judgment alterations for impacted events
and trends, then became the data used to generate our three scenarios.
our application of cross-impact analysis, we defined each of our three scenarios
- Stable, Turbulent, and Chaotic - in terms of a trigger probability value.
For example, we set the trigger value for the Stable Scenario at a
probability level of .80. AR events with a probability of .80 or greater were
said to occur in the year their probabilities reached the .80 threshold.
Each event's occurrence then set off a chain of alterations in the values
of other impacted events and trends.
manipulated trigger values to obtain rich data sets for each scenario.
For the Turbulent Scenario, we set the trigger probability value at .60
and set the trigger value for the Chaotic Scenario at .35. The trigger values
then produced scenarios defined by the events in the environment.
approached the task of writing scenarios from the vantage point of contemporary
institutional historians in the year 2000 writing about the university's
environment and internal developments during the 1990s.
Our starting point for each scenario was the event history.
developing scenarios, we established an organizational and conceptual pattern
that we followed throughout the scenario writing. First, we separated our discussion of events and trends.
Events were handled as the driving forces of the scenarios, with trends
as the descriptors. We began our
scenarios with a narrative overview, laying out the major events that occurred
over the decade of the nineties. For
example, in the Turbulent Scenario, we described the events occurring during the
nineties in the following overview:
This report surveys the history of developments affecting Arizona State University and its affirmative action efforts over the decade of the 1990s. The report chronicles the interplay of broad external forces, as well as the more specific socioeconomic and demographic factors shaping the environment for ASU's recruitment and retention of racial/ethnic minority and female students, faculty, and staff.
Economic depression was the dominant theme of the decade. Starting in 1994, the depression had wide-ranging impact lasting throughout the rest of the decade.
One effect of the depression was the significant rise in demand for public-funded human services . . . and rising gang activity.....
The Arizona legislature responded to the problems of increased demand and lower revenues by focusing on accountability and increasingly intervening in the operations of state agencies...
Nationwide, demands for increased accountability, educational outcomes, and raised academic achievement standards were major educational policy issues throughout the decade.... As the decade closes, debate continues on the issues of the impact of financial aid cuts on minority student access and the racial bias of new eligibility requirements.
this backdrop, Arizona State continued its enrollment growth pattern.
The body of the scenario then dealt with trends, describing their levels and interrelationships over the decade in both narrative and graphics. To handle the large number of trends in our Delphi process, we organized our scenarios into five subsections: Economics, Demographics, Characteristics of the Workforce and Workplace, Education, and Focus on ASU. Here we discussed the implications for the organization of the contextual trends described in other sections, setting the stage for strategic policy analysis and planning.
task was to weave this large group of trends into a coherent image of the
decade. In preparing our scenarios,
however, we were struck immediately with a number of trends that appeared to be
contradictory and even conflicting. For
example, the trend levels for one scenario indicated a substantial increase in
the number of women head-of-household families living below the poverty level-a
not surprising, negative picture of women's economic status.
At the same time, however, trend levels for the same scenario painted a
more positive picture of substantially more women in graduate programs.
Added to this were a number of other trends, some indicating a worsening
in the economic and career status of women, others indicating improvement.
first reaction to contradictions of this kind was to check our data. Finding that our output correctly reflected the Delphi
judgment values plus the cross-impact estimations, we then were left to puzzle
over issues involved in the economic status of women. What we concluded was that trend projections in our output
might well reflect a widening opportunity gap, separating the futures of those
women with marketable skills and education from the futures of those with little
skills or training.
of the Three Scenarios
Stable Scenario generally portrayed a steady state future.
Only six events occurred in this scenario, generating few cross-impact
alterations of the Delphi estimations of trends.
Three of these events related directly to higher education.
In this scenario, the economy remained relatively stable, with only a
slight downward trend in 1994-95. Demographic
trends in the Stable Scenario included a rise in the numbers of the local
population living below the poverty level and an increase in the proportion of
women head-of-household families. These
and other trends resulted in a dramatic rise in demand for social services.
Characteristics of the Workforce and Workplace reflected the widening
opportunity gap for both racial/ethnic minorities and women between those with
marketable skills and education and those without.
Throughout the decade, the latter group was portrayed as experiencing an
increasing dislocation and economic instability.
in the Stable Scenario reflected rising college tuition costs and declining
federal financial aid dollars. While
Arizona's racial/ethnic minority student college enrollments rose substantially
in this environment, most of this increase was in the state's community
colleges. The state's universities
garnered only a small 10 percent of this gain.
In this environment, the Focus on ASU section described how Arizona State
University faced the decision of whether to cap enrollment and the question of
whether this decision would affect minority student access.
Turbulent Scenario portrayed a decade characterized by an economic
recession/depression and its related social problems. In this decade, fourteen events occurred, including a major
depression in 1994, a fossil fuel crisis
of 1973-74 proportions in 1999, and legislation passed by the Arizona
legislature tying funding increases in education to educational outcomes in
1998. One result of the economic
recession/depression was a dramatic increase in demand on publicly funded social
services. The Arizona legislature
responded to increased demand and lower revenues by focusing on accountability
and by increasingly intervening in agency operations, including the state's
trends reflected the economic downturn in a dramatic increase in homelessness
and poverty. Characteristics of the
Workforce and Workplace in the Turbulent Scenario reflected the economic
depression with an increase in unemployment and a widening of the opportunity
gap. The impact of the depression
on the individual, however, was buffered partly by education, experience, and/or
skills. The Focus on ASU section of
the scenario included the following narrative:
The effects of the
depression on ASU were complex. While
the decline in federal financial aid put the university out of reach for the
severely economically disadvantaged, overall enrollment continued to climb
during the first half of the decade. As
enrollments climbed, debate centered around the decisions to cap main campus
enrollments and establish a second branch campus in the East Valley....
In this environment, ASU's affirmative action hiring tasks became more complex. On the one hand, the depression improved the size of the pools of racial/ethnic minority and women faculty and management personnel by as much as 33% during the period. Yet, on the other hand, salaries in the higher education sector became less competitive with the outside marketplace during the decade. While the effects of the depression on ASU were perhaps less severe than for many higher education institutions, the decline in appropriations and enrollment caps translated into only a few new positions and minimum hiring opportunities.
Chaotic Scenario, we portrayed the decade of the nineties as a time of
dislocation and disruption. Driving
the Chaotic Scenario were nineteen events, including "a major depression
occurs," "all nonmilitary federal government student financial aid
funds are cut by at least 50 percent," "a major regional conflict
involving U.S. military troops erupts," and "the U.S. experiences a
dramatic flood of refugees from Mexico, Central, and/or South America." The
background section of the scenario described government's response:
with problems of unprecedented magnitude, state and local governments responded
with haphazard and often drastic policies and measures . . . in an environment
of dwindling resources for higher education, demands for increased
accountability, educational outcomes measures, and raised academic achievement
standards were major educational policy responses.
depression of the Chaotic Scenario was reflected in a number of trends described
in the Economics section of the scenario. The
rate of new job creation in Arizona, for example, fell by 20 percent over the
decade. At the same time, business
relocation to the Sunbelt fell by 26 percent, and the resale value of homes in
metro-Phoenix declined by 27 percent. Not
surprisingly, the social problems described in the Stable and Turbulent
Scenarios deepened in the Chaotic nineties. In the Characteristics of the Workforce and Workplace section
of the Chaotic Scenario, we detailed how the worldwide depression in an era of
technological revolution brought changes to the workforce by raising the social
security retirement age to 70 and changing the distribution of industry in the
metro-Phoenix area. In addition,
the depression of the Chaotic Scenario reflected a substantial change in the
opportunity gap described in the Stable and Turbulent Scenarios. While opportunities remained steady throughout this decade
for skilled and educated women, opportunities for skilled and educated
minorities declined substantially.
impact of the Chaotic Scenario conditions on education and Arizona State
University was catastrophic. Tuition
and funding cuts not only took their toll on the disadvantaged, but also
devastated enrollments of whites and racial/ethnic minorities throughout the
higher education system. For
Arizona State University, the severe depression conditions of the Chaotic
Scenario brought faculty and staff layoffs, a downscaling of programs, and an
abandonment of earlier affirmative action goals and strategies.
development and use of scenarios at Arizona State University was an experiment
for us in terms of both (1) the dynamics of the process within our organization
and (2) the methodology of writing scenarios.
Our experience developing scenarios taught us a number of lessons that
may be helpful to other institutions contemplating the development of scenarios
for strategic planning.
and the Organization
changes within the university over the course of the project made it difficult
for us to assess fully how the process works within the organization.
Early in the project, the associate vice president for student affairs, a
key supporter and organizer of the process, took a position as a vice president
at another university. At about the
same time, after eight years, the university president announced his retirement. However, despite the administrative disruption during the
project, in the year since the project's completion, we see that several of our
organizational objectives were met. The
process of developing and using scenarios accomplished the following:
is no cookbook for writing scenarios. While
we rather facetiously noted earlier that scenario writing is often referred to
as an "art form," it is a highly creative process.
Scenario writing requires a narrative writer unafraid of taking risks,
making generalizations, and going out on a limb.
While cross-impact analysis does provide an organized set of data for
scenario writing, there is a creatively ambiguous gap still remaining between
the data and the written story. From
the vantage point of hindsight, we see that our scenario development and writing
would have been aided if we had
Planning with Scenarios
addition, we discovered that development and use of scenarios add a number of
dimensions that are particularly important to planning in a time of rapid
sum, our experience at Arizona State University showed us that scenario
development is a viable tool, particularly for expanding the thinking of
planners and focusing the environmental-scanning effort.
However, the institutionalization of scenario development and use across
the organization is, for us, still a question.
If the process can be streamlined and the benefits more directly related
to specific decisions, we believe scenarios can make an important contribution
in better preparing colleges and universities for a new era that promises to be
dramatically different from the past.
A. Whiteley is a senior researcher, John
D. Porter is director, and Nelle Moore
is research analyst, all in the Office of
Institutional Analysis at Arizona State University.
James L. Morrison is professor of education at the University of North Carolina, Chapel
Henk A. and van Doom, Joseph W. M. "Scenarios in an Organizational
19, 1987, 669-677.
Becker, Howard S. 'Scenarios: Importance to Policy Analysts." Technological Forecasting and Social Change, 18, 1983, 95-120.
Boucher, Wayne 1. "Scenarios and Scenario Writing." Nonextrapolative Methods in Business Forecasting: Scenarios, Vision, and Issues Management. Edited by Jay S. Mendell. Westport, CT: Quorum Books, 1985.
The Council of State Governments. The Dynamic West: A Region in Transition. IL: Westrends, The Council of State Governments, 1989.
Selyn. "INTERAX-AN Interactive
Model for Studying Future Business Environments: Part I." Technological Forecasting and Social Change, 17, 1980,
David M. and Murdick, Robert G. "Manager's Guide to Forecasting." Harvard
Business Review, 64, 1986, 110-120.
and Strategic Management. Boston,
MA: Butterworths, 1987.
David L. and Huss, William R. "Building Scenarios for an Electric
G.A. "Energy Scenarios -The Sizewell Experience." Long
Range Planning, 19, 1986, 94-101.
Herman and Weiner, Anthony J. The Year
2000. London: Macmillan, 1967.
Robert E. and Klein, Harold E. "The Use of Multiple Scenarios by U.S.
Industrial Companies: A Comparison Study, 1977-1981." Long
Range Planning, 16, 1983, 94-101. Martino,
Joseph P. Technological Forecasting for
Decision-Making, 2d ed. New
Thomas V. and Adams, Charles F. "Policy Analysis and Simulation System for
Educational Institutions." Paper presented at the annual meeting of the
American Educational Research Association, Chicago, IL, 1985. (ERIC Document
Reproduction Service No. ED o265 632).
Jay S. Nonextrapolative Methods in Business Forecasting: Scenarios, Vision, and
Issues Management. Westport,
CT: Quorum Books, 1985.
Millett, Steven M. and Randles, Fred. "Scenarios for Strategic Business Planning: A Case History for Aerospace and Defense Companies." Interfaces, 16, 1986, 64-72.
Morrison, James L. and Mecca, Thomas V. "Managing Uncertainty: Environmental Analysis/ Forecasting in Academic Planning." Higher Education: Handbook of Theory and Research, Volume V. Edited by John C. Smart. New York, NY: Agathon Press, 1989, 354-390.
Morrison, James L.,
Renfro, William L., and Boucher, Wayne 1. Futures
Research and the Strategic Planning Process: Implications for Higher Education.
ASHE/ERIC Higher education Research Reports, Number 9, 1984.
Steven P. "How to Develop and Use Scenarios." Long
Range Planning, 20, 1987, 105-114.