On July 1, the White House issued a fact sheet summarizing its new report on civilian casualties from drones, also known as unmanned aerial combat vehicles, or UACVs. Evidently, this is a summary taken from a much longer report that has not been publically released. The release of the fact sheet has triggered virtually the same debate over the accuracy of the data and the effectiveness and humanitarian impact of such strikes. This is the same debate that has taken place since the first Gulf War, narrowly focused on drone strikes and casualties without any meaningful military or strategic context.

Civilian casualties and collateral damage do matter, killing key terrorist or insurgent figures is a valid issue and the reaction of the nearby population to such strikes may also matter. To this end, the United States should make every practical effort to limit civilian casualties, to limit collateral damage and to ensure that the targets of its strikes are valid hostile targets. But these goals are only part of the impact of using drone/UACV strikes in war. The problem with both the public portions of the White House summary and with the following media and think tank debates over its contents is that their focus is so narrow that they say virtually nothing about the real issues involved. In fact, they become something close to a debate over the need for some form of strike-by-strike “perfect war.”

Limited coverage

To begin with, drone strikes are only one small element of both the struggles against terrorists and the broader struggles against insurgents, and casualties are only one of the costs in terms of human suffering. Much of the debate over the July 1 report on the use of drones focuses on how accurate the estimate of civilians killed happens to be.

The White House summary fact sheet on the report makes an estimate for the period from Jan. 20, 2009 to Dec. 31, 2015, and does so only for 473 strikes by the US government against terrorist targets outside areas of active hostilities, and the assessed number of combatant and noncombatant deaths resulting from those strikes. The summary ignores the use of drones and other forms of air power in “areas of active hostilities,” currently including Afghanistan, Iraq and Syria. It never attempts to define combatant or the reasons for targeting given individuals. It simply uses the term “combatant” and “noncombatant,” without any supporting context to explain “combatant,” although it does say that:

“Noncombatants are individuals who may not be made the object of attack under applicable international law. The term “noncombatant” does not include an individual who is part of a belligerent party to an armed conflict, an individual who is taking a direct part in hostilities, or an individual who is targetable in the exercise of US national self-defense. Males of military age may be noncombatants; it is not the case that all military-aged males in the vicinity of a target are deemed to be combatants.”

It provides a very general description of the method used to estimate casualties and addresses the uncertainties involved in this method. The report estimates that the number of combatant deaths ranged from 2,372 to 2,581, and that the number of noncombatant deaths ranges from 64 to 116.

It also notes that nongovernmental organizations “generally estimate significantly higher figures for noncombatant deaths than is indicated by United States government,” based on information that:

“Can include both aggregate data regarding noncombatant deaths as well as case studies addressing particular strikes, and generally rely on a combination of media reporting and, in some instances, field research conducted in areas of reported strikes:

• Reports of noncombatant deaths resulting from US strikes against terrorist targets outside areas of active hostilities vary widely, such reporting generally estimates significantly higher figures for noncombatant deaths than is indicated by US government information. For instance, for the period between Jan. 20, 2009 and Dec. 31, 2015, nongovernmental organizations’ estimates range from more than 200 to slightly more than 900 possible noncombatant deaths outside areas of active hostilities.
• Nongovernmental organization reports of counterterrorism strikes attributed to the US Government – particularly their identification of noncombatant deaths – may be further complicated by the deliberate spread of misinformation by some actors, including terrorist organizations, in local media reports on which some nongovernmental estimates rely.

Although the US government has access to a wide range of information, the figures released should be considered in light of the inherent limitations on the ability to determine the precise number of combatant and noncombatant deaths given the non-permissive environments in which these strikes often occur. The US government remains committed to considering new, credible information regarding noncombatant deaths that may emerge and revising previous assessments, as appropriate.”

There are good reasons to ensure that every feasible effort is made to properly identify and target terrorist and extremist leaders, and to minimize civilian casualties and collateral damage to civil facilities. That said, it is critical to understand that this is only one small part of the effort needed to evaluate the use of drones in combat and the tradeoffs between efforts to minimize civilian losses and winning or losing a war.

To drone, or not to drone?

It equally is important to note just how selective the scope of the White House document is. No mention is made of the role of unmanned aerial vehicles in gathering intelligence and targeting, or of unmanned aerial combat vehicles in carrying out strikes in the combat zones in Afghanistan, Iraq and Syria. It does not address the relative role of manned versus unmanned aircraft in carrying out strikes against key combatants, nor does it discuss the relative efficiency of these two very different forms of airpower in killing combatants without killing civilians.

Like previous such reports, it never addresses the key tradeoffs between having a “man in the loop” in drone strikes versus “an aircrew in the cockpit” in manned airstrikes.

This is a key issue for two reasons. First, some critics of drones seem to feel that it is either unfair not to put a pilot into harms way or that civilian casualties are higher using drones/UACVs. However, it is unclear that any substantive analysis of the real-world issues involved with such arguments has been carried out or has been made public. The same is true of relative total cost per sortie and strike – including all of the battle management and systems costs – and relative effectiveness per sortie. There also does not seem to be a clear understanding of the fact that an aircraft using standoff precision-guided munitions in a climate of total air superiority is not really engaged in direct combat.

Second, a great deal has been said about the fact that civilians in the area of drones are hostile to their use. This usually ignores the fact that civilians in such areas are equally unlikely to be pleased with strikes by manned aircraft or by artillery strikes, especially if they are sympathetic to the combatants being targeted, lose people they know or lose property. It also assumes that their attitudes distinguish between a drone and a precision-guided weapon; that the reports of their attitudes are accurate; and that civilians on the ground can judge whether such a strike will help bring them lasting security or will have other benefits that offset their costs.

The linkage

Focusing only on drone/UACV strikes “against terrorist targets outside areas of active hostilities” also ignores the critical role that such strikes have played when they are used in combination with piloted airstrikes, and particularly with US Special Forces and other counterterrorism forces.

Being able to use drones and UACVs to shape a combined air and ground attack against key terrorist or insurgent leaders has played a critical role in supporting Special Forces and allied elite combat units. Ignoring their value in joint warfare and particularly in commando-style raids ignores one of the key tools used in defeating insurgents in Iraq between 2006 and 2011 and by elite strike units in Afghanistan.

It is also important to note that most of the strikes “against terrorist targets outside areas of active hostilities” in the report were carried out in Pakistan. Some were made in covert cooperation with Pakistani officials and military and at their request, generally against targets of benefit to Pakistan’s internal security. Others were made to deal with the fact the Pakistani Inter-Services Intelligence – supposed US allies – were giving sanctuary and support to the Taliban, Haqqani network and Al Qaeda.

These latter strikes would have led to a far higher level of confrontation with Pakistan if they had been carried out with manned aircraft. It also is questionable whether it was ever possible to be as certain of the presence or non-presence of civilians with manned aircraft. Last, the alternative would have been to deny the United States any way to strike such targets, giving the targets total sanctuary, and deny the United States some active leverage over Pakistan.

Measuring the lasting value

The White House fact sheet also does not address the effectiveness of such strikes, which is the key justification for launching them. President Barack Obama’s claims to have killed more terrorists than any previous administration, and the lists of key figures who have been killed over time, say nothing about the end effect. Instead, these claims are the equivalent of mounting the heads of game animals on a wall.

The White House fact sheet does not address any aspect of the value of such strikes in terms of weakening terrorist networks, weakening morale of the command structure or weakening the morale of insurgent forces. The debates over such strikes never seem to address any intelligence analysis of their lasting impact on hostile forces, terrorist networks and insurgent operations. This, too, is a critical issue in judging the real-world merits of both drone/UACV strikes on such targets in comparison to every other effort to kill or capture them.

If intelligence can confirm that drone/UACV strikes have a clear benefit in degrading, disrupting or defeating terrorist and combatant organizations and forces inside and outside areas of active hostilities, this needs to be made public in as much detail as possible without compromising ongoing operations. Striking at such targets is not a matter of collecting trophies; it is a matter of achieving important combat outcomes. The justification for civilian losses is military effectiveness – not a higher elite body count. The White House and Department of Defense have failed to make this clear to the public.

Again, this issue involves the tradeoff between using both drone/UACV systems and piloted systems, particularly in any form of operation that involves precisions strikes outside areas of combat, or “in the rear.” Focusing on drones/UACVs without examining this tradeoff with aircraft with aircrews simply does not make any more sense in evaluating the lasting impact of such strikes than in any other aspect of assessing the value of drones relative to other combat systems.

This is all too clear from the United Nations casualty reporting on Afghanistan. Airpower continues to play a role in checking the Taliban, and a very low role in producing total civilian casualties. A United Nations Assistance Mission in Afghanistan report, “Afghanistan Annual Report 2015: Protection of Civilians in Armed Conflict,” shows a grim rise in civilian casualties from 2,969 killed and 5,669 injured in 2013 –before the withdrawal of most US and allied combat forces – to 3,700 killed and 6,833 injured in 2014 – the year most forces left – to 3,545 killed and 7,457 injured in 2015.

It also shows that only 3 percent of civilian deaths or injuries were caused by any form of airpower in 2015, compared with 13 percent of civilian deaths or injuries due to targeted and deliberate killings; 17 percent due to complex and suicide attacks; 21 percent due to IEDs; 37 percent due to ground engagements; and 5 percent other.

The UN Assistance Mission in Afghanistan attributed 62 percent of all civilian casualties to antigovernment elements and 17 percent to pro-government forces (14 percent to Afghan national security forces, 2 percent to international military forces and 1 percent to pro-government armed groups). Seventeen percent of all civilian casualties resulted from ground engagements between antigovernment elements and Afghan national security forces and could not be attributed to one specific party, and around 4 percent of civilian casualties resulted from unattributed explosive remnants of war.

This essay shows that singling out airpower, and any portion of airpower such as drone/UACVs, without looking at the overall pattern in civil casualties, human suffering and the extent to which airpower does or does not provide more protection and progress toward victory than other forms of combat, makes no sense at all. It also risks imposing a totally unrealistic standard – the search for “perfect war” in the use of airpower – while no matching constraints are imposed on any other form of combat, regardless of its effectiveness and impact on civilians.

Judging by the day

It is even more dangerous to assess the impact of drones/UACVs, or any other instrument of war, in terms of impact on civil casualties and the outcome of the fighting by the strike or by the day, and to do so without regard to the total course of the fighting and its impact on the total “butcher’s bill” in civilian losses.

The civil war in Syria is a horrible case in point. It may be debatable whether a decisive use of airpower early in the war could have produced a stable end in a relatively short period of time. What is not debatable is the cost of not using some form of decisive force.

Estimates by the US Central Intelligence Agency say that Syria is now a nation with 17.1 million people (its prewar population in 2011 was estimated at around 23 million). While no precise numbers exist, the number of Syrian civilians killed since 2011 has almost certainly risen to somewhere between 300,000 and 400,000. The number injured is usually roughly twice as high.

The UN refugee agency estimates that at the end of 2015, some 6.5 million people in Syria were internally displaced, including multiple times, and many others continued to flee to neighboring countries. More than 13.5 million people inside the country were in need of humanitarian assistance at that time and some 4.8 million were refugees. Some estimates put Syria’s gross domestic product at the end of 2015 at well under 30 percent of its already low level in 2011.

War is a terrible business at best, and so is de facto occupation by extremist terrorists and insurgents. Today, terrorists and insurgents, as well as some states like the Assad regime in Syria, fight largely using forms of asymmetric warfare, where civilians are used as weapons. On the one hand, there are both state and nonstate actors that ruthlessly attack civilians using suicide bombs, mass murder of minorities, booby traps, killing and torture of dissidents, the siege of civilian areas and other forms of murder and repression.

On the other hand, the same state and nonstate actors use civilians as human shields, travel with civilians to prevent attacks upon them, use civilian facilities and keep civilians in them or nearby, and keep civilians as mass hostages in urban areas and camps. They also make every effort to exploit all forms of US and allied attacks, exaggerate civilian losses, shift all of the responsibility to the attacker and find political ways to limit the use of any effective system that can strike against them.

This, too, has critical implications for the assessment of drones/UACVs and the use of other forms of airpower and military force.

First, the human costs of war need to be evaluated in terms of the overall course of the fighting. Reducing the use of drones/UACVs and airpower to minimize casualties in single strikes can become a critical limit on the ability to use force in a way that leads to a quick, decisive outcome; an outcome that can sharply reduce total civilian costs and losses over time. The wrong kind of restrictions, focused solely on the costs of individual strikes, can easily result in massive increases in the total butcher’s bill in war.

Second, the use of civilians in war may prevent the effective defeat of terrorists and insurgents. It denies or partially denies the United States and its allies the use of a critical tool in warfare, shifts the nature of the battle in the insurgents’ direction and can give sanctuary to key elements of the enemy. Decisive force does not necessarily mean ground forces. It does, however, mean sufficient military power to win, and to win quickly enough to minimize the total damage to civilians both from the war and in the form of whatever peace will follow.

Third, the more that the use of civilians as human shields and hostages is successful, the greater the incentive to repeat and expand their use. Placing unrealistic limits on the use of drones/UACVs, or any other form of force, scarcely discourages the use of human beings as tools of war. It instead creates a strong incentive.

And fourth, it creates a structure in which nations such as Syria and Russia can use airpower and other elements of force by far less demanding standards. Nations such as Pakistan can manipulate their support of terrorists and extremists, and grant them selective sanctuary, with near impunity. Once again, the cumulative human cost to civilians becomes much higher, and in ways that favor the abusive and more extreme use of force.

Waging war is intrinsically horrible, even when fought by conventional armies along lines close to the ideals set forth in The Hague Conventions and other, more demanding interpretations of the “laws” of war. The justification for war does not lie in the way individual aspects of war are waged, but in its net outcome and its cumulative costs and benefits. Today’s wars are also fought on asymmetric lines where at least one side is composed of state and nonstate actors that are willing to use almost any means to win, regardless of its impact on the civilians involved. This is no excuse for a lack of intelligent restraint on the part of the United States and its allies, but restraint that ends in increasing the total human cost of war is no restraint at all.

Anthony H Cordesman is chair in strategy at the Center for Strategic and International Studies (CSIS) in Washington. This essay was first published at www.csis.org.