Commentary: Strategy as a Profession

in the Future Security Environment

Andrew W. Marshall

Revised and updated version of Marshalls essay,

Strategy as a Profession for Future Generations, in
Marshall, J. J. Martin and Henry S. Rowen, eds., On Not
Confusing Ourselves: Essays on National Security Strategy
in Honor of Albert and Roberta Wohlstetter, Boulder, CO:
Westview Press, 1991, pp. 302-311.

The future is always full of uncertainties. A common error is

to underestimate the scale and multiplicity of the uncertainties.
This is a general failing that Nassim Taleb in his book, The Black
Swan, explores in detail.1 Here we are concerned with the national
security area. In this case, as elsewhere, some aspects of the future
are more predictable than others, and good assessments and
strategies take whatever advantage they can of this. Demographic
trends, relative rates of economic growth are some examples of
relatively more predictable aspects of the future. Also cultural
beliefs in different societies are more stable than other aspects of
the future.
But big changes are also common, indeed major shocks can
occur, and tend to be under-represented in forecasts of the future
not only for the reasons that psychologists tell us about, but in
the national security area because of the pressures of political
correctness. Some topics, some future scenarios, may tend to be
avoided, almost as taboo for a variety of reasons.
We need a strategy, or strategies, that both takes account of our
best assessment of the competition we are involved in, now and in
the future, and in some way takes account of the uncertainties of
the future situation. As I will address below, Albert Wohlstetter
was especially adept in his strategic thinking, particularly on
this score. And Roberta Wohlstetter in her book on Pearl Harbor
stresses the inevitable uncertainty of the future. We will never
know, ahead of time, the future. I have found it useful to think in
terms of the following model: there are the players, all with their
individual goals, resources, distinctive culture, and strategies;
and there is the context, which none of the players controls, for
example, technology, climate, etc. There are long-term trends in
many of these variables, and enduring asymmetries between the

625
players. A good strategy would have to accommodate in some
way all of this, reflect the trends that are changing the situation,
as well as exploit some asymmetry that provides the basis for
advantages he has in achieving his goals. Strategies can involve
coalitions, and obviously they must address adversaries. And, in
some way, they must aim to limit the risks that the uncertainties
pose.
Richard Rumelt in his forthcoming book has an excellent
characterization of strategies as solutions to solve complex
problems. One of the virtues of Rumelts discussion is that it
provides real clarity about how the word strategy should be
used. In practice, the word strategy tends to be used in too many
ways. In particular I would note that in the national security
area, which is the main focus here, there is a constant tendency to
think of military strategy as related principally to the application
of resources in a possible future war and the general guidance
for more detailed planning for specific contingencies. The result
is that there is relatively little discussion of strategies for the
peacetime management of our military organizations and for the
allocation of resources over time so as to develop more efficient,
effective, competitive military forces with appropriate doctrines
and concepts of operations. Most statements of national security
strategy tend to be just long lists of desirable goals with little to
say about how these goals might be achieved. Good examples of
fully developed national security strategies are thus very few.
There is, then, a special problem in the national security area.
Given the existence of nuclear weapons, the highest priority
objective for the United States has been deterrence of large-
scale war. In this we have been largely successful. Therefore,
the strategic management problem in our national security
establishment was for a long time the peacetime competition to
preserve and indeed enhance in the future our ability to deter
the Soviet Union from actions adverse to our interests. Now this
definition of our priority objective may need serious amendment
as we move into a different world. The discernible aspects of this
world are: the rise of Asia and decline of Europe, a long, extended
struggle with Islamic extremists, wider proliferation of weapons,
including nuclear weapons, and continued rapid scientific and
technological changes.
With new problems, new thinking will be required. It is not
that the uncertainty is higher. There were lots of uncertainties in
the late 1940s and the 1950s, indeed throughout the Cold War.

626
But there are new players, new options, and the natures of the
competitions are different. We will need to be as serious about
strategy as we were in the early stages of the Cold War. Finding
the right people and organizing the right sorts of teams will be
important.
It is clear that some people among us seem more readily able to
address issues of strategy, in particular the strategic management
of our national security efforts. They have a willingness and a
self-confidence to address the larger issues than do others. They
appear to bring a very different perspective to the discussion of
what our strategy ought to be. How do they get this way? What
sort of training is useful? This is what I want to address in the next
two sections.

What Environments Produce Strategists?

This is a question that deserves extensive study. The best I

can do is to draw upon my experience in and observations of
the environment at the RAND Corporation in the 1950s and
early 1960s and my later experience in government in the period
1972 to the present. One disadvantage of focusing on RAND as
a producer of strategists is that it clearly biases the discussion
toward an analysis of the development of people whose role has
been advising, in the sense that Herb Goldhamer used in his
book, The Adviser.2 There are other routes to becoming a strategist,
including those who reach high positions in the military services
or enter government service from other career lines such as the
law or investment banking. But the case of RAND is perhaps of
special interest because it did provide in the 1950s and early 1960s
an environment that produced a number of people who are now
acknowledged as major strategic thinkers.

The RAND Experience

There was something special about the RAND environment

from the late 1940s through most of the 1960s. For one thing,
especially in the late 1940s and the 1950s, there was a sense of
being on the leading edge, of dealing with the centrally important
problems. The invention of nuclear weapons and several other
technology developments at the end of World War II produced a
situation that was quite new, one in which the issue of what our
strategy should be was extremely important. Another aspect of

627
this situation, given the large increase in destructive power nuclear
weapons introduced, was that there were no experts. Two small
weapons had been used at the very end of World War II; what
larger numbers of weapons and more powerful weapons might
do to change the nature of war was unclear. Nobel prizewinners
were no better than graduate students in thinking about the
relevant issues, and at meetings and working groups at RAND in
the early days there was no hierarchy. This was an ideal situation
for younger people (the average age of the professional staff at
RAND in 1950 was about twenty-eight), who were immediately
treated as equals and valued for what they could contribute to
the discussions. This is a rare situation, certainly not characteristic
of academia or normal organizations, and it led to the rapid
development of individuals who were willing to address the
broadest issues of national security. There was also a sense of
having a preferred position with respect to access to information
on the new developments taking place in weaponry, in particular
in the design of nuclear weapons, their delivery systems, and
other relevant technology.
Two other things favored the development of strategic think-
ing and innovation at RAND, and the willingness of the people
there to address the highest level national strategy issues. One
was the freedom RAND had to select the problems and the issues
on which it worked. This is very different from the environment
in contract studies organizations, especially now. The other was
the presence of several remarkable men who set the intellectual
tone and style of much of the broader strategies analysis that
began in the early 1950s. Two I would name are Charles Hitch
and John Williams, the heads respectively of the Economics and
the Mathematics Divisions. Apart from their own intellectual
contributions, their cultivation of full-ranging discussion, their
intellectual fairness, and their interest in the development of
younger people and of new methods of analysis all favored the
fullest examination of all issues of U.S. national security.
One of the interesting things that happened at RAND was
the success of the economists in assuming a leading role in the
direction of a number of important studies and, more generally,
in shaping the way in which RAND addressed national security
issues. Initially the economists were brought into what had been
largely a technological organization to deal with what was called
the military worth issue. It had become clear to the technical
people that they needed some assistance in thinking about the

628
objectives that military weapon systems were to achieve. There
was also some interest in the economics of defense, especially
as it dealt with issues of mobilization, and in the targeting of an
opponents industrial capacity and assessing damage to industrial
societies from strategic bombing. The economists soon played a
much larger and more central role in managing and directing a
number of the successful studies. Why was this?
Herman Kahn and I used to discuss this puzzle. We had a
number of hypotheses. For one thing the economics of the situation,
broadly conceived, were important. What things cost, the level
of resources that nations are able to devote to defense over an
extended periodthese all shape ones views as to the kinds of
weapon systems and forces that are desirable and feasible. But
another advantage the economists had was that they knew from
their own experience that experts could be wrong. Indeed, they
also knew that much discussion of economic problems is foolish
and that many widely-held views, even among responsible
people, are faulty. The experience of engineers and physicists is
different. In those fields there are real experts who are much more
likely to be right than are others. Economists, therefore, were
more intellectually comfortable in the situation that existed with
respect to nuclear warfare, in which there were no experts.
One of the people in the economics department who was
the first to lead and manage a large RAND study was Albert
Wohlstetter. Beginning in the early 1950s, he examined a set of
issues connected with the basing of long-range bombers. I want to
note what seems to me one of the major innovations or inventions
Albert made in the conduct of that study. In previous large RAND
studies, the practice had been to lay out a number of alternative
systems or programs at the very beginning of the study. The study
itself focused on evaluating which of the alternative systems was
the most cost-effective.
Alberts approach was different. He started with a few
alternatives to the existing plan or program, but as the study
went on he evolved improved alternatives. He was also less rigid
than had been reflected in the earlier practice in setting down the
criteria, the objective functions, the measures of effectiveness at
the beginning of the study, and then simply sticking with them.
His evolutionary approach developed additional criteria and tests
of performance as more understanding of the problems and the
issues emerged. And a wider range of situations within which the
alternative possible solutions could be tested grew as the study

629
went on. This was, in my judgment, a crucial invention for doing
these kinds of studies, because one would learn much more about
the nature of the issues and the problems, how one ought to look
at them, and what criteria were relevant as one went further along
in the studies. Also, this way of conducting the analysis had the
advantage of inventing additional and better alternative solutions
to examine as one went along. Alberts study was in many ways
emblematic of the kind of good strategic analysis I wrote about at
the beginning of this essay: it accepted certain structural elements
of the situations, and then sought measures to both limit and
mitigate effects of the uncertainty about the future.3
Another aspect of the situation at RAND that was exceptionally
favorable to strategic thinking and innovation during the early
period was the practice of inviting first-rate people to come and
spend the summer. This created an environment in which the
important thing was to try to tap into the very best talent in the
whole country. The objective was not to do the best that RAND
could do with its existing staff, but in a sense to do an analysis that
was the best that the country as a whole could accomplish. By its
very nature, any organization is limited in the amount and variety
of talent, backgrounds, and insights that it can include among
its staff. This attitude of searching for the very best people and
drawing on the best talent is a key to excellence in broad thinking
about any problem or issue. Unfortunately, most organizations
do not operate this way.
Another way in which Albert was especially good was in
reaching outside Rand to get the best technical advice. In the
mid-1950s the experts, at Rand and a DoD advisory group on
physical vulnerability, believed that no structures could be built
to withstand blast overpressures exceeding something like 25
psi. Albert recruited Paul Weidlinger, an innovative structural
engineer, to design hardened structures for protecting aircraft and
missiles to withstand overpressures far beyond this limit. Herman
Kahn was also involved because of his knowledge of the physics
of nuclear weapons effects. This led, after a long argument and
tests, to a major shift in views of what was possible.
The RAND of the 1950s and early 1960s was a remarkable
place, both for the talent it recruited and for its atmosphere and
intellectual dynamic. It was also remarkable for its boldness in
addressing broader questions of strategy. It is, therefore, not
surprising that some interesting and influential people developed
there.

630
The U.S. Government

The next experience that is perhaps relevant comes from my

time in government. Beginning in the middle 1970s, I was involved
in attempts to initiate strategic planning activities in the Depart-
ment of Defense including some strategic planning experiments.
In particular, James Roche, then a U.S. Navy Commander, and I
wrote several papers during 1975-1976 to promote strategic think-
ing in the Defense Department. We also sponsored contractor
research on some aspects of strategic planning. This experience
led me to believe that, while systems analysis had been a liberating
force during its early development, by the middle 1970s it had
become a constraint on thinking strategically. People who were
systems analysts found it difficult to address the sorts of questions
that we felt needed to be considered in strategic planning. People
with a business background or a combination of business school
and military service seemed to be among the best at taking up and
addressing the questions we wanted dealt with.
We saw it as a vaccination problem: some backgrounds
promoted strategic thinking and others seemed to inoculate
people against it. Why is that? To some extent, the systems analysts
had by that time developed routine approaches to analysis and
perhaps had ceased paying sufficient attention to the complex
consequences of acquiring the systems they dealt with. James
Schlesinger commented to me a number of years ago that systems
analysis proceeds by trivializing the measurement of effectiveness
while perfecting the analysis and estimate of costs. Programmatic
actions, the acquisition of particular weapon systems, the adoption
of a new concept of operations, and the setting of new objectives
for military forces have complex consequences, including their
effects upon the beliefs, actions, and resource allocation patterns
of potential opponents. Most of these consequences are not
usually considered in the standard kinds of analysis. One result
is that the top leaders of the Department of Defense often get
remarkably little assistance from their staffs when truly strategic
decisions are addressed. This is because the focus of the work of
the staffs, the criteria they use, and their measures of effectiveness
are too narrow to account for the considerations that top-level
decisionmakers in fact want to consider, are concerned with, and
take into account as best they can.

631
 Some decisions have larger and different consequences
than others. For example, a decision to pursue or create a major
strategic defense capability is different from a choice among
several alternative programs for the next generation of fighter
aircraft. The former involves going into a new business for the
U.S. military (although it is a business we once were in), the latter
the continuation of an existing business. Different issues are
involved, different forms of analysis seem needed, but existing
analysis methods tend to treat the two types of decisions the
same way. Part of the problem may be that much if not all of
the existing analysis methodology was developed to assist in
procurement or operational planning decisions. Other methods
of analysis are necessary when the questions are more like: What
businesses should I be in? What are my competitive advantages?
One advantage people from the business world or business
schools may have is that they are used to addressing these kinds
of questions, though often with analysis methods that are less
systematic.

What Backgrounds and Experiences Are Conducive to Strategic

Thinking?

There is no specific set of disciplines that must be mastered to

be a strategist. People who think strategically come from a number
of different backgrounds. Among those whom I have met, and
feel that I know personally, the best academic backgrounds seem
to be economics, business school, applied technology (especially
for those who have been in the business world), and in some cases
political science. But what seems to be central is a cast of mind
that is questioning, eclectic, able to address the broadest kinds
of issues and goals, and able to formulate appropriate ways of
achieving these goals. A high tolerance for the uncertainty that
necessarily accompanies any effort to think forward five, ten, or
twenty years is required. For many people, some period of intense
involvement in an important, large-scale project or enterprise has
proved to be crucial.
World War II was such an experience for a number of people
and, indeed, there may be a generational factor at work: living
in interesting times may contribute to being a good strategist.
People who were involvedeven if only in staff positions or on
the peripheriesin some major decisionmaking body connected
with that war had a special quality about them. Experiences in

632
World War II clearly had a significant impact on a number of
the people who were at RAND during the 1950s. Because they
contained many people with World War II experience the Truman
and Eisenhower administrations had a character to them that
favored strategic thinking. This characteristic of administrations
has gradually eroded since the late 1950s.
The changes that we now see in the security environment of
the United States are forcing another major effort of rethinking
our situation, our goals, and our strategies. It might, therefore,
be a period in which a new generation of strategic thinkers will
emerge as a result of the critical experiences they will go through
in the next decade.
Turning to the question of what kind of academic study or
professional training might be useful, I would start with economics
and business school training, especially business schools that
have strong programs in business policy and strategy. My
recommendation about economics is, however, a guarded one.
Since the 1940s and 1950s, economics training has become too
mathematical, too focused on the acquisition of particular analytic
tools that are not, in fact, of much use in the national security
area. Something like the first courses in graduate school may be
enough. They are important, however, because people who do not
have a sense of macroeconomics and the fundamental tradeoffs
that societies have to make, find it difficult to think clearly about
the long-term implications of devoting large, possibly excessive,
percentages of gross national product (GNP) to military uses.
In the early 1980s, when the first initiatives were taken
within the Defense Department to encourage application of a set
of ideas that later were labeled as competitive strategies, I had
a discussion with the chief of one of the military services. His
reaction to the idea of designing some military programs so as to
impose increased costs upon the Soviets was negative, or at least
cautious. He had two arguments against focusing on increasing
Soviet costs or expenditures.
The first was that the Soviets would simply spend the extra
money, there being no reasons for them not to do so; the second
was that our own budgets fluctuate so much that it was unwise
to stimulate a competition which we ourselves might not sustain.
The second of these arguments has real merit to it. The first shows
an unawareness of the long-term consequences for the Soviets
of high levels of military expenditures or of possible tradeoffs
between individual programs the Soviets might be compelled to
make, since resources always are limited.

633
 Another virtue of economics training, or for that matter
business school training, is that a modest amount of mathematics
is acquired, as is some sense of the importance of technology
and an ability to interact more effectively with technologists and
hard scientists. This was one of the advantages the economists
had over the political scientists at RAND in the early 1950s:
quantitative analysis was something the economists were used
to, and their interest in or ability to discuss and understand what
the technologists were up to was somewhat better than that of the
political scientists.
Demography is another area that deserves much more
attention than it has had in the past in the development of strategy.
The relationship of demography to political and military behavior
is likely to be an area of increased importance and attention.
Demography is often brought into discussions of strategy and
broad national policy, but only in the most obvious and limited
ways. William McNeill a few years ago wrote a small volume
addressing some of the broader relationships of demography to
political behavior.4 As in other of his works, he provides a number
of hypotheses and sketches out areas that deserve considerably
more attention.
Additional fields of interest are cultural anthropology,
ethnology, and some areas of psychology. In some ways a
new understanding of man is emerging, based on study of the
evolution of man and human society and on new analyses of the
biology of man, in particular the functioning of the brain. How
men process information, make decisions, and behave are central
issues on which much new knowledge exists and more will be
available in the future.
But above all, if I had a suggestion to make, it would be
that people study, in any case at least read, history of all kinds:
military history, of course, but also economic and technological
history. The history or analysis of past wars is a major antidote
to the narrow focus of many existing methods of analysis of
defense issues. Most discussion of strategy and defense programs
is, if anything, too focused on technology and weaponry and not
enough on the other factors that often dominate actual warfare.
Also, if one considers the extended competition between states
such as Rome and Carthage, the issue of why the Romans won in
the end may shed interesting light on the key variables that need
to be considered in our conceptions of strategy.

634
 Another factor of great importance is to understand the
differences in the ways in which other nations are likely to
perceive situations and react to them. Specialized studies of the
strategic cultures of Russia, China, India, Japan, Iran, and the
European nations and many others are of great use. Some of this
can be gained by reading the history of these nations, especially
the development of their military and other national security
organizations. Other aspects relate to the particular cultural
characteristics of these societies.

The Future of Strategy

We are at a major turning point in history. Uncertainty

about what the future competitive environment will be like is
especially pronounced. There are at least three major issues that
our defense or national security strategy must deal with. There
is the problem of radical Islam, which both poses an immediate
threat and has the potential to be a long-running problem. Any
serious strategy dealing with this problem will have to have a
substantial nonmilitary component. A second issue is the potential
emergence of a strong hostile China. A major problem of strategy
here is setting and articulating in some definitive way the goals
for the U.S., or a picture of what, ideally, we would like to see Asia
as a region look like in 20 or 30 years. The third major strategic
issue, I believe, is the likely proliferation of WMD (particularly
nuclear weapons) and long-range strike systems. We can of
course try to prevent proliferation, but any realistic strategy must
take account of the possibility that these efforts will fail and that
the future world will have many more nuclear powers, some of
whom would employ weapons in ways very different from how
we have tended to focus on.
Of course, a defense or national security strategy for the long
term must deal with all of these problems. It must attempt to shape
the future security environment where possible, and develop
hedges against the emergence of particular threats or problems.
There is also pronounced uncertainty about the character of future
warfare: new kinds of weapons systems are being developed,
which in turn will require the development of new doctrines, new
concepts of operations, and new kinds of military organizations
to exploit fully the new technologies. What our strategy should
be for the more complex competition that is emerging will
require consideration of many aspects of the changing security

635
environment and changing technology. We will need to know
much more than we now do about the emerging regional powers,
as well as about the likely major actors, their strategic orientation,
their strengths, and their weaknesses.
It is hoped that new centers of strategic thought and innova-
tion will arise and a new generation of strategists and military
innovators will develop to deal with these problems.

ENDNOTES - Marshall

1. Nassim Nicholas Taleb, The Black Swan: The Impact of the

Highly Improbable, New York: Random House, 2007.

2. Herbert Goldhamer, The Advisers, New York, NY: Elsevier,

1978.

3. Theory and Opposed-Systems Design (1968), Wohlstetters
essay on the theory and design of competitive systems in an earlier
part of the present volume, reflects this approach.

4. William H. McNeill, Population and Politics Since 1750,

Charlottesville, VA: University Press of Virginia, 1990.

636