United States/United Kingdom, Report on the Conduct of the Persian Gulf War

A. Department of Defense Report to Congress on the Conduct of the Persian Gulf War

[Source: ILM, vol. 31 (3), 1992, pp. 612-644. Available on http://www.globalsecurity.org/military/library/report/1992]



The law of war with respect to targeting, collateral damage and collateral civilian casualties is derived from the principle of discrimination; that is, the necessity for distinguishing between combatants, who may be attacked, and noncombatants, against whom an intentional attack may not be directed, and between legitimate military targets and civilian objects. Although this is a major part of the foundation on which the law of war is built, it is one of the least codified portions of that law.

As a general principle, the law of war prohibits the intentional destruction of civilian objects not imperatively required by military necessity and the direct, intentional attack of civilians not taking part in hostilities. The United States takes these proscriptions into account in developing and acquiring weapons systems, and in using them in combat. Central Command (CENTCOM) forces adhered to these fundamental law of war proscriptions in conducting military operations during Operation Desert Storm through discriminating target selection and careful matching of available forces and weapons systems to selected targets and Iraqi defenses, without regard to Iraqi violations of its law of war obligations toward the civilian population and civilian objects.

Several treaty provisions specifically address the responsibility to minimize collateral damage to civilian objects and injury to civilians. Article 23(g) of the Annex to Hague IV prohibits destruction not “imperatively demanded by the necessities of war,” while Article 27 of that same annex offers protection from intentional attack to “buildings dedicated to religion, art, science, or charitable purposes, historic monuments, hospitals, and places where the sick and wounded are collected, provided they are not being used at the time for military purposes.” Similar language is contained in Article 5 of Hague IX, while [...] in the 1954 Hague Cultural Property Convention [...] cultural and civilian objects are protected from direct, intentional attack unless they are used for military purposes, such as shielding military objects from attack.

While the prohibition contained in Article 23(g) generally refers to intentional destruction or injury, it also precludes collateral damage of civilian objects or injury to noncombatant civilians that is clearly disproportionate to the military advantage gained in the attack of military objectives, as discussed below. As previously indicated, Hague IV was found to be part of customary international law in the course of war crimes trials following World War II, and continues to be so regarded.

An uncodified but similar provision is the principle of proportionality. It prohibits military action in which the negative effects (such as collateral civilian casualties) clearly outweigh the military gain. This balancing may be done on a target-by-target basis, as frequently was the case during Operation Desert Storm, but also may be weighed in overall terms against campaign objectives. CENTCOM conducted its campaign with a focus on minimizing collateral civilian casualties and damage to civilian objects. Some targets were specifically avoided because the value of destruction of each target was outweighed by the potential risk to nearby civilians or, as in the case of certain archaeological and religious sites, to civilian objects.

Coalition forces took several steps to minimize the risk of injury to noncombatants. To the degree possible and consistent with allowable risk to aircraft and aircrews, aircraft and munitions were selected so that attacks on targets within populated areas would provide the greatest possible accuracy and the least risk to civilian objects and the civilian population. Where required, attacking aircraft were accompanied by support mission aircraft to minimize attacking aircraft aircrew distraction from their assigned mission. Aircrews attacking targets in populated areas were directed not to expend their munitions if they lacked positive identification of their targets. When this occurred, aircrews dropped their bombs on alternate targets or returned to base with their weapons.

One reason for the maneuver plan adopted for the ground campaign was that it avoided populated areas, where Coalition and Iraqi civilian casualties and damage to civilian objects necessarily would have been high. This was a factor in deciding against an amphibious assault into Kuwait City.

The principle of proportionality acknowledges the unfortunate inevitability of collateral civilian casualties and collateral damage to civilian objects when noncombatants and civilian objects are mingled with combatants and targets, even with reasonable efforts by the parties to a conflict to minimize collateral injury and damage.

This proved to be the case in the air campaign. Despite conducting the most discriminate air campaign in history, including extraordinary measures by Coalition aircrews to minimize collateral civilian casualties, the Coalition could not avoid causing some collateral damage and injury.

There are several reasons for this. One is the fact that in any modern society, many objects intended for civilian use also may be used for military purposes. A bridge or highway vital to daily commuter and business traffic can be equally crucial to military traffic, or support for a nation’s war effort. Railroads, airports, seaports, and the interstate highway system in the United States have been funded by the Congress in part because of US national security concerns, for example; each proved invaluable to the movement of US military units to various ports for deployment to Southwest Asia (SWA) for Operations Desert Shield and Desert Storm. Destruction of a bridge, airport, or port facility, or interdiction of a highway can be equally important in impeding an enemy’s war effort.

The same is true with regard to major utilities; for example, microwave towers for everyday, peacetime civilian communications can constitute a vital part of a military command and control (C2) system, while electric power grids can be used simultaneously for military and civilian purposes. Some Iraqi military installations had separate electrical generators; others did not. Industries essential to the manufacturing of CW, BW and conventional weapons depended on the national electric power grid.

Experience in its 1980-1988 war with Iran caused the Government of Iraq to develop a substantial and comprehensive degree of redundancy in its normal, civilian utilities as back-up for its national defense. Much of this redundancy, by necessity, was in urban areas. Attack of these targets necessarily placed the civilian population at risk, unless civilians were evacuated from the surrounding area. Iraqi authorities elected not to move civilians away from objects they knew were legitimate military targets, thereby placing those civilians at risk of injury incidental to Coalition attacks against these targets, notwithstanding efforts by the Coalition to minimize risk to innocent civilians.

When objects are used concurrently for civilian and military purposes, they are liable to attack if there is a military advantage to be gained in their attack. (“Military advantage” is not restricted to tactical gains, but is linked to the full context of a war strategy, in this instance, the execution of the Coalition war plan for liberation of Kuwait.)

Attack of all segments of the Iraqi communications system was essential to destruction of Iraqi military C2. C2 was crucial to Iraq’s integrated air defense system; it was of equal importance for Iraqi ground forces. Iraqi C2 was highly centralized. With Saddam Hussein’s fear of internal threats to his rule, he has discouraged individual initiative while emphasizing positive control. Iraqi military commanders were authorized to do only that which was directed by highest authority. Destruction of its C2 capabilities would make Iraqi combat forces unable to respond quickly to Coalition initiatives.

Baghdad bridges crossing the Euphrates River contained the multiple fiber-optic links that provided Saddam Hussein with secure communications to his southern group of forces. Attack of these bridges severed those secure communication links, while restricting movement of Iraqi military forces and deployment of CW and BW warfare capabilities. Civilians using those bridges or near other targets at the time of their attack were at risk of injury incidental to the legitimate attack of those targets.

Another reason for collateral damage to civilian objects and injury to civilians during Operation Desert Storm lay in the policy of the Government of Iraq, which purposely used both Iraqi and Kuwaiti civilian populations and civilian objects as shields for military objects. Contrary to the admonishment against such conduct contained in Article 19, GWS, Articles 18 and 28, GC, Article 4(1), 1954 Hague, and certain principles of customary law codified in Protocol I (discussed below), the Government of Iraq placed military assets (personnel, weapons, and equipment) in civilian populated areas and next to protected objects (mosques, medical facilities, and cultural sites) in an effort to protect them from attack. For this purpose, Iraqi military helicopters were dispersed into residential areas; and military supplies were stored in mosques, schools, and hospitals in Iraq and Kuwait. Similarly, a cache of Iraqi Silkworm surface-to-surface missiles was found inside a school in a populated area in Kuwait City. UN inspectors uncovered chemical bomb production equipment while inspecting a sugar factory in Iraq. The equipment had been moved to the site to escape Coalition air strikes. This intentional mingling of military objects with civilian objects naturally placed the civilian population living nearby, working within, or using those civilian objects at risk from legitimate military attacks on those military objects.

The Coalition targeted specific military objects in populated areas, which the law of war permits; at no time were civilian areas as such attacked. Coalition forces also chose not to attack many military targets in populated areas or in or adjacent to cultural (archaeological) sites, even though attack of those military targets is authorized by the law of war. The attack of legitimate Iraqi military targets, notwithstanding the fact it resulted in collateral injury to civilians and damage to civilian objects, was consistent with the customary practice of nations and the law of war.

The Government of Iraq sought to convey a highly inaccurate image of indiscriminate bombing by the Coalition through a deliberate disinformation campaign. Iraq utilized any collateral damage that occurred including damage or injury caused by Iraqi surface-to-air missiles and antiaircraft munitions falling to earth in populated areas in its campaign to convey the misimpression that the Coalition was targeting populated areas and civilian objects. This disinformation campaign was factually incorrect, and did not accurately reflect the high degree of care exercised by the Coalition in attack of Iraqi targets.

For example, on February 11, a mosque at Al-Basrah was dismantled by Iraqi authorities to feign bomb damage; the dome was removed and the building dismantled. US authorities noted there was no damage to the minaret, courtyard building, or dome foundation which would have been present had the building been struck by Coalition munitions. The nearest bomb crater was outside the facility, the result of an air strike directed against a nearby military target on 30 January. Other examples include use of photographs of damage that occurred during Iraq’s war with Iran, as well as of prewar earthquake damage, which were offered by Iraqi officials as proof of bomb damage caused by Coalition air raids.

Minimizing collateral damage and injury is a responsibility shared by attacker and defender. Article 48 of the 1977 Protocol I provides that:

in order to ensure respect for and protection of the civilian population and civilian objects, the Parties to the conflict shall at all times distinguish between the civilian population and combatants and between civilian objects and military objectives and accordingly shall direct their operations only against military objectives.

Paragraph one of Article 49 of Protocol I states that “’Attacks’ means acts of violence against the adversary, whether in offense or defense.” Use of the word “attacks” in this manner is etymologically inconsistent with its customary use in any of the six official languages of Protocol I. Conversely, the word “attack” or “attacks” historically has referred to and today refers to offensive operations only.

Article 49(1) otherwise reflects the applicability of the law of war to actions of both attacker and defender, including the obligation to take appropriate measures to minimize injury to civilians not participating in hostilities.

As previously indicated, the United States in 1987 declined to become a party to Protocol I; nor was Protocol I in effect during the Persian Gulf War, since Iraq is not a party to that treaty. However, the language of Articles 48 and 49(1) (except for the erroneous use of the word “attacks”) is generally regarded as a codification of the customary practice of nations, and therefore binding on all.

In the effort to minimize collateral civilian casualties, a substantial responsibility for protection of the civilian population rests with the party controlling the civilian population. Historically, and from a common sense standpoint, the party controlling the civilian population has the opportunity and the responsibility to minimize the risk to the civilian population through the separation of military objects from the civilian population, evacuation of the civilian population from near immovable military objects, and development of air-raid precautions. Throughout World War II, for example, both Axis and Allied nations took each of these steps to protect their respective civilian populations from the effects of military operations.

The Government of Iraq elected not to take routine air-raid precautions to protect its civilian population. Civilians were not evacuated in any significant numbers from Baghdad, nor were they removed from proximity to legitimate military targets. There were air-raid shelters for less than 1 percent of the civilian population of Baghdad. The Government of Iraq chose instead to use its civilians to shield legitimate military targets from attack, exploiting collateral civilian casualties and damage to civilian objects in its disinformation campaign to erode international and US domestic support for the Coalition effort to liberate Kuwait.

The presence of civilians will not render a target immune from attack; legitimate targets may be attacked wherever located (outside neutral territory and waters). An attacker must exercise reasonable precautions to minimize incidental or collateral injury to the civilian population or damage to civilian objects, consistent with mission accomplishment and allowable risk to the attacking forces. The defending party must exercise reasonable precautions to separate the civilian population and civilian objects from military objectives, and avoid placing military objectives in the midst of the civilian population. As previously indicated, a defender is expressly prohibited from using the civilian population or civilian objects (including cultural property) to shield legitimate targets from attack.

The Government of Iraq was aware of its law of war obligations. In the month preceding the Coalition air campaign, for example, a civil defense exercise was conducted, during which more than one million civilians were evacuated from Baghdad. No government evacuation program was undertaken during the Coalition air campaign. As previously indicated, the Government of Iraq elected instead to mix military objects with the civilian population. Pronouncements that Coalition air forces would not attack populated areas increased Iraqi movement of military objects into populated areas in Iraq and Kuwait to shield them from attack, in callous disregard of its law of war obigations and the safety of its own civilians and Kuwaiti civilians.

Similar actions were taken by the Government of Iraq to use cultural property to protect legitimate targets from attack; a classic example was the positioning of two fighter aircraft adjacent to the ancient temple of Ur as depicted in the photograph in Volume II, Chapter VI, “Off Limits Targets” section on the theory that Coalition respect for the protection of cultural property would preclude the attack of those aircraft. While the law of war permits the attack of the two fighter aircraft, with Iraq bearing responsibility for any damage to the temple, Commander-in-Chief, Central Command (CINCCENT) elected not to attack the aircraft on the basis of respect for cultural property and the belief that positioning of the aircraft adjacent to Ur (without servicing equipment or a runway nearby) effectively had placed each out of action, thereby limiting the value of their destruction by Coalition air forces when weighed against the risk of damage to the temple. Other cultural property similarly remained on the Coalition no-attack list, despite Iraqi placement of valuable military equipment in or near those sites.

Undoubtedly, the most tragic result at this intentional commingling of military objects with the civilian population occurred in the February 13 attack on the Al-Firdus Bunker (also sometimes referred to as the Al-’Amariyah bunker) in Baghdad. Originally constructed during the Iran-Iraq War as an air raid shelter, it had been converted to a military C2, bunker in the middle of a populated area. While the entrance(s) to a bomb shelter permit easy and rapid entrance and exit, barbed wire had been placed around the Al-Firdus bunker, its entrances had been secured to prevent unauthorized access, and armed guards had been posted. It also had been camouflaged. Knowing Coalition air attacks on targets in Baghdad took advantage of the cover of darkness, Iraqi authorities permitted selected civilians apparently the families of officer personnel working in the bunker to enter the Al-’Amariyah Bunker at night to use the former air raid shelter part of the bunker, on a level above the C2 center. Coalition authorities were unaware of the presence of these civilians in the bunker complex. The February 13 attack of the Al-’Amariyah bunker a legitimate military target resulted in the unfortunate deaths of those Iraqi civilians who had taken refuge above the C2 center.

An attacker operating in the fog of war may make decisions that will lead to innocent civilians’ deaths. The death of civilians always is regrettable, but inevitable when a defender fails to honor his own law of war obligations or callously disregards them, as was the case with Saddam Hussein. In reviewing an incident such as the attack of the AI-’Amariyah bunker, the law of war recognizes the difficulty of decision making amid the confusion of war. Leaders and commanders necessarily have to make decisions on the basis of their assessment of the information reasonably available to them at the time, rather than what is determined in hindsight.

Protocol I establishes similar legal requirements. Articles 51(7) and 58 of the 1977 Protocol I expressly prohibit a defender from using the civilian population or individual civilians to render certain points or areas immune from military operations, in particular in an attempt to shield military objectives from attack or to shield, favor or impede military operations; obligate a defender to remove the civilian population, individual civilians and civilian objects under the defender’s control from near military objectives; avoid locating military objectives within or near densely populated areas; and to take other necessary precautions to protect the civilian population, individual civilians and civilian objects under its control against the dangers resulting from military operations.

It is in this area that deficiencies of the 1977 Protocol I become apparent. As correctly stated in Article 51(8) to Protocol I, a nation confronted with callous actions by its opponent (such as the use of “human shields”) is not released from its obligation to exercise reasonable precaution to minimize collateral injury to the civilian population or damage to civilian objects. This obligation was recognized by Coalition forces in the conduct of their operations. In practice, this concept tends to facilitate the disinformation campaign of a callous opponent by focusing international public opinion upon the obligation of the attacking force to minimize collateral civilian casualties and damage to civilian objects a result fully consistent with Iraq’s strategy in this regard. This inherent problem is worsened by the language of Article 52(3) of Protocol I, which states:

In case of doubt whether an object which is normally dedicated to civilian purposes, such as a place of worship, a house or other dwelling or a school, is being used to make an effective contribution to military action, it shall be presumed not to be so used.

This language, which is not a codification of the customary practice of nations, causes several things to occur that are contrary to the traditional law of war. It shifts the burden for determining the precise use of an object from the party controlling that object (and therefore in possession of the facts as to its use) to the party lacking such control and facts, i.e., from defender to attacker. This imbalance ignores the realities of war in demanding a degree of certainty of an attacker that seldom exists in combat. It also encourages a defender to ignore its obligation to separate the civilian population, individual civilians and civilian objects from military objectives, as the Government of Iraq illustrated during the Persian Gulf War.

In the case of the Al-Firdus bunker, for example repeatedly and incorrectly referred to by the Government of Iraq and some media representatives as a “civilian bomb shelter” the Coalition forces had evidence the bunker was being used as an Iraqi command and control center and had no knowledge it was concurrently being used as a bomb shelter for civilians. Under the rule of international law known as military necessity, which permits the attack of structures used to further an enemy’s prosecution of a war, this was a legitimate military target. Coalition forces had no obligation to refrain from attacking it. If Coalition forces had known that Iraqi civilians were occupying it as a shelter, they may have withheld an attack until the civilians had removed themselves (although the law of war does not require such restraint). Iraq had an obligation under the law of war to refrain from commingling its civilian population with what was an obviously military target. Alternatively, Iraq could have designated the location as a hospital, safety zone, or a neutral zone, as provided for in Articles 14 and 15, GC. [...]

B. Minutes of evidence taken before the Defence Committee – House of Commons UK – on Wednesday, March 6, 1991

[Source: House of Commons Defence Committee Report on “Preliminary Lessons of Operation Granby”, Minutes of evidence taken before the Defence Committee, March 6, 1991, p. 38]

Mr Home Robertson (Former Secretary of State for Defence)

  1. Can I come back to a question which I should have asked at the very beginning when we were talking about joint command structure? I apologize for coming back at the fag end on this important question of the allocation of missions and selection of targets. Was there always consensus between yourself and your counterparts on that subject or was there any occasion when you decided, for whatever reason, that it would not be appropriate for the Royal Air Force to attack a particular target?

Air Vice Marshal Wratten (Air Vice Marshal W. J. Wratten, CB, CBE)

Yes, there were such occasions. In particular, when we were experiencing collateral damage, such as it was, and some of the targets were in locations where with any weapon system malfunction severe collateral damage would have resulted inevitably, then there were one or two occasions but I chose not to go against those targets, but they were very few and far between and they were not – and this is the most important issue – in my judgment and in the judgment of the Americans of a critical nature, that is to say, they were not fundamental to the timely achievement of the victory. Had that been the case, then regrettably, irrespective of what collateral damage might have resulted, one would have been responsible and had a responsibility for accepting those targets and for going against them. But towards the end there were, I think, two occasions when I chose not to, when I chose to go against alternative targets. [...]


  1.  Do you accept the US definition with regard to targeting?
  2. Which measures said to have been taken by the US correspond to Protocol I? Which ones go beyond what is required by Protocol I? And which ones are below the standards set by Protocol I?
  3. Under IHL, which reasons given by the report in relation to collateral damage and injury are pertinent? And which one is unacceptable? (P I, Arts 51, 52 and 57)
    1. Neither the Conventions nor the Protocols mention the principle of proportionality as such. From which source is this principle derived? Are the consequences of the principle of proportionality codified in IHL? Does this concept allow some attacks which would be normally prohibited by IHL? Could the concept of military necessity be used by one party to the conflict to justify collateral damage? (P I, Arts 51(5)(b) and 57(2)(a)(iii))
    2. Is the report correct when it states that IHL “precludes collateral damage of civilian objects or injury to noncombatant civilians that is clearly disproportionate to the military advantage gained in the attack of military objectives”? (P I, Arts 51(5)(b) and 57(2)(a)(iii))
    3. Can a factor weighed in the proportionality test be the overall campaign objective, as stated in the report, such as the liberation of Kuwait? (P I, Arts 51(5)(b) and 57(2)(a)(iii))
  4. In international armed conflict, which precautionary measures must be taken by the parties before launching an attack? (P I, Arts 48, 50, 51, 57 and 58)
  5. Does a military objective become immune from attack if it is situated among the civilian population? (P I, Art. 57) According to US officials, Iraq systematically used this tactic. What was the reaction of the Coalition towards this situation? Did the Coalition forces target some military objectives although they expected disproportionate civilian losses? Did the Coalition forces always reach a consensus on the targets chosen for attack in Iraq?
    1. Can it rightly be argued that the Iraqi electric power grid was a legitimate military objective? (P I, Preamble, para. 5, and Art. 52(2))
    2. Does the concept of military advantage allow the Coalition forces to determine if an object is a military objective for the sole purpose of the Coalition war plan, namely the liberation of Kuwait? Would the advantage be assessed differently if the aim of the Coalition forces was not to liberate Kuwait, but to occupy a territory in violation of the UN Charter? (P I, Art. 52(2))
    3. Would an attack on the two fighter aircraft located next to the temple of Ur have been lawful even though the temple risked being destroyed? (P I, Arts 52(2) and 53)
    4. Was the Al-’Amariyah bunker a legitimate military objective if its description in the report is accurate? What should the US forces have done if they had known that there were civilians in the bunker? (P I, Arts 52(2) and 57)
    1. Is the alleged commingling of military objectives and civilian objects by Iraq a violation of Protocol I? Which examples of commingling of military objectives and civilian objects violate IHL and which ones do not? Does the reference to Art. 28 of Convention IV in the report concern attacks on Iraq, Kuwait, or both? Are only Kuwaiti civilians or also Iraqi civilians protected persons under Arts 4 and 28 of Convention IV? What are the legal responsibilities for the attacker if the defender uses civilians or civilian objects to shield military objectives? When is the attack prohibited? Which additional precautionary measures have to be taken? (GC IV, Arts 18(5) and 28; P I, Art. 51(7) and (8))
    2. Is “minimizing collateral damage and injury” a responsibility shared by the attacker and the defender? Do the defender and the attacker have a responsibility not to position military objectives among the civilian population? Do they have to build air-raid shelters for the civilian population if there are military objectives among them? Should the Iraqi government have evacuated the inhabitants of Baghdad to protect the civilian population? (P I, Arts 48, 51, 57 and 58)
    3. Are Arts 51(7) and 58 of Protocol I customary international law? Do these two provisions entail the same level of obligations? [See ICRC, Customary International Humanitarian Law]
    4. Does Art. 58 of Protocol I compel the defender to remove the civilian population from places near military objectives?
  8. If the presumption in Art. 52(3) of Protocol I did not exist, what would an attacker do in case of doubt about a military objective? In such a situation, may he attack this objective? Does the defender have “the burden for determining the precise use of the objective”? If military objectives are disguised as civilian objects, would this be a violation of IHL ? Which forms of camouflage are unlawful? (P I, Arts 52(1) and (2) and 57)
  9. If the targets discarded by Air Vice Marshall Wratten had been “fundamental to the timely achievement of the victory” could he really have accepted them, as he stated, “irrespective of what collateral damage might have resulted”? (P I, Arts 51(4) and (5), 52(2) and 57)
  10. Does the fact that neither the US nor Iraq have ratified Additional Protocol I imply that its provisions referred to above were irrelevant? Did the US and UK simply apply pre-existing customary law? Or was their assessment of customary law influenced by Protocol I?