27 April 2006


Mary  O. McCarthy, "The Mission to Warn: Disaster Looms," Defense Intelligence Journal; 7-2 (1998), 17-31. http://cryptome.org/mccarthy-mtw.htm

Mary McCarthy, "The National Warning System: Striving for an Elusive Goal," Defense Intelligence Journal 3-1 (1994), 5-19. http://cryptome.org/mccarthy-nws.htm

Source: Hardcopy Defense Intelligence Journal.

Defense Intelligence Journal; 7-2 (1998), 33-44


Charles E. Allen

Mr. Allen is the Assistant Director of Central Intelligence for Collection, a position to which he was appointed in May of 1998. He has served with the Central Intelligence Agency (CIA) since 1958, holding a variety of positions of increasing responsibility both in analytic and managerial capacities, including service as the as Chief of the CIA Counterterrorist Center. Mr. Allen served as the National Intelligence Officer for Terrorism and Narcotics. From 1988 to 1994 he was the NIO for Warning.

The views expressed in this article are those of the author and do not reflect the official policy or position of the Central· Intelligence Agency, the Intelligence Community or the US Government.

Strategic warning emerged as a special topic of study following three catastrophes; Pearl Harbor, the imposition of the Berlin Blockade in 1949, and the Korean War in 1950. As a result, in 1953, the Director of Central Intelligence (DCI), Walter Bedell Smith, established a new full-time warning organization -- a 24-hour National Indications Center that reported to the Watch Committee of the United States Intelligence Board. DCI Smith defined the mission of that organization as being to avoid strategic surprise. His first directive on warning was the genesis of a national warning strategy to accomplish that mission.

Despite the dedication of resources and effort to avoiding surprise, we all know that the United States has suffered a number of strategic warning disasters in the post - World War II era. This article will review some of those instances of strategic surprise and describe how the warning function of the Intelligence Community evolved as a result of them. It will then report how the warning staff of the Community performed in the period leading up to the Iraqi Invasion of Kuwait.

Warning in the 1960s and 1970s

In the early years, the primary topic of warning was war, either by actions of the USSR, China, or North Korea. The warning organizations established at that time were proscribed from studying other areas closely, except as regional threats that contained the potential to escalate to global conflict.

The chief technique used by the new warning organizations was to compile and apply check lists -- indicator lists -- of potentially threatening military activities. Military mobilization was considered the single most important indicator on which to base warning.

While following this warning paradigm, the US suffered several strategic surprises, including the Soviet invasion of Czechoslovakia. One of the most devastating was the Intelligence Community's failure to warn of the Arab- Israeli War of October 1973.

The failure in 1973 was the catalyst for a far-reaching reconsideration of the Community's national warning strategy. The impetus for change, however, ironically, came not from the Intelligence Community but from a House Permanent Select Committee on Intelligence (HPSCI) review of warning conducted in 1978. One of the principal recommendations of the HPSCI study was that the Intelligence Community should establish a single point of focus for warning. This recommendation led to the establishment of the position of the National Intelligence Officer (NIO) for Warning and the establishment of an interagency National Warning Staff.

Warning Since the 1980s

As a result, during the 1980s, a new warning strategy was evolved, which kept the best of the old techniques, but steered in new directions. The mission was redefined as using intelligence warning to help keep the country safe -- not just to avoid surprise. This meant alerting not only intelligence but also policy officials as early as possible of potential damage to US interests and threats to US territory and citizens.

Under this rubric, the NIO for Warning and his staff were charged with monitoring a wide range of threats in all parts of the world, including potentially damaging political, social, and economic changes; weapons developments; and, of course, the threat of warfare in any region. In issuing warnings for most of these threats, the NIO for Warning and the National Warning Staff relied on and assisted other elements of the Intelligence Community. Functionally, this broader approach to monitoring threats to US interests led to the development of threat management as a complement to crisis management. The idea was not just to avoid surprise, it was to help avoid damage before a crisis erupted.

The goals of threat management were to warn early, even if information were ambiguous; to base the warning on reliable evidence to ensure its persuasiveness; and to discriminate diagnostically among bluffs, exercises, demonstrations, and genuine national war preparations. These goals have proven difficult to implement across-the-board in the Intelligence Community because, by its nature, the Community prefers to be right before issuing warning.

We now understand that warning is a repetitive process that is not completed until national policymakers accept the threat as genuine or reject the evidence as not persuasive. In the latter case, warning must be repeated as and when more persuasive evidence is obtained.

As for techniques, warning officers use an analytical approach that concentrates on the underlying structure of various threats--the phenomenology of threats. This approach includes a variety of sampling and structured analytical techniques, including indicator lists, and applies them to the total process by which countries prepare for war, or political systems become unstable. The warning process is designed to be safe, rather than waiting to be right, thereby giving decisionmakers time to clarify, avoid, deter, or manage threats, using prudent measures. These measures can be diplomatic, political and economic, if persuasive warning is given early enough.

Lessons Learned

The development of a new warning strategy in the 1980's facilitated accomplishment of the mission in ways we did not foresee. First, we found that, by looking at the total war process, we could effectively differentiate warning of national war preparations, which included disruptions to civilian social and economic activities, from warning of attack or from the early development of capabilities by national military forces to launch a future attack. In practice, we also found we could further refine warning of attack by distinguishing between general attack preparations, which usually feature the buildup of forces, and their supporting logistics and final attack preparations, which are distinguished by forward movement of forces to attack positions.

Using sampling and other techniques from the field of operations research that help elucidate the underlying structure of a threat, we also discovered we could warn earlier and more reliably than we thought possible. We learned, for example, that countries with centrally planned economic systems -- such as the Soviet Union, North Korea, Syria, Iraq, and even India -- prepare for war by first initiating measures across the spectrum of social, economic, political, and military life that divert resources away from civilian life and into the armed forces. We further learned that, without large logistic buildups, combat force movements will always remain important but ambiguous indicators of intentions.

One implication of these findings is that we can apply more intelligence information from all sources, including press and low-grade human sources, more effectively in support of warning judgments. Another important finding is that policymakers tend to be more receptive to warning when it provides enough time for developing and employing prudent options to manage a developing threat.

Warning in Practice

By using this approach to warning, US intelligence agencies, the NIO for Warning, and the National Warning Staff have helped to avoid crises and to control escalation in a number of situations. The India-Pakistan military buildup in late 1986 and early 1987 was headed toward a potential conflict brought about, principally, by a misperception of each other's intentions and military capabilities improvements. US intervention in response to early warning, especially that provided by John Bird, the NIO for Warning at that time, succeeded in defusing the crisis.

In early 1990, when I was the NIO for Warning, we again warned of potential conflict between India and Pakistan as the Kashmir insurgency began to develop. Between January and May, as the violence in Kashmir worsened and New Delhi and Islamabad began to deploy military forces, the National Warning Staff, led by the very able John McCreary of the Defense Intelligence Agency (DIA), issued five warnings about the potential for a major conflict.

Other agencies of the Intelligence Community, which initially disagreed with us strongly on the probability of conflict, also eventually responded and, in May, warned of the high probability of war. Thus, in response to warnings, US efforts to defuse the crisis succeeded. There is evidence that some government leaders, both in India and Pakistan, had expected war to break out in June.

In this period, we also expanded our repertoire of issues about which we warn and found that, by using new analytical techniques, we could warn effectively of instability and low-intensity conflicts. By watching the process of breakdown of authority, we warned in September 1989 of the development of serious instability in Azerbaijan and Armenia in the USSR and of the development of a crisis in Lithuania. Both warnings were given about three months before these developments reached crisis proportions. But, nothing we did came close to the dramatic warnings involving Iraq and Kuwait during those fateful days of July and August 1990.

Warning of the Threat from Iraq

My Staff and I first became concerned about the threat from Iraq in the final months of the Iran-Iraq War in 1988. Our concern was stimulated by Iraq's successful use of chemical weapons in an offensive mode against Iranian forces. After Iraq used chemical weapons against civilians at Hallabjah in March 1988, I issued warnings that chemical weapons would be used in subsequent offensive operations and judged that they could prove decisive in ending the war.

In 1988 and 1990, my Staff and I remained concerned about Iraqi intentions, primarily in the context of an Arab-Israeli conflict. This was largely because Saddam Husayn did not demobilize his large standing force and persisted in efforts to develop chemical and biological weapons and long-range delivery systems. In mid-1989, my Staff assessed that Iraqi military capabilities could be decisive in a future Arab war against Israel, a conclusion that was not accepted at the time by Intelligence Community experts with whom we consulted.

In January 1990, the US Central Command and the Joint Staff requested the National Warning Staff's assistance in reviewing indicators of a potential Iraqi attack against Kuwait and Saudi Arabia. Using the techniques we had developed, my Staff, under John McCreary's leadership, rated the threat as plausible, though not imminent at that time. They assessed that Iraq would take about 12 days to prepare, assemble, and move a corps-size attack force into position for such an attack. They also assessed that several months of preparation would precede such an attack and would be accompanied by a political crisis. I might add at this point that, later, there were reports that Iraq had prepared the attack force for its eventual move into Kuwait over about a three-month period. Additionally, the initial movement of Iraqi forces was reported 12 days before the actual attack against Kuwait.

In June and early July 1990, the National Warning Staff participated in a number of meetings with the Director for Strategic Plans and Policy (J-5) of the Joint Staff concerning contingency planning for US forces to respond to an Iraqi attack against Kuwait and Saudi Arabia. National Warning Staff officers provided indicators and scenario events for the J-5 plans, a further refinement of the intelligence provided to the Central Command in January.

The Kuwait Crisis: 17 July to 2 August 1990

Saddam Husayn's 17 July denunciation of Kuwait and the United Arab Emirates over longstanding grievances again focused my Office on the possible threat from Iraq. Our concern intensified on 19 July, when information was received that Saddam Husayn had ordered two elite Republican Guard divisions to the Kuwait border.

Because of my deep concern that Saddam was prepared to reassert Iraq's longstanding claims to Kuwait, I directed my Staff to begin consultations with Intelligence Community analysts to confirm the deployment. I personally contacted collection agencies to marshal collection assets to confirm the movement of the Iraqi forces. On 21 July, we confirmed, by national technical means, the first movements of Republican Guards elements and, on 23 July, we detected the start of a large logistic buildup just north of the Kuwaiti border. Reporting of large truck movements stretching southward from Baghdad to Basrah indicated that civilian resources were being diverted to sUpport large-scale military movement, a powerful indicator that the military preparations were not a bluff or an exercise.

On 23 July, after consultations with CIA and DIA analysts, I directed my Staff to begin writing a special warning report. The intelligence indicated that Iraq had initiated its war preparation process just as we expected it, virtually mirroring the scenario events and indicators that my Staff had provided to Central Command and the J-5.

To support the warning report, the National Warning Staff conducted a review of the political indicators to determine Iraq's motives. That review concluded that Saddam's objectives included, inter alia, the removal of the government of Kuwait and would not likely be satisfied by political negotiations and compromise.

On 25 July, I issued a "warning of war" memorandum which stressed that Iraq had nearly achieved the capability to mount a corps-size operation capable of defeating Kuwaiti forces and of occupying much of Kuwait. This report rated the chances of a "military incursion" at better than 60 percent. It further stressed that, even with major Kuwaiti concessions, the chances of some form of Iraqi military action remained significant -- 25 percent.

The Chairman of the National Intelligence Council (NIC), after reviewing the report, stated that the warning appeared to be "OBE," -- that is, "overtaken-by-events". He cited a cable which had just been received from the US Ambassador in Iraq who recounted Saddam's assurance that he had no intention of taking military action because of forthcoming talks between Iraq and Kuwait in Jeddah and Baghdad. In response I stated that Saddam could be engaging in deception, given the massive buildup of Iraqi forces along the border and the costs the Iraqi government was incurring to complete it. The Chairman of the NIC agreed and concurred in the issuance of the warning report.

The report was hand-carried to our standard list of customers, which included all of the senior policy officials of the Department of Defense, the Joint Chiefs of Staff and White House, including: Carnes Lord, the National Security Adviser to the Vice President; William Working, the Presidential Assistant for Intelligence Liaison on the National Security Council; and Richard Haass, the President's Special Assistant for Near East and South Asian Affairs on the National Security Council. Additionally, copies were sent to the Assistant Chiefs of Staff for Intelligence of the Services well as to the Directors of the Defense Intelligence and National Security Agencies.

On 28 July, the National Warning Staff began compiling a standard warning chronology and data base, tracing the development of events during the past year. Copies of this document were provided to intelligence production organizations and warning officers in the Intelligence Community. I also continued to assist, daily, in collection tasking in order to make the most effective use of collection assets to track the military buildup and locate Iraqi units in support of warning.

My sense of concern about the Intelligence Community's analytical direction was heightened by a 31 July teleconference of Intelligence Community analysts. They judged that no major military action was likely. Moreover, they assessed that, if military action were to occur, likely, it would be to seize only northern oilfields and offshore islands in Kuwait and that it was unlikely that Iraq would attempt to occupy all of Kuwait. From a threat process perspective, the indicators of a large-scale military movement were obvious and were continuing to build. From my perspective, the Community focused too strongly on the distinction between capabilities and intentions. In our approach, capabilities reflect intentions because they show where the real costs are being paid. Critical to our judgments was the large logistic buildup which indicated that disruption of civil life was occurring inside Iraq to accomplish the large military buildup north of Kuwait.

On 1 August, by 0645 EDT, national technical means indicated that the Iraqis had accomplished those actions which we assessed as final attack indicators. These included movement of armored and mechanized infantry brigades and artillery battalions, arrayed in attack formations, to within two kilometers of the Kuwait border; the forward movement of ground attack aircraft to airfields near Kuwait; and the massing of approximately 50 attack helicopters.

At that early hour, I called Mr. William Working and Mr. Richard Haass of the National Security Council; Commander Andrew Riddile of the Vice President's Office;. Assistant Secretary of State, Richard Clarke; the Office of Lieutenant General Lee Butler, J-5; and officers in my own agency, telling them that I was giving them "warning of attack," and that there could be no further warning. I directed my Staff to advise senior defense officials that this was the final warning. In some instances, I was unable to talk to some senior officers directly and could only speak to lower ranking staff I called Commander Riddile twice, requesting him to inform Mr. Lord and the Vice President. In some cases, we were not successful. My Staff's efforts to warn the Director of Current Operations of the Joint Staff (J-33) received a cold response. I then directed my Staff to begin writing a final "warning of attack" report and arranged to brief personally William Working at the White House.

At 0930 EDT, I went to the White House to brief Mr. Working on the latest developments with Thomas Lewis, an officer from the National Photographic Interpretation Center (NPIC) and John McCreary. Mr. Working was sufficiently impressed that he arranged an immediate briefing for Dr. Haass. The briefing was presented to Dr. Haass at 10 15 EDT. He did not seem as convinced as I that a massive Iraqi attack against Kuwait was only hours away.

At 1900 EDT, the National Warning Staff distributed the "warning of attack" report to a number of Defense and White House consumers. At that hour, we were only able to reach an Assistant Secretary of Defense, Dr. Henry Rowan; the Director of Current Operations (J-33) of the Joint Staff, Rear Admiral Fitzgerald; the Army's Assistant Chief of Staff for Intelligence, Lieutenant General Eichelberger; and the White House Situation Room. General Eichelberger, to his credit, read" the report carefully and indicated that he agreed that an attack was imminent. The assessment was on the mark, except we underestimated, by a good margin, how quickly Iraqi forces could attack Kuwait. We had suggested that the attack might take a few days, depending upon Kuwaiti resistance. Some officials who read the report commented that they expected either a continued demonstration of force or, if military action were taken, it would be a limited incursion and not an occupation of Kuwait.


In retrospect, my Staff and I viewed Iraq's invasion of Kuwait as a classic warning episode. Indicators were assessed early and accurately. Our prior assessment of time lines for war preparations proved correct. Collection was honed and focused; coordination with the analytic community was constant; and policy officials were informed of our conclusions at each major stage in the development of the threat, personally as well as in writing. Nevertheless, the warning messages of the NIO for Warning -- both warning of war and warning of attack -- were not heeded, either by senior intelligence officials or policymakers

As a postscript, at a session of senior military and civilian officers at the Pentagon, General Butler, the J-5, stated that the NIO for Warning had provided warning of war and warning of attack, but that he was not taken seriously because senior US officials talked with, and accepted the judgment of a number of leaders in the Middle East as well as the Soviet Union, all of whom were of the opinion that Saddam did not intend to attack.

After our disastrous history in warning -- Czechoslovakia in 1968; the October war in the Middle East in 1973; Afghanistan in 1979; and others -- the Intelligence and Policy Communities still seemed, in the aftermath of the Iraqi invasion, to lack a fundamental understanding of the causes of strategic surprise. The factors that inhibited effective early warning in this crisis corresponded closely to those that led to warning failure during the October 1973 War in the Middle East.

In both crises, separated by 17 years of improvements in intelligence collection and training, most intelligence analysts and policy officials underestimated the political intentions of the aggressor; minimized the aggressor's possible military objectives and the probability of war; downplayed the significance of the military indicators of threat; and gave excessive weight to the, ultimately, mistaken opinions of the foreign leaders and foreign intelligence services most directly threatened by aggression. In 1973, no warnings were issued by any intelligence agency. In this crisis, the NIO for Warning and the National Warning Staff provided repeated warnings, but these could not compete with the more reassuring messages provided by others.

HTML by Cryptome.