|
The
Warning-Response Problem
and Missed Opportunities in Preventive Diplomacy
Alexander
L. George and Jane E. Holl
May 1997
A Report to the Carnegie Commission on Preventing Deadly Conflict
Carnegie Corporation of New York
 Carnegie
Corporation of New York established the Carnegie Commission on Preventing
Deadly Conflict in May 1994 to address the looming threats to world
peace of intergroup violence and to advance new ideas for the prevention
and resolution of deadly conflict. The Commission is examining the
principal causes of deadly ethnic, nationalist, and religious conflicts
within and between states and the circumstances that foster or deter
their outbreak. Taking a long-term, worldwide view of violent conflicts
that are likely to emerge, it seeks to determine the functional
requirements of an effective system for preventing mass violence
and to identify the ways in which such a system could be implemented.
The Commission is also looking at the strengths and weaknesses of
various international entities in conflict prevention and considering
ways in which international organizations might contribute toward
developing an effective international system of nonviolent problem
solving.
Commission
publications fall into three categories: Reports of the Commission,
Reports to the Commission, and Discussion Papers. Reports of the
Commission have been endorsed by all Commissioners. Reports to the
Commission are published as a service to scholars, practitioners,
and the interested public. They have undergone peer review, but
the views that they express are those of the author or authors,
and Commission publication does not imply that those views are shared
by the Commission as a whole or by individual Commissioners. Discussion
papers are similar to Reports to the Commission but address issues
that are more time-sensitive in nature.
Additional
copies of this report may be obtained free of charge from the Commission's
headquarters:
Carnegie
Commission on Preventing Deadly Conflict
2400 N Street, N.W.
Sixth Floor
Washington, D.C. 20037-1153
Tel.: (202) 429-7979 Fax: (202) 429-9291
E-mail: pdc@carnegie.org
THE
ARGUMENT that nothing can be done to prevent genocide or other forms
of mass violence is increasingly unconvincing. Genocide on the scale
of Bosnia or Rwanda can be anticipated and prevented. Early warning
is a prerequisite both for any prudent decision to act and for effective
action itself.
In
this concise essay, Alexander George and Jane Holl argue that leaders
need the kind of warning that will induce them to act preventively,
not simply warning that a bad situation is getting worse. Leaders
tend to put off hard decisions as long as they can, and this has
often resulted in delay or paralysis in dealing with developing
crises. To prevent violent conflicts, leaders must overcome this
initial policy paralysis.
The
events that could trigger widespread violence are usually different
from the events that trigger a preventive response from outside
parties. It would not, for example, have been possible to give an
unambiguous, precise warning that a plane crash in Central Africa
would precipitate the slaughter of nearly one million people. But
many earlier indications of the possibility of genocide in Rwanda
in 1994 were ignored, and no preventive plan of action was in place.
As George and Holl point out, outside parties must become more receptive
to warning.
Early
warning will not ensure successful preventive action unless there
is a fundamental change of attitude by governments and international
organizations. Third parties should not simply wait for unambiguous
disasters and mass slaughter before they take preventive action.
Rather, a systematic and practical early warning system should be
combined with consistently updated contingency plans for preventive
action that provide leaders with a repertoire of responses. This
would be a radical departure from the present system, where when
a trigger event sets off an explosion of violence, it is usually
too difficult, too costly, and too late for a rapid and effective
response. This early warning system would be a crucial component
of the international preventive framework envisioned by the Carnegie
Commission on Preventing Deadly Conflict.
The
Commission has always defined broadly the groups that would participate
in such a system. States, nongovernmental organizations, business
enterprises, religious leaders, scientific groups, the media, and
international organizations all have a role to play in providing
early warning and in responding to warning. Logically, early warning
should be given first to those who can take action. This generally
means governments and groups likely to be immediately involved in
the crisis, governments and leaders nearest to the scene of conflict,
the United Nations (particularly the member states of the UN Security
Council), and regional organizations. Religious hierarchies may
also be warned, particularly of situations in which local religious
leaders and institutions could play positive roles. In addition,
those who can induce governments, organizations, and agencies to
act (the media, business communities, and concerned publics) should
be kept informed of badly deteriorating situations. Public expectations
that governments will act responsibly to ward off disasters are
a significant factor in motivating preventive actions.
This
is one of several studies of the warning–response problem that
the Commission is sponsoring. The role of leaders in responding
to warning will be illuminated by Boutros Boutros-Ghali, George
Bush, Jimmy Carter, and Mikhail Gorbachev in a collection of essays
to be published in the fall. Professor Bruce Jentleson, director
of the University of California, Davis, Washington Center, is editing
a set of 11 case studies of preventive diplomacy in the post–Cold
War world that will appear in the Commission book series published
by Rowman & Littlefield. Recognizing the important and sometimes
ambiguous role of the media in providing warning and motivating
response, the Commission has asked three distinguished journalists,
Tom Gjelten, Nik Gowing, and Robert Manoff, to present their views
in an essay collection. The Commission is also sponsoring several
studies of instruments—sanctions, incentives, mediation, and
the military, for example—to enlarge and sharpen leaders' repertoire
of responses.
The
Commission is grateful to Alex George and Jane Holl for advancing
the thinking on this crucial aspect of conflict prevention.
David
A. Hamburg
Cyrus R. Vance
Cochairs
SPECIALISTS
MAY DISAGREE on the scope of preventive diplomacy and, more broadly,
preventive measures of various kinds. They may differ also in their
assessment of policies and strategies to ward off undesirable events.
There is no disagreement, however, on the importance of obtaining
early warning of incipient or slowly developing crises if preventive
action is to have any chance of success.
The
end of the Cold War has diminished neither the importance nor the
challenge of obtaining early warning. Indeed the intelligence community
today monitors and analyzes an increasing number of factors, in
addition to traditional indicators of potential conflict, such as
environmental degradation, economic conditions, and population trends.
The increased complexity of gathering, sorting, and analyzing data
for early warning results from the pressing need to respond quickly,
efficiently, and effectively to rapidly changing global events.
In an era of increasing demands on limited resources, the task is
all more difficult.1
In
recent years the problem of obtaining early warning has received
a great deal of attention not only within the United Nations, regional
organizations, and governments, but also from nongovernmental organizations
(NGOs) and research specialists.2 However,
the more difficult problem of marshaling timely, effective responses
to warning has received much less systematic attention. A major
objective of this paper is to highlight this need for more emphasis
on developing effective responses for preventive action of various
kinds.3 This paper also emphasizes that
the design and management of early warning systems should be intimately
connected with the task of responding to warning. We base this view
on the belief that an improved capacity to know about and correctly
interpret events early will improve the responses that are brought
eventually to bear—a belief that is shared by a range of policy
professionals, government officials, and informed publics. The paper
does not offer specific policy recommendations for overcoming the
gap between early warning and effective response; rather, it provides
a conceptual approach through which to analyze the problem. We conclude
the paper with a discussion of how warning and response interact
in policymaking. When successful, that interaction can help avert
violence. When unsuccessful, the result is often looked upon as
a "missed opportunity." We discuss such missed opportunities,
but with reservations, not least because of the dangers associated
with counterfactual analysis. However, well-crafted examinations
of missed opportunities for preventive diplomacy can be useful in
bringing to light and learning from past warning–response failures.
Too
much of the considerable effort to develop improved warning indicators
has been divorced from the problem of linking available warning
with appropriate responses. One explanation for this separation
may stem from the stark lines drawn between collection and analysis
in the intelligence community.4 Perhaps
there is reason for this separation, for this approach may be traced
to the increased professionalization of the intelligence field,
where intelligence analysts assiduously ward off any hint that they
"do policy." They focus their efforts instead on improving
the ways in which information is acquired and analyzed.5
Another explanation may lie in the very difficulty of policymaking
in today's international environment. It may simply be beyond the
capacity of any single office or agency to stay abreast of global
developments in such a way as to anticipate, craft, launch, and
manage intricate, multilateral policy responses.
But
whatever the institutional causes of the warning–response gap,
expectations that governments will act responsibly to help ward
off possible crises are quite real.6
These expectations arise, in part, because an increasingly mobile
world population combined with the explosion of global communications
(the so-called CNN effect) have helped create and inform attentive,
expert, and often activist communities in many countries who know
about problems before they become violent. In part as a consequence,
it has become less plausible for government officials to try to
explain away policy missteps or failures by pointing to the lack
of timely or correctly evaluated intelligence, although the urge
remains almost irresistible.7
The
complexity of world events combined with the compressed time span
within which decision makers are expected to craft and articulate
a policy to deal with unfolding crises make it harder, yet at the
same time more necessary, for intelligence analysts and policymakers
to work within an integrated "warning–response" framework.
Indeed, the need for such an integrated approach was the fundamental
lesson drawn from the surprise attack on Pearl Harbor and provided
the starting point for post–World War II efforts to design
systems and procedures for avoiding such a lapse.8
As
with the need to respond effectively to avoid a surprise attack,
preventive action to deter the outbreak of various post—Cold
War crises also demands an integrated warning–response framework.
Yet, for such crises, the warning–response problem is often
more complicated and difficult than for avoiding surprise attack.
In the latter case, policymakers have already determined that some
set of observable hostile actions would be an unmistakable threat
and have the strongest possible incentives to acquire timely warning
and to respond to that threat in some way. The same cannot be said
for many lesser contingencies, such as ethnic conflicts or patterns
of gross human rights abuses. Since situations of this kind—even
in crisis—pose a much less grave threat to the interests of
a third party, policymakers are often less inclined to demand early
warning or to take it seriously and respond to it.9
But
one may wonder whether there have been many crises for which no
warning was available, however misperceived, misjudged, or ignored.
Experts predicted war in Bosnia even as the Vance plan brought a
cessation of hostilities between Croatia and Serbia in 1992. The
violent spasm in Rwanda in 1994 was anticipated months in advance,
although the magnitude of the killing was not precisely foreseen.
Even Saddam Hussein's precipitous invasion of Kuwait in 1990 was
no great surprise to those who watch events in the Middle East closely.10
If events such as in Bosnia, Kuwait, and Rwanda are known (and increasingly
knowable, given the rapidly contracting nature of global interactions),
why are they not prevented? No simple answer is possible, yet a
partial explanation may lie in the examination of how warnings are
recognized and transmitted to policymakers, and with policymakers'
assessments of the implications of such warnings for action.
Receptivity
to warning has been a problem not only for conflicts that occur
on the margin of states' interests but also for situations threatening
a surprise attack. Although the reasons for inadequate receptivity
and response to warning differ in some ways for these two types
of threats, it will be useful first to review experience with the
problem of receptivity to warning of possible surprise attack and,
related to this, to unexpected diplomatic initiatives that trigger
the possibility of war. Properly scrutinized and evaluated, this
historical experience may be suggestive for the design and use of
warning–response systems for preventive action for other types
of crises.
Experimental
research provides a useful starting point for analysis of factors
that impede receptivity to warning. Laboratory studies of difficulties
in perception of stimuli provide useful analogies to the problem
of receptivity to warning of emerging threats in the international
arena. The results of perception experiments, however, do not encourage
hopes for easy or complete solutions to this problem. Studies of
a person's ability to recognize a stimulus that is embedded in a
stream of other stimuli have shown at least three factors to be
important:
- The
"signal-to-noise" ratio—i.e., the strength of the
signal relative to the strength of the confusing or distracting
background stimuli
- The
expectations of observers called upon to evaluate such signals
- The
rewards and costs associated with recognizing and correctly appraising
the signal
One
might assume that the stronger the signal and the weaker the background
"noise," the easier it should be to detect the signal;
weak signals are simply not picked up. However, even controlled
laboratory tests reveal the task of correct signal detection to
be more complicated than this. The results of perceptual experiments
that deal with relatively simple psychophysical auditory or visual
stimuli indicate that detection of a signal is not simply a function
of its strength relative to background "noise." Indeed,
the effect of a signal's strength on the ability to identify it
can be less important than the second and third variables mentioned
above.
The
complex environment of international affairs only complicates matters
further, adding domestic and international overlays to the basic
"map" of the crisis situation. A decision maker's expectations
and the rewards and costs associated with recognition of the signal
may be more important in determining receptivity to and correct
appraisal of information about an emerging threat.
But
while expectations regarding both the emerging crisis and the potential
responses play a key role in a decision maker's receptivity to warning,
the logic of warning and the logic of response conflict. The logic
of warning can be summarized as "the sooner the better."
However, policymakers generally prefer to put off hard choices as
long as possible. Thus, even if a leader expects a situation to
deteriorate, additional information or warning to this effect may
not prompt preventive action.
Because
policy choices in a crisis are often so difficult to make, individuals
(as well as small policymaking groups and organizations) may discredit
information that calls into question existing expectations, preferences,
or policies. It is well known that discrepant information of this
kind is often required, in effect, to meet higher standards of evidence
and to pass stricter tests of admissibility than new information
that supports existing expectations and policies. As a result, it
is disconcertingly easy at times for policymakers and their intelligence
specialists to discount discrepant information or to interpret it
in such a way as to protect a preferred hypothesis or policy. In
the United States, the establishment of multiple intelligence organizations,
with their capacity for redundancy and rich detail, was designed,
in part, to counter this tendency. Yet the habit persists. Indeed,
not only is the discrepant information still discounted, but entire
intelligence organizations can be discounted.12
The
"reward-cost" aspect of correct signal detection, too,
can sharply reduce the policymaker's receptivity to information
of emerging threats, for early warning does not necessarily make
for easy response. On the contrary, warning often forces policymakers
to confront difficult or unpalatable decisions. One means for
avoiding such difficult decisions is to reduce one's receptivity
to warning signals. Moreover, the policy "background"
against which new information is judged can strengthen the tendency
to ignore or downgrade incoming information that challenges existing
beliefs or exacerbates decision dilemmas. Thus, once policy decisions
have been made within the government, they tend to acquire a momentum
of their own and the support of vested interests. Top-level decision
makers are often reluctant to reopen policy matters that were decided
earlier with great difficulty; to do so, they fear, can be taken
as an indirect admission of policy failure and easily plunge the
government once again into the turmoil of decision making.
Psychological
mechanisms of this kind have contributed to a number of important
intelligence and policy failures. Among them was the Truman administration's
pronounced lack of receptivity to the ample warning available in
the spring of 1950 of the forthcoming North Korean attack on South
Korea. As studies have shown, had the warning been taken more seriously,
the administration might have weighed more carefully whether the
perceived stakes in Korea warranted U.S. military intervention.13
If an affirmative answer to this fundamental question had emerged,
the administration might have undertaken to deter North Korea. As
it was, the North Koreans acted as they did on the mistaken notion
that the United States would not intervene militarily on behalf
of South Korea. Thus, the Korean War, with all of its fateful consequences,
qualifies as a genuine example of war-through-miscalculation. It
was a war that might well have been avoided had Washington been
more receptive to warning and acted upon it.14
This
case illustrates how information processing within the U.S. policymaking
system was impeded and distorted both by the expectations or mind-set
of the administration and by the costs that greater receptivity
to incoming information of the emerging threat would have entailed.
Taking available warning seriously always carries the "penalty"
of deciding what to do about it. In this case, it would have
required President Truman and Secretary of State Dean Acheson to
reconsider the earlier decision that they had made in 1949 to draw
a line defining U.S. security interests in the Far East to exclude
Formosa, South Korea, and Indochina. The exclusion of Formosa was
part of the administration's policy of disengaging from the Chinese
Nationalists, a decision that was far more controversial within
the administration and with the public than the exclusion of South
Korea. So much so that a reversal of the existing policy of no military
commitment to South Korea in response to the warning of a possible
North Korean attack would have been politically inconceivable unless
Truman and Acheson had also been willing—which they were not,
prior to the North Korean attack—to extend a new commitment
to the Chinese Nationalist regime on Formosa as well.
As
this case and others show, the policy background at the time warning
becomes available may subtly erode the policymaker's receptivity
to it. A similar misfortune occurred later in the Korean War. During
September and early October 1950, the administration eased itself
into a commitment to occupy North Korea and to unify it with South
Korea. But when repeated warnings came in that such a move would
trigger Communist Chinese military intervention, the administration
found itself so locked into its more ambitious war policy that it
dismissed the warnings as a bluff. To give credence to the worrisome
indications of a forthcoming Communist Chinese intervention carried
with it the cost of reconsidering and abandoning the war policy
that had given rise to the danger. In this critical situation, wishful
thinking contributed to the administration's grossly defective information
processing. Once again the result was that Washington was taken
by surprise when the Chinese launched their massive offensive in
late November. A new war resulted that neither side had wanted,
one that might have been avoided had Washington not misperceived
and misjudged the evidence of Chinese intentions. 15
Similarly,
in the spring of 1948, most American policymakers refused to take
seriously the possibility of a Soviet blockade of West Berlin despite
mounting tension and the fact that the Soviets had recently imposed
a temporary blockade of Western ground access to the city. Some
of the same psychological dynamics that interfered with optimal
processing of incoming information in the cases already described
can be seen here, too. For U.S. policymakers to have taken available
warning of a possible Soviet blockade of West Berlin seriously would
have carried with it the "cost" of having then to face
up to and resolve difficult, controversial policy problems.
At
the time an American commitment to West Berlin did not yet exist.
Officials within the administration were badly divided over the
wisdom of attempting to defend the Western outpost that lay deep
in Soviet-occupied East Germany. Under these circumstances, it was
easier to believe the Soviets would not undertake serious action
against West Berlin than it was to decide beforehand what the American
response should be to such an eventuality. In this case, fortunately,
although American policymakers were surprised by the Soviet blockade,
Truman dealt with the crisis without backing down or going to war.16
The
August 2, 1990, Iraqi invasion of Kuwait offers a more recent example
of the difficulty of correctly reading an adversary's signals. By
mid-July of 1990, U.S. intelligence had identified the buildup of
some 35,000 Iraqi troops and 300 tanks on Kuwait's border. At the
same time, Iraq was bringing charges before the Arab League that
Kuwait had, among other things, broken OPEC oil production quotas
and stolen oil from Iraqi territory. In compensation, Iraq demanded
an increase in the price of oil (from $18 to $25 a barrel), $2.4
billion from Kuwait, and a moratorium on Iraqi debts to other Arab
states stemming from the Iran–Iraq War. Should the demands
not be met, Saddam Hussein threatened that he would "have no
choice but to resort to effective action to put things right and
ensure the restitution of [Iraqi] rights." 17
Through the latter portion of July, U.S. intelligence continued
to monitor Iraqi troop advancements. By the end of the month, 100,000
troops had been assembled on the Kuwaiti border, accompanied by
strategic deployments of ammunition and supplies. These moves, together
with other ominous signs, such as the continued buildup of biological
and chemical weapons and strong evidence of a nuclear weapons development
program, highlighted the threat posed to the region and vital U.S.
interests.18
Analysis
of Iraqi intentions differed within the intelligence and diplomatic
communities. Even the Kuwaitis at first believed Hussein was merely
bluffing to gain economic concessions. Analysts tracking the situation
within both the Central Intelligence Agency (CIA) and the Defense
Intelligence Agency (DIA) eventually concluded (by July 25 and July
30, respectively) that Iraq intended to invade Kuwait. Even at this
late date, however, high-ranking officials in the intelligence and
military communities remained skeptical of the invasion analysis,
believing instead that Iraq was likely to make only a limited border
crossing.19
American
diplomatic response to the Iraqi troop movements was equivocal.
Bush administration officials repeatedly stated that the U.S. had
no defense treaties with Kuwait or other Arab states threatened
by Iraq. The U.S. ambassador to Iraq reportedly told Hussein that
"we have no opinion on the Arab–Arab conflicts, like your
border disagreement with Kuwait."20
At no point was Iraq told what the consequences would be should
it attack Kuwait or other Gulf states. Many now believe that the
absence of a clear response led Iraq to believe that its invasion
of Kuwait would be met with little resistance by the international
community, and more specifically, the U.S. 21
These
several lessons of historical experience regarding lack of receptivity
and inadequate response to warning of surprise military or diplomatic
actions are applicable also to the different kinds of threats in
the post–Cold War world that effective preventive action must
address.
The
Rwandan conflict offers another, brutal, example of the difficulties
associated with generating effective responses to the types of conflict
dominating the post–Cold War era—situations that do not
threaten a nation's vital interests.
"Most
leading activists believe that the government has compiled lists
naming people to be assassinated when circumstances require."22
So reported Africa Watch in a 1992 report highlighting human rights
abuses and tensions between the Hutu majority and Tutsi minority
in Rwanda. Beginning on April 6, 1994, these lists were used as
part of a killing spree that would, in a matter of weeks, take the
lives of nearly one million people. The significant presence of
international organizations (the UN and the Organization of African
Unity) and representatives of key donor countries (including France,
Belgium, and the United States) ensured that warning of the developing
crisis was received by prominent actors in the international community.
Despite this significant presence and ample evidence of deteriorating
circumstances in Rwanda, there was an acute failure to respond.
A number of factors contributed to this failure. According to one
report:
There
existed an internal predisposition on the part of a number of the
key actors to deny the possibility of genocide because facing the
consequences might have required them to alter their course of action.
The mesmerization with the success of Arusha [The 1993 peace accord
between the Hutu- dominated government and the Tutsi-dominated Rwandan
Patriotic Front] and the failure of Somalia together cast long shadows
and distorted an objective analysis of Rwanda.23
Among
the more obvious warning signals were "hate radio" broadcasts
directed at Tutsis and moderate Hutus, continued training of Hutu
militia units, and government-sponsored killings. Yet none of the
major outside actors formulated, let alone articulated, a response
to the potential outbreak of widespread violence. According to Human
Rights Watch consultant Alison Des Forges, a particularly important
event was the February 1994 murder of a moderate Hutu cabinet member
by government soldiers. Des Forges noted, "when they [Hutu
extremists] saw they could get away with that kind of violence .
. . it encouraged them to go ahead with the larger operation."24
While
the foregoing discussion of receptivity to warning has been necessarily
brief, it indicates that the impediments are numerous and that they
cannot be easily eliminated. For this reason, most specialists have
urged that the problem of securing and analyzing warning should
be linked closely with the problem of deciding what responses are
appropriate and useful in the light of the available warning, however
equivocal or ambiguous it may be. While high-confidence warning
is desirable, often it is not available. But neither is high-confidence
warning always necessary for making useful responses to the possibility
of an emerging crisis.
Indeed,
this discussion of receptivity to warning of emerging threats applies
also to information about favorable developments elsewhere in the
world that offer opportunities for foreign policymakers to
advance positive goals. For many purposes, policymakers do not need
or require high-confidence forecasts of emerging opportunities in
order to explore and facilitate such openings and possibly to turn
them to account. Thus, for example, following the Soviet invasion
of Czechoslovakia in 1968 and the enunciation at that time of the
Brezhnev Doctrine, policymakers in Washington (as well as other
observers) speculated that these events may have increased China's
anxiety regarding a Soviet invasion. Was this anxiety (which its
ongoing border conflict with the Soviet Union could only have heightened)
sufficient to make China interested in d§tente with the United
States? We cannot be sure of Chinese thinking at that time, but
the point that deserves emphasis here is that it did not require
a forecast that could confidently predict Beijing's readiness for
d§tente to make it worthwhile for Washington to explore and
encourage the possibility discreetly. Sensible steps could be taken
to reinforce and activate any disposition for d§tente on
the part of the Chinese. From the standpoint of U.S. policy, the
matter of possible d§tente was "actionable" even
in the face of considerable uncertainty as to China's readiness
and conditional willingness to reorient its policy towards the United
States.25
We
have noted that policymakers are often not inclined to take early
warning seriously or to act upon it in situations that pose the
possibility of severe ethnic and religious conflicts, humanitarian
disasters, or gross human rights violations.26
A number of reasons exist for this passivity. The first is the relatively
low stakes perceived to be at risk. At an early stage in their development,
such contingencies simply are not perceived to pose grave threats
to a given state's national interests. Moreover, whether a low-level
conflict or incipient crisis will escalate in ways that would eventually
engage major interests of individual states or the international
community often remains problematical and difficult to forecast.
Second,
despite efforts to improve early warning indicators of possible
flare-ups, such events are likely to remain equivocal, subject to
considerable uncertainty, and capable of diverse interpretations.
It is not that potential major trouble spots cannot be identified;
rather, the problem lies in understanding such situations well enough
to forecast which ones are likely to explode and when. Experts and
observers are likely to differ in their estimates of how serious
a low-level situation will become, with what probability, and how
soon.
Third,
early warning indicators typically do not speak for themselves;
they require analysis and interpretation. But the kinds of knowledge
and theories needed for this purpose may be in short supply. As
noted earlier, specialists have worked more on improving possible
indicators than on developing better theories and models to assess
and predict the significance of the indicators.27
Fourth,
even in a case in which there is relatively good warning, policymakers
may be reluctant to credit the warning and to take preventive action
because they have been subjected too often to the "cry wolf"
phenomenon. Oddly enough, intense policy concerns that actions may
be seen as premature or unnecessary—revealing an embarrassing
policy naāvet§, or worse, the possible unneeded commitment of scarce
resources—generate a real wariness of "false triggers."28
These policymakers, typically preoccupied with a battery of other
problems that require urgent attention, often give only the barest
attention to new, low-level crises that may never develop into serious
concerns.
Fifth,
and related to this, overload induces passivity. Given the large
number of simmering crises, and given the ever-growing limitation
of resources, policymakers find it impractical to respond with preventive
actions to all of them, thinking that is reinforced by the general
lack of knowledge regarding what efforts would be effective. Early
warning of an equivocal, uncertain nature in such situations is
insufficient for costly or risky responses.
Thus,
in many ethnic and religious conflicts, humanitarian crises, or
severe human rights abuses, timely or accurate warning may not
be the problem at all. Rather, for one reason or another, as
noted, no serious response is likely to be taken solely on
the basis of early warning simply because a simmering situation
that threatens to boil over may not be deemed important enough to
warrant the type and scale of effort deemed necessary to prevent
the hypothetical catastrophe. Moreover, this reaction can occur
not only when what is at stake is only dimly perceived or not foreseen
at all, but also if the coming crisis is fully and accurately
anticipated.
Indeed,
sixth and finally, it may be that a reluctance to act in the face
of warning at times results not because warning is not taken
seriously, but rather because decision makers take it very seriously
but are nonetheless deterred by the prospects of a "slippery
slope," that is, inexorable—and potentially intractable—involvement
in an already nasty problem. This dilemma is particularly poignant
for political leaders who must weigh incurring political costs now
(in addition to the human and material costs that action entails)
for benefits that will accrue downstream, if at all, with no guarantees
that they would be given credit for preventing a disaster, now a
non-event. Thus, even in cases where the prospect of a catastrophe
is taken seriously, there may be a lack of "political will"
to take timely and effective action.
Numerous
observers have noted that governments often ignore an incipient
crisis until it has escalated into a deadly struggle or a major
catastrophe. All too often political leaders find it difficult to
persuade their people to support potentially costly and risky operations
before a disaster actually occurs. As one report put it:
People
throughout the world tend to be guided by the media—and they
are predominantly Western media—in determining when a problem
warrants international action. Television coverage of a situation
has become, for many, a precondition for action. Yet for most commercial
networks, the precondition for coverage is crisis. There has to
be large-scale violence, destruction, or death before the media
takes notice. Until that happens, governments are not under serious
internal pressure to act. And by then, the international community's
options have usually been narrowed, and made more difficult to implement
effectively.29
But
as noted earlier, even when events that could precipitate a major
humanitarian or violent crisis are perceived in a timely manner
and accurately evaluated, decision makers will often still defer
taking preventive action. As we have seen, this inaction is either
because the warning is not taken seriously, for the reasons mentioned,
or because the warning is taken very seriously but decision makers
are loath to confront the unpalatable choice of responses facing
them. Particularly for the complex and seemingly intractable disputes
that have characterized much of the violence of the post–Cold
War period, it may be less the unfolding crisis that conditions
how a decision maker processes warning than the implications of
that crisis for action.
However
a policymaker responds to warning, that response entails costs and
risks of its own: indeed some responses could even be quite harmful.
There is clearly a need to search for responses to warning that
are useful in the situation without posing unacceptable costs. Even
ambiguous warning, for example, gives policymakers more time to
consider what to do: to step up efforts to acquire more information
about the situation, to rehearse the decision problem that they
would face if the warning proves to be correct, to spell out the
likely consequences if the equivocal warning to which low probability
is assigned proves to be genuine, to review their commitments and
contingency plans, and—not least in importance—to seize
the opportunity to avert a possible dangerous crisis. Thus, even
ambiguous warning provides an opportunity to deal with the conflict
situation and/or the misperceptions associated with it before it
leads to a violent conflict.
Nevertheless,
it is a truism to note that policymakers prefer to receive unequivocal
warning before deciding whether and how to respond. But, as noted
earlier, high-confidence early warning is seldom available, and
it can be highly disadvantageous if policymakers defer action altogether
until more conclusive warning is available. It is precisely because
unambiguous warning is so difficult to obtain that policymakers
must confront the question of what types of response are useful
and acceptable, even though the warning is uncertain or equivocal.
As
noted earlier, once the problem of warning is linked with its implications
for action, it becomes significantly redefined. Early warning of
a possible crisis is desirable not in and of itself but insofar
as it provides decision makers with an opportunity to make a timely
response of an appropriate kind that might be otherwise impossible.
Warning gives the decision maker time to decide what to do and then
to prepare to do it. Warning provides an opportunity to avert the
expected crisis, to modify it, or to redirect it into some less
dangerous and less costly direction. On occasion, warning may provide
an opportunity to deal with a conflict-of-interest situation or
misperceptions before they lead to a military conflict.
Consideration
of the warning–response problem requires that we introduce
another dimension into the analysis at this point. Since response
to warning is never without cost or risk, the development of warning–response
systems, contingency response options, or ad hoc responses
requires careful consideration of the possible costs as well as
of the expected benefits of each option, weighed, of course, against
the costs and benefits of inaction. At the same time, there are
undoubtedly some responses to early warning of an equivocal and
ambiguous character that are less costly than others. One could,
for example, quietly intensify the collection of intelligence and/or
begin discreet consultations with selected allies in order to clarify
an uncertain situation before "going public" with more
assertive measures, such as placing forces at increased readiness.
Admittedly,
some low-cost responses may make only a limited or uncertain contribution
to dealing with a troublesome situation. There may be, in other
words, a trade-off between responses that promise a great deal but
are costly and risky, and responses of a more modest but still usefulkind
that do not pose large costs and risks. The experience with trade-offs
of this kind in dealing with the problem of surprise attack may
be suggestive. In part, the trade-off dilemma in these cases can
be dealt with by developing a calibrated warning–response system,
one in which the level-of-readiness response increases with the
level or urgency of warning.
For
special historical reasons related to the trauma of Pearl Harbor,
as noted earlier, American analysts concerned with the warning problem
have focused attention primarily upon the danger of a surprise all-out
military attack. Lesser types of threats and crises associated with
the broader, and in many ways, more complex tasks of preventive
diplomacy and preventive actions have not yet received as much systematic
attention in efforts to develop warning–response systems. Thus,
the major uses of warning contemplated by the U.S. planners in the
past have focused upon (a) the use of warning to alert military
forces in order to reduce their vulnerability and to shorten their
response time; and (b) the use of warning to reinforce deterrence
by signaling to the adversary a strong and credible commitment to
respond.
A
broader range of threats and types of crises should engage the interest
of policymakers and specialists on crisis anticipation. Similarly,
a broader range of response options than the two uses of warnings
noted above should be developed.30
A longer, more diversified list of possible uses of warning would
include, but are not limited to, the following (general response
options are listed here without attempting to judge their utility
in any particular situation):
- Gather
more information about the situation. Step up collection of intelligence
and public information.
- Reduce
vulnerabilities. Alert forces and citizens abroad to reduce their
exposure and susceptibility to attacks of all kinds. Increase
readiness of standby forces and alert special forces for contingency
operations.
- Reinforce
commitments. Strengthen deterrence, whenever necessary, by signaling
credible "red lines" that should not be crossed, using
diplomatic means and, if necessary, military demonstrations.
- Engage
the targeted state in sustained dialogue. Establish clear
and reliable channels for exchange of communications.
- Take
measures to reduce potential political/diplomatic/economic costs
that could result from the emerging crisis in the domestic or
international arena.
- Conduct
consultations with key states and allies. Raise the issue in the
United Nations and other appropriate international forums.
- Undertake
a public information campaign to inform populations at home and
abroad of the unfolding circumstances. Prepare publics for possible
coercive diplomacy or military action.
- Conduct
a decision rehearsal, i.e., rehearse the decision problem that
one would be confronted with if the warning proved justified.
A rehearsal involves (a) assessing the damage to important interests
should the crisis erupt (something that policymakers have done
very poorly in some past crises); and (b) anticipating the political
and psychological pressures that are likely to be brought to bear
upon policymakers should the crisis occur.
- Consider
and, if necessary, clarify one's commitment to take action should
the crisis emerge. Warning can have the useful function of encouraging
policymakers to identify and assess the complex interests that
may be jeopardized if the crisis develops. Such a review may also
result in a timely redefinition or clarification of existing commitments,
identifying and separating issues that are peripheral and negotiable
from those that are central.
- Review,
update, and rehearse existing contingency plans. Improvise new
policy options tailored to the emerging crisis, taking into account
potential actions of other states with interests at stake.
- Initiate
formal negotiations, efforts at conciliation, or mediation. On
many occasions, for example, the UN secretary-general's office
responds to early warning by sending out fact-finding missions
or by extending "good offices."
The
preceding list of response options characterizes in general terms
the types of responses available to decision makers and is intended
for illustrative purposes. More specific options must be identified
in policy planning tailored to the type of situation and problem
that is envisaged by the warning. Obviously, different types of
incipient crises will require identification of different response
options.31
This
brief list should not obscure the implied steps that each measure
entails. For example, using military demonstrations to underscore
one's seriousness of purpose must be balanced against the desire
to control the level of engagement (and avoid a "slippery slope").
So
much of this list seems like straightforward policymaking. What
we mean to emphasize, however, is the need for an explicit effort
to map various responses to anticipated developments—before
those developments occur—and to associate particular response
options more closely with foreseeable cues. 32
Those
who call attention to failures to take timely, appropriate actions
in response to early warning of an emerging crisis often refer to
them as missed opportunities. The clear implication is that it might
well have been possible to avoid or limit the development of a major
crisis—whether a violent ethnic or religious conflict, a humanitarian
catastrophe, or a gross human rights violation—if only the
international community or an external actor had intervened.
A
word of caution may be in order. "Missed opportunities"
implies that the "misses" constitute important policy
failures of various kinds. Indeed, it is difficult, if not
impossible, to avoid the analytic conclusion that such "failures"
contributed measurably to a worsened situation on the ground. This
assumption, that a crisis situation is the measure against
which policy decisions and their aftermath are judged, may contribute
to analytic clarity, but it fails to represent adequately all of
the factors that constrain policy decisions—especially in times
of crisis. Indeed, as we have tried to illustrate, factors unrelated
to the crisis situation (domestic elections, credibility and other
strategic concerns, or other international problems) can affect
a decision maker's receptivity to warning more than the circumstances
causing the alarm—even when warning is "loud and clear."
Moreover, these other factors are frequently perceived by decision
makers not only to be legitimate to take into account, they are
often seen as more legitimate considerations than circumstances
on the ground. Indeed, decision makers most closely associated with
many of these so-called missed opportunities resulting in policy
"failures" often strongly resist that indictment, arguing
instead that their action (or wise restraint) was in the best interest
of the public that they serve. Thus, even as the following discussion
focuses on the crisis situation as the main measure of the effectiveness
of actions taken (or not), we recognize the tensions that exist
within the full context of these situations.
The
assertion that a missed opportunity occurred is an example of counterfactual
reasoning, a practice that is very frequently resorted to in everyday
life as well as in serious analysis of historical outcomes. However
widespread and indeed indispensable, counterfactual analysis is
recognized to be a very weak, problematical method. This is not
the occasion to discuss recent efforts by scholars to identify requirements
for more disciplined uses of counterfactual reasoning.33
Suffice it to say that statements that missed opportunities occurred
in cases of failure of preventive diplomacy must be evaluated carefully
to distinguish highly plausible from implausible or barely plausible
claims. Efforts to do so are necessary not merely to improve historical
analysis of cases in which preventive diplomacy was not attempted
or was ineffectual; more rigorous counterfactual analysis is necessary
also to draw correct lessons from such failures.
A
useful start in this direction can be made by distinguishing different
types of missed opportunities. The following is a provisional (no
doubt incomplete) listing:
- Cases
in which there was no response to warning by policymakers,
who either ignored the warning or regarded it as insufficiently
reliable, too equivocal, or uncertain (Example: Iraq's 1990 invasion
of Kuwait).
- Cases
of inadequate analysis of ample warning indicators, and,
thus, an inaccurate forecast of what was to occur (Examples: the
1979 Iranian revolution; the North Korean attack on South Korea
in June 1950).
- Cases
of inadequate response to warning, either too slow or too
weak (Examples: slow international response to the developing
crisis in Somalia; slow, graduated sanctions against Serbia).
- Cases
of misused opportunity involving responses of a misconceived,
harmful, inappropriate character (Example: European Union recognition
of Croatia without securing a prior guarantee of the rights and
interests of its substantial Serbian minority).
- Cases
of inconsistent responses (Example: In the unfolding crisis
in Yugoslavia, European countries were often at cross-purposes,
such as in 1991 when they tried to serve as mediator between Serbia
and Croatia while pushing international recognition of Croatia
and the imposition of sanctions on Serbia.)
- Cases
of incomplete response to a complex crisis (Example: Somalia,
where the international community undertook to deliver humanitarian
assistance but refused to engage in peace enforcement efforts.)
- Cases
of contradictory responses (Example: Efforts by some states
to install peacekeepers in Nagorno-Karabakh undermined by other
states opposed to such a move.)
In
addition to some such typology of different types of possible missed
opportunities, we need, as noted earlier, some way of assessing
the merits of claims that there was indeed a missed opportunity
to avoid the disaster that followed. Counterfactuals are a way of
rewriting history (exploring the possibility of an alternative outcome)
by conducting a mental experiment—i.e., "if only this
rather than that had been done, the outcome would have been quite
different." Some counterfactual assertions are more plausible
than others. Those of us who believe in the necessity for timely
responses to early warning may inadvertently exaggerate the plausibility
of a missed opportunity in cases that developed into major conflicts
or severe humanitarian catastrophes.
Several
suggestions can be made for assessing the plausibility of assertions
of a missed opportunity. A basic distinction needs to be made between
two connotations of "opportunity." One use of the term
implies that a significantly better/good outcome would surely
have been achieved if it were not for . . ., or if only this rather
than that had been done. A weaker connotation of the term "opportunity"
is that a better outcome was possible; it might have
been achieved if . . . A still weaker connotation states merely
that a better outcome was possible but without indicating what might
have been done to secure it. In making assertions of a missed opportunity,
and of course, in evaluating such claims, it is important to keep
this distinction in mind. Frequently, critics who identify a missed
opportunity blur this distinction.
Admittedly,
it is often difficult to judge the degree of confidence that can
be ascribed to what appears to have been a missed opportunity. Practitioners
who engage in efforts at preventive diplomacy may well regard these
distinctions as an academic exercise. It must be recognized that
those who engage in preventive actions often do so without demanding
of themselves that they be able to predict outcomes with high confidence;
they make what they regard to be appropriate efforts and use what
leverage they have to influence the course of events. They reason
that when the stakes are high, one must make efforts to influence
the course of events even when prospects of success are highly uncertain.
It is only human to believe that adverse outcomes might have been
avoided or moderated, if only . . . .
Such
explanations for what may be dubious claims on behalf of a particular
missed opportunity leave us with the task of developing reasonable
ways of evaluating them. To construct a good counterfactual analysis
of a missed opportunity one needs to start with a good explanation
of the actual outcome of the case at hand. This step is important,
obviously, because the counterfactual changes what are thought to
be the critical variable(s) that presumably accounted for the historical
outcome. If one has an erroneous/unsatisfactory explanation for
it, then the counterfactual analysis that argues that a better outcome
was possible, "if only . . .," is likely to be flawed.
Both the historical explanation and the counterfactually derived
alternative to it are likely to be more correct or plausible if
they are supported by relevant generalizations (and theory).
In
formulating hypothetical missed opportunities and in evaluating
them, at least two questions need to be addressed: First, was the
alternative action possible at the time and known to be possible,
or was it something that one sees only in retrospect. If the latter,
then the claim of a missed opportunity is weakened since it rests
on the argument that alternative action could have and should have
been seen at the time. Missed opportunities that rest too heavily
on hindsight carry less plausibility but, of course, such claims
should not be dismissed if one wants to draw useful lessons from
such experiences. An after-the-fact identification of an action
or strategy not known or considered at the time can still be useful
in drawing lessons.
Missed
opportunities differ, too, depending upon whether the alternative
is a simple, circumscribed action or whether it is a sequence of
actions over time. In the latter case, counterfactual reasoning
involves a long, complex chain of causation involving many variables
and conditions, all of which would have to fall into place at the
right time for the missed opportunity to be realized. The plausibility
of a missed opportunity is enhanced, in contrast, when the chain
of causation is shorter and less complicated. A missed opportunity
is obviously less plausible when it rests on the belief or expectation
that a different set of actions could have occurred over time and
overcome a series of obstacles, thereby achieving a successful outcome.
The
second question: Was there at least one or a few decisive turning
points? Those who take a "path dependent" view of history
point to the importance of "branching points" in a developing
situation. At such points, once events start down a certain path,
all possible future outcomes are not equally probable. If an analyst
who asserts that there was a missed opportunity does not provide
a plausible scenario of how the outcome would have been more favorable,
then it is not yet a strong candidate for a plausible missed opportunity.
Those
of us interested in assessing possible missed opportunities more
rigorously may find it useful, if not indeed necessary, to keep
such distinctions in mind. At the same time, we believe that the
difficulties of assessing missed opportunities should not discourage
us from efforts to do so. It is not that we are interested in rewriting
history per se. Rather, careful study of possible missed opportunities
is necessary if we are to learn from experience.34
This
paper has argued that policymakers must cultivate an integrated
strategy that develops potential responses with anticipated warnings.
The need to do so will only increase as publics increasingly expect
their governments to do something about crises that they surely
see coming. We believe that it has become implausible for Western
governments to claim that they "didn't know" that something
on a scale of Bosnia or Rwanda could happen. Similarly, claims that
"nothing could be done" ring hollow when coming from such
advanced, wealthy states. These states cannot prevent every conflict,
but they would do well to strengthen their ability to act responsibly
and in a timely manner.
The
authors are grateful for the insightful comments and suggestions
of a number of scholars and practitioners in the field, particularly
Andrew Bennett, Barry Blechman, Herman Cohen, Bruce Jentleson, Daniel
Kaufman, Mary McCarthy, Crystal Nix, Enid Schoettle, John Stremlau,
Ted Raphael, Jim Reed, and Greg Treverton. The authors would also
like to thank Brian George for his assistance in preparing the text
of this paper.
1.
The changing dimensions of information gathering and response are
highlighted in two Foreign Affairs articles by Joseph Nye.
Nye notes that as warning indicators become more diffuse and complex,
"information about what is occurring becomes a central commodity
of international relations, just as the threat and use of military
force was seen as the central power resource in an international
system overshadowed by the potential clash of superpowers."
Nye suggests that international coalitions will, in the future,
be based on "the ability quickly to reduce the ambiguity of
violent situations, to respond flexibly, and to use force, where
necessary, with precision and accuracy." Joseph Nye and William
Owens, "America's Information Edge," Foreign Affairs
75, no. 2 (March/April 1996), pp. 20–36; and Joseph Nye,
"Peering into the Future," Foreign Affairs 73,
no. 4 (July/August 1994), pp. 82–93. See also Gregory F. Treverton,
"Estimating Beyond the Cold War," Defense Intelligence
Journal 3 (1994), pp. 5–20.
2.
See for example, Pauline H. Baker and John A. Ausink, "State
Collapse and Ethnic Violence: Toward a Predictive Model," Parameters
26, no. 1 (1996); "Early Warning of Communal Conflicts
and Humanitarian Crises," Proceedings of a workshop held at
the Center for International Development and Conflict Management,
University of Maryland, November 5–6, 1993, Journal of Ethno-Development
4, no. 1 (July 1994); Ted Robert Gurr, "Early Warning Systems:
From Surveillance to Assessment to Action," in Preventive
Diplomacy, ed. Kevin M. Cahill (New York: Basic Books, 1996);
Robert Kennedy, "Warning for National Response in the 21st
Century," Strategic Outreach Conference Report, Strategic
Studies Institute, US Army War College (August 18–19, 1994);
Mary McCarthy, "The National Warning System: Striving for an
Elusive Goal," Defense Intelligence Journal 3, no. 1
(1994); Jurjen van der Vlugt and Klaas van Walraven, "Conflict
Prevention and Early Warning in the Political Practice of International
Organizations," Netherlands Institute of International Relations
(February 1996); and Robert I. Rotberg, ed., Vigilance and Vengeance:
NGOs Preventing Ethnic Conflict in Divided Societies (Washington,
DC: Brookings Institution Press, 1996).
A
number of countries are struggling to improve their capacity to
foresee, and respond to, humanitarian and political crises in the
post–Cold War era. For an example of national efforts to understand
and develop a response to longer-term changing international circumstances,
see Canada 21: Canada and Common Security in the Twenty-First
Century (The Center for International Studies, University of
Toronto, 1994).
For
an incisive, documented analysis of the warning–response gap
in dealing with humanitarian emergencies, see "Global Humanitarian
Emergencies 1995," released by the United States Mission to
the UN (January 1995). The so-called "Norwegian Model"
offers an example of successful government-NGO cooperation to overcome
this gap. The framework for Norwegian efforts is provided by the
Norwegian Emergency Preparedness System (NOREPS) and Norwegian Resource
Bank for Democracy and Human Rights (NORDEM), which provide flexible
stand-by arrangements and foster close cooperation between government,
voluntary, and academic agencies.
3.
This need was succinctly recognized in the report of The Commission
on Global Governance: "Although the need for collection, analysis,
and dissemination of information cannot be overemphasized, an
even more important task is to initiate action on the basis of information
providing early warning of possible conflicts." Our
Global Neighborhood (Oxford: Oxford University Press, 1995),
p. 98, emphasis added.
4.
This division is clearly demonstrated by the overwhelming—and
often criticized—intelligence community emphasis on information
gathering at the expense of analysis. According to a recent report,
90 percent of the classified intelligence budget of U.S. agencies
is used for the collection of data, while less than 10 percent goes
toward the analysis of this information. See the report of the Twentieth
Century Fund Task Force on the Future of U.S. Intelligence, In
From the Cold (New York: Twentieth Century Fund Press, 1996).
Also, the report of the Commission on the Roles and Capabilities
of the United States Intelligence Community, Preparing for the
21st Century: An Appraisal of U.S. Intelligence (March 1, 1996).
5.
While this seems true of policy professionals, senior policymakers
seem to be more divided on the optimal degree of separation between
intelligence and policy. The 1995 debate over John Deutch's appointment
as director of Central Intelligence with Cabinet rank points up
the ambivalence. While most observers applauded his appointment
as DCI, many intelligence experts have been critical of the decision
to award him Cabinet rank, arguing that its political nature could
impede Deutch's ability to offer objective intelligence, especially
in times of crisis. Political analysts appear to have a greater
understanding of the president's motives. For a general discussion
of the separation between collection and analysis on the one hand
and the policymaker on the other, see Richard K. Betts, "Policy-makers
and Intelligence Analysts: Love, Hate or Indifference?" Intelligence
and National Security 3, no. 1 (January 1988), pp. 184–189.
6.
A survey of recent newspaper articles and editorials bears this
point out. From Chechnya to Yugoslavia to Rwanda, members of the
press and public have argued that these conflicts were preventable
and have lamented the lack of initiative taken by leading governments
and international organizations to head off such disasters.
7.
Indeed, lack of timely or adequate intelligence has been blamed
for everything from Iraq's invasion of Kuwait in 1990 to the failure
to warn airline passengers of information that terrorists have targeted
a particular carrier. See, for example, Jim Wolf, "Embattled
CIA Nominee Acknowledges U.S. Intelligence Failures," Washington
Post, October 4, 1991; John Mintz, "Intelligence Blamed
in Saudi Attack; Military Officials Dispute Hill Report on Fatal
Dhahran Bombing; Hill Report Faults U.S. Deference to Saudi Hosts,"
Washington Post, August 15, 1996; David B. Ottaway, "U.S.
Considers Slugging It Out With International Terrorism; Aides Split
on Whether to Target Groups or States That Sponsor Them," Washington
Post, October 17, 1996.
8.
Gordon W. Prange with Katherine V. Dillon and Donald Goldstein,
Pearl Harbor: The Verdict of History (New York: McGraw-Hill,
1986), see especially pp. 549–551.
9.
Although that is not to say that policymakers do not want to know
that such events are imminent, only that when compared to direct
national threats, these contingencies are simply deemed less important.
10.
As will be discussed later in the paper, there were a number of
short-term and long-term signals that demonstrated the threat of
military action by Iraq. Though many in mid-1990 may have been surprised
by the timing and scope of Iraq's action, the mobilization of its
army on the Kuwaiti border was no secret. For a detailed discussion
of these warning signals, see Bruce Jentleson, With Friends Like
These: Reagan, Bush, and Saddam, 1982–1990 (New York: W.
W. Norton & Company, 1994), especially pp. 167–176.
11.
This section draws from Chapter 20 of A.L. George and R. Smoke,
Deterrence in American Foreign Policy: Theory and Practice
(New York: Columbia University Press, 1974); and from A.L. George,
"Warning and Response: Theory and Practice," in International
Violence: Terrorism, Surprise and Control, ed., Yair Evron (Hebrew
University of Jerusalem, 1979). For a useful summary and interpretation
of the results of laboratory studies, see Joseph De Rivera, The
Psychological Dimension of Foreign Policy (Columbus, Ohio: C.E.
Merrill Publishing Co., 1968), pp. 53ff.
12.
A number of off-the-record discussions with analysts and policymakers
familiar with the relationship between the intelligence and policy
communities have noted the sometimes notorious competition among
the various intelligence agencies for primacy in informing the policy
process. Assessments of potential crisis situations find agencies
at times in sharp disagreement with one another regarding the likely
outcome. Such sustained disagreement often leads, over time, to
the marginalization of the agency that is at odds with the intelligence
agency offering the estimates that reinforce the policy inclinations
of the key decision makers.
13.
See, for example, Glenn D. Paige, The Korean Decision (New
York: Free Press, 1968), especially pp. 349–352. Ample warning
is acknowledged by Truman in Years of Trial and Hope, 1946–1952,
vol. 2, Memoirs (Garden City, NY: Doubleday, 1956), p. 331.
14.
Burton I. Kaufman, The Korean War (New York: Alfred A. Knopf,
1986), p. 33.
15.
Allen Whiting concluded that both Washington and Beijing miscalculated
the danger of each other's position: the U.S. underestimated Chinese
warnings against moving across the 38th parallel; China, for its
part, overestimated the threat to its security of UN troops crossing
the 38th parallel; in China Crosses the Yalu (Stanford, CA:
Stanford University Press, 1960), see especially pp. 163–172.
Whiting updated the study, making use of newly declassified materials,
in his "The U.S.–China War in Korea," in Avoiding
War: Problems of Crisis Management, ed., A. L. George (Boulder,
CO: Westview Press, 1991), pp. 103–125.
16.
Avi Shlaim, The United States and the Berlin Blockade, 1948–1949
(Berkeley, CA: University of California Press, 1983), pp. 415–416.
17.
Baghdad Radio, July 17, 1990, cited in Lawrence Freedman and Efraim
Karsh, The Gulf Conflict 1990–1991 (Princeton, NJ: Princeton
University Press, 1993), p. 48.
18.
"Kuwait: How the West Blundered," in The Gulf War Reader,
eds., Christopher Cerf and Micah L. Sifry (New York: Times Books,
1991), pp. 99–106. Originally published in The Economist
(September 29, 1990).
19.
Jentleson, With Friends Like These, pp. 173–174 (see
note 10).
20.
A number of sources have included discussions of, and excerpts from,
the meeting between Ambassador Glaspie and Saddam Hussein, including
Jentleson, With Friends Like These, pp. 169–171; and
Freedman and Karsh, The Gulf Conflict 1990–1991, pp.
51–55.
21.
See, for example, Freedman and Karsh, The Gulf Conflict 1990–1991,
p. 63; Jentleson, With Friends Like These; "Kuwait:
How the West Blundered," in The Gulf War Reader.
22.
Africa Watch, "Rwanda: Talking Peace and Waging War: Human
Rights Since the October 1990 Invasion" (February 27, 1992).
23.
John Erikkson, The International Response to Conflict and Genocide:
Lessons from the Rwanda Experience, Synthesis Report (Copenhagen:
Steering Committee of the Joint Evaluation of Emergency Assistance
to Rwanda, March 1996), p. 21.
24.
Holly J. Burkhalter, "The Question of Genocide: The Clinton
Administration and Rwanda," World Policy Journal 11,
no. 4 (Winter 1994/95), pp. 44–54, originally quoted in Keith
Richburg, "Witnesses Describe Cold Campaign of Killing in Rwanda,"
Washington Post, 8 May 1994. For additional studies of the
Rwandan genocide, see Howard Adelman and Astri Surhke, The International
Response to Conflict and Genocide: Lessons from the Rwanda Experience,
Study 2: Early Warning and Conflict Management (Copenhagen: Steering
Committee of the Joint Evaluation of Emergency Assistance to Rwanda,
1996); Alain Destexhe, Rwanda and Genocide in the Twentieth Century
(New York: New York University Press, 1995); Gerard Prunier, The
Rwanda Crisis: History of a Genocide (New York: Columbia University
Press, 1995). For an incisive critique of bureaucratic phenomena
that accompanied the UN's sluggishness, written by the person specializing
on Rwanda in the U.S. Mission to the United Nations, see Michael
N. Barnett, "The Politics of Indifference at the United Nations:
The Security Council, Peacekeeping and Genocide in Rwanda,"
forthcoming in Cultural Anthropology.
25.
Likewise, it is not difficult to find excuses not to act. In Rwanda,
for example, even in the aftermath of genocide, the limited U.S.
contribution of armored personnel carriers to UNAMIR was delayed
for two months while the State Department haggled with the UN over
compensation for the vehicles.
26.
Such reluctance may be a Western phenomenon. In an essay prepared
for the Carnegie Commission on Preventing Deadly Conflict, former
Soviet president Mikhail Gorbachev argues that the response tendency
in many former bloc states, as well as other repressive regimes,
tends to be exactly the opposite: knee-jerk crackdowns. The essay,
"Nonviolent Leadership," is part of a volume that the
Commission is preparing on the role of leadership in preventing
deadly conflict.
27.
This difficulty is highlighted, for example, in Ted Robert Gurr
and Barbara Harff, "Conceptual, Research and Policy Issues
in Early Warning Research: An Overview," The Journal of
Ethno-Development 4, no. 1; also, Janie Leatherman and Raimo
Vyrynen, "Structure, Culture and Territory: Three Sets of Warning
Indicators," paper prepared for the 36th Annual Convention
of the International Studies Association, Chicago, IL, February
21–25, 1995.
28.
The "cry wolf" dilemma in warning and response is cogently
summarized by William Zartman: "The biggest problem in the
early warning debate is not whether an event is preceded by warning
signals but whether warning signals are followed by an event. There
are many more prior indications than there are ensuing events; many
warning signals simply fizzle and seemingly impending events work
themselves out . . . What is needed is tornado warnings that announce
tornados but also that do not announce non-tornados. The corridors
of policy-makers reverberate with cries of `wolf!'" I. William
Zartman, "Preventing State Collapse: The Argument," draft
paper, Working Group on Collapsed States, The Johns Hopkins University,
Washington, DC, November 1996.
Other
observers have noted that while greater vigilance in warning and
response does result in an increase in "cry wolf" outcomes,
errors of this type are preferable to extremely costly failures
resulting from lack of attention to developing events.
29.
Our Global Neighborhood, pp. 94–95 (see note 3).
30.
A number of experienced intelligence and policy specialists have
endorsed the need for developing a response "repertoire"
that includes a wide array of responses, some small, possibly covert,
and low cost, others large, public, and more costly. The rationale
that "the response must fit the warning" is a simple one,
but not one easily realized. The response repertoire, of course,
should include the many different responses that can be made by
nongovernmental organizations. See, for example, the remarks by
John Brinkerhoff in Strategic Outreach Conference Report: Warning
for National Responses in the 21st Century, SSI/Strategic Studies
Institute, U.S. Army War College (August 18–19, 1994). See
also the study by Michael Lund, Preventing Violent Conflicts:
A Strategy for Preventive Diplomacy (Washington, D.C.: United
States Institute of Peace Press, 1996).
31.
A recent World Bank report provides a very interesting example of
the possible utility of differentiating early warning signals with
respect to identifying (a) different possible adverse consequences,
and (b) appropriate preventive actions for each such consequence.
Nat J. Colletta, Markus Kostner, and Ingo Wiederhofer, The Transition
from War to Peace in Sub-Saharan Africa (Washington, DC: The
World Bank, 1996), see especially pp. 32–38.
32.
These cues do not necessarily need to occur to be foreseeable; indeed
early preventive measures may mean that they never occur.
33.
See, for example, Philip E. Tetlock and Aaron Belkin, Counterfactual
Thought Experiments in World Politics (Princeton, NJ: Princeton
University Press, 1996).
34.
This is the objective of a study for the Carnegie Commission on
Preventing Deadly Conflict, Opportunities Missed, Opportunities
Seized: Preventive Diplomacy in the Post–Cold War World,
directed and edited by Bruce W. Jentleson, to be published in late
1997 by Rowman & Littlefield (Boulder, CO). Some problems of
obtaining and using warning by NGOs in humanitarian crises are discussed
in Robert I. Rotberg, ed., Vigilance and Vengeance (see note
2).
David
A. Hamburg, Cochair
President Emeritus
Carnegie Corporation of New York
Cyrus
R. Vance, Cochair
Partner
Simpson Thacher & Bartlett
Gro
Harlem Brundtland
Former Prime Minister of Norway
Virendra
Dayal
Former Under-Secretary-General and
Chef de Cabinet to the Secretary-General
United Nations
Gareth
Evans
Deputy Leader of the Opposition and Shadow Treasurer
Australia
Alexander
L. George
Graham H. Stuart Professor Emeritus of International Relations
Stanford University
Flora
MacDonald
Former Foreign Minister of Canada
Donald
F. McHenry
University Research Professor of Diplomacy and International Affairs
Georgetown University
Olara
A. Otunnu
President
International Peace Academy
David
Owen
Chairman
Humanitas
Shridath
Ramphal
Cochairman
Commission on Global Governance
Roald
Z. Sagdeev
Distinguished Professor
Department
of Physics
University of Maryland
John
D. Steinbruner
Senior Fellow
Foreign Policy Studies Program
The Brookings Institution
Brian
Urquhart
Former Under-Secretary-General for Special Political Affairs
United Nations
John
C. Whitehead
Chairman
AEA Investors Inc.
Sahabzada
Yaqub-Khan
Former Foreign Minister of Pakistan
Chairman, Board of Trustees
Aga Khan International University -- Karachi
Pakistan
Special
Advisor to the Commission
Herbert S. Okun
Visiting Lecturer on International Law
Yale Law School
Former U.S. Representative to the German Democratic Republic and
to the UN
Jane
E. Holl, Executive Director
Morton
Abramowitz
Acting President
International Crisis Group
Ali
Abdullah Alatas
Minister for Foreign Affairs
Republic of Indonesia
Graham
T. Allison
Douglas Dillon Professor of Government
John F. Kennedy School of Government
Harvard University
Robert
Badinter
President Emeritus
Constitutional Council of France
Carol
Bellamy
Executive Director
United Nations Children's Fund
Harold
Brown
Counselor
Center for Strategic and International Studies
McGeorge
Bundy*
Scholar-in-Residence
Carnegie Corporation of New York
Jimmy
Carter
The Carter Center of Emory University
Lori
Damrosch
Professor of Law
Columbia University School of Law
Francis
M. Deng
Senior Fellow
Foreign Policy Studies Program
The Brookings Institution
Sidney
D. Drell
Professor and Deputy Director
Stanford Linear Accelerator Center
Stanford University
Lawrence
S. Eagleburger
Senior Foreign Policy Advisor
Baker Donelson Bearman & Caldwell
Leslie
H. Gelb
President
Council on Foreign Relations
David
Gompert
Vice President
National Security Research
RAND
Andrew
J. Goodpaster
Chairman
The Atlantic Council of the United States
Mikhail
S. Gorbachev
The Gorbachev Foundation
James
P. Grant**
Executive Director
United Nations Children's Fund
Lee
H. Hamilton
United States House of Representatives
Theodore
M. Hesburgh
President Emeritus
University of Notre Dame
Donald
L. Horowitz
James B. Duke Professor of Law and Political Science
Duke University School of Law
Michael
Howard
President
International Institute for Strategic Studies
Karl
Kaiser
Director
Research Institute of the German Society for Foreign Affairs
Nancy
Landon Kassebaum Baker
Baker Donelson Bearman & Caldwell
Sol
M. Linowitz
Honorary Chairman
The Academy for Educational Development
Richard
G. Lugar
United States Senate
Michael
Mandelbaum
Christian A. Herter Professor of American Foreign Policy
The Paul H. Nitze School of Advanced International Studies
The Johns Hopkins University
Robert
S. McNamara
Former U.S. Secretary of Defense
William
H. McNeill
Professor Emeritus of History
University of Chicago
Sam
Nunn
Partner
King & Spalding
Olusegun
Obasanjo
President
Africa Leadership Forum
Sadako
Ogata
The High Commissioner for Refugees
United Nations
Javier
P§rez de Cu§llar
Former Secretary-General
United Nations
Condoleezza
Rice
Provost
Stanford University
Elliot
L. Richardson
Milbank, Tweed, Hadley & McCloy
Harold
H. Saunders
Director of International Affairs
Kettering Foundation
George
P. Shultz
Distinguished Fellow
Hoover Institution on War, Revolution, and Peace
Stanford University
Richard
Solomon
President
United States Institute of Peace
James
Gustave Speth
Administrator
United Nations Development Programme
Desmond
Tutu
Archbishop Emeritus of Cape Town
James
D. Watkins
President
Consortium for Oceanographic Research and Education
Elie
Wiesel
President
The Elie Wiesel Foundation for Humanity
I.
William Zartman
Jacob Blaustein Professor of International Organizations and Conflict
Resolution
Director of the African Studies and Conflict Management Programs
The Paul H. Nitze School of Advanced International Studies
The Johns Hopkins University
*
Deceased September 1996.
**
Deceased February 1995.
ALEXANDER
L. GEORGE is Graham H. Stuart Professor Emeritus of International
Relations at Stanford University. A leading academic specialist
on deterrence, crisis prevention and management, and coercive diplomacy,
Dr. George came to Stanford in 1968 after 20 years at the Rand Corporation,
where he had been head of the social science department. The most
recent of his many scholarly publications are Limits of Coercive
Diplomacy (1994) and Bridging the Gap: Theory and Practice
of Foreign Policy (1993). Deterrence in American Foreign
Policy, which he coauthored with Richard Smoke, won the Bancroft
Prize in 1975. In 1983 he was the recipient of a MacArthur Foundation
Five-Year Prize Award.
JANE
E. HOLL is executive director of the Carnegie Commission on
Preventing Deadly Conflict. Prior to joining Carnegie, Holl was
a career officer in the United States Army, serving most recently
as director for European Affairs on the National Security Council
staff. She holds a Ph.D. in political science from Stanford University.
|
