As top Pentagon officials, including Undersecretary of Defense for Policy Colin Kahl and Defense Secretary Lloyd Austin, consider the release of a final DoD Instruction (DoDI) on civilian harm by the end of this year, it remains to be seen whether DoD will re-institute Non-Combatant Casualty Cut-Off Values (NCVs), or an alternative non-numerical level of gauging the possibility of civilian casualties before U.S. airstrikes and other actions that may cause such casualties.
Pentagon officials and advocates for civilian harm mitigation have discussed the future of NCV, which the Pentagon removed from its doctrine in 2018 in a move that has not gained wide attention. Numerical NCVs gained currency in Operation Inherent Resolve (OIR) in Iraq and Syria during the Obama administration. The targeting process ostensibly took into account urban density, the value of a target and surrounding structures, such as residences, government buildings, apartments, mosques and bridges, to provide insights to commanders on whether to strike or hold off.
If a strike was likely to kill more persons than specified under the NCV, the strike commander would have to appeal to higher authority, whether the combined air operations center (CAOC) commander, the combatant commander, the defense secretary, or even the president.
Austin could offer a unique perspective on civilian casualties, as he was the head of U.S. Central Command (CENTCOM) from 2013 to 2016 during the Obama administration. While Austin wanted to avoid airstrikes over Iraq and Syria that caused civilian casualties, he expressed the difficulty of doing that and determining such casualty numbers without a significant U.S. ground presence. Austin also was known to give flexibility to commanders in the field who wanted to give the go ahead for an airstrike, but who were limited by a given NCV.
Sarah Holewinski, the director of Human Rights Watch’s Washington, D.C. office, said that NCVs are “critical for ensuring that there is some conversation around civilian casualties and what is acceptable and what is not acceptable when going into a particular theater, and without having an NCV, or something like that, you don’t have the forcing mechanism to have that conversation.” Holewinski served as the first senior advisor on human rights to the Chairman of the Joint Chiefs of Staff from 2016 to 2018–a position that is no longer in place.
DoD asserts that its collateral damage estimation (CDE) process is rigorous, that DoD routinely updates its CDE methodology with weapons and strike analysis, and that no need exists for operational effectiveness statistics on CDE aids, including the former Digital Precision Strike Suite Collateral Damage Estimation (DCiDE) algorithm and the current Digital Imagery Exploitation Engine (DIEE) tool, which incorporated DCiDE and resulted in a more streamlined and automated CDE process, according to DoD.
“Given that civilian casualties in military operations can result from a wide range of factors, including the conduct of the adversary, circumstances unique to particular operations, and the variety of DoD efforts to reduce the risk of civilian casualties, it’s generally not practical to develop statistics isolating the effectiveness of a particular tool, like DCiDE or DIEE, in reducing civilian casualties,” Mike Howard, a Pentagon spokesman, wrote in an email.
But Larry Lewis, the director of the Center for Autonomy and Artificial Intelligence at the non-profit research group CNA and an author of a 2010 civilian casualty study for the Joint Staff, said that the Pentagon needs to have such statistics to calibrate its CDE tools.
“I strongly disagree that it’s not practical to develop statistics on a CDE tool’s effectiveness,” Lewis wrote in an email. “Not only is it practical, it is needed in light of differences we observe between predicted and actual collateral damage outcomes.”
Marc Garlasco, the military advisor for PAX, a Dutch civilian protection organization, said that the Pentagon has not squared such predicted and actual civilian casualty numbers in his more than 20 years of work on the topic inside and outside the Pentagon. In 2011, Garlasco was the head of civilian protection at the United Nations Assistance Mission in Afghanistan (UNAMA), and he said he was at the Kabul headquarters of the former International Security Assistance Force (ISAF), the NATO-led mission in the country, every day to scrutinize ISAF targeting.
“Because there are no ground investigations (they were conducted during ISAF but stopped in 2015), the number of anticipated deaths is never checked against actual deaths,” per Garlasco. “In the few instances where it may be, the lack of investigations or incomplete investigations render it meaningless. DoD no longer talks to victims and witnesses as they used to. They have ceded this responsibility to NGOs and the U.N. I have conducted investigations of airstrikes in non permissive environments while with the U.N. and I must say if the U.N. can do it, then the DoD surely can.”
The apparent lack of U.S. verification of pre-strike estimates of civilian casualties with the actual numbers of post-strike casualties reduced confidence in the former NCV measure.
“If DoD does not validate the pre strike number of anticipated civilian deaths with a post strike analysis how do they know their models are any good?,” Garlasco asked. “There may be instances when they can use more force because their numbers are too high. Similarly, there are instances when they should use less force or alternatives because their numbers are too low. This plays into why the NCV served little purpose. If you can’t trust the number of anticipated civilian deaths in your model, what good is modeling them? Added to that is where they get the numbers. When I did targeting in 2003 for Iraq, we used historical census figures. I don’t know the last time Iraq conducted a census, but I am sure we used very old data that meant the models had a built-in bias. Hopefully, DoD is improving where it gets civilian data. I was working with a U.S. university to assist DoD in developing tools to improve the data –things like aggregating cell phone user data to see when people were in an area that was to be targeted. Using real time data they would have a better understanding of the real world situation on the ground.”
Replacing NCV
Garlasco wrote in an email that civilian harm mitigation (CHM) has replaced NCV.
CHM “is a cycle that exists to help mitigate civilian harm and includes positive actions taken to protect civilians during preparation (training), operational planning (creating a civilian casualty mitigation team–CCMT), employment (operations including casualty tracking & analysis), assessment (including investigations and accountability), response to civilian harm (including amends and changes to Tactics, Techniques, and Procedures), and the integration of lessons learned that feed the training cycle,” per Garlasco. “The U.S. military, NATO, and other allies have transcended the NCV with Civilian Harm Mitigation.”
Holewinski said that she believes the upcoming DoDI will not be a far-reaching reform but a note-taking of current practices.
“We believe it will be mostly focused on investigations of civilian harm and making amends for civilian harm, ex gratia compensation payments,” Holewinski said of the DoDI on a recent JIB/JAB podcast.
“None of that, of course, goes to real protection,” she said. “So, what do you do when you go into a new conflict? Do your operational planners have to take into account the civilian population and in what ways? Do your battle damage assessments after you’ve hit something include civilians and possible civilian casualties? Are you going to be talking with victims on the ground? Are you going to be talking with civil society about the data that they know? Are you going to make sure that your standards for considering evidence are a bit more broad so that you’re not just so narrowly defining what you consider to be evidence that will either go to showing civilian casualties or not? So a lot of things will not be covered by this [DoD] instruction, which means that we’re really still lacking a comprehensive policy on something the United States says is as important as civilian protection.”
Garlasco said that he hopes that the DoDI will help institutionalize the best practices of civilian harm mitigation.
“I am hopeful the new policy will further embrace CHM and CCMTs,” he said. “I am among those on a team of NGOs that have been advising the DoD on this DoDI and I am hopeful they will accept our recommendations.”
Ex Gratia Payments
While DoD said that it has authorized $3 million annually for ex gratia payments, the Pentagon did not disburse any funds last year for such payments. “I do not understand how we can authorize the military to make amends for harm they have caused, yet they choose not to do so,” Garlasco said. “I hope Congress will intervene.”
Combatant commands use the guidance in the classified Chairman of the Joint Chiefs of Staff Instruction (CJCSI) 3160.01C—No Strike and the Collateral Damage Estimation Methodology, a process modified since its adoption more than a decade ago.
CENTCOM used the guidance of CJCSI 3160.01C during OIR against ISIS targets in Iraq and Syria. Analysts perform civilian pattern of life analyses and use algorithms to inform the targeting process.
Under CDE, targeting planners examine the possible effects of various weapons to propose recommendations to lessen collateral damage to comply with rules of engagement (ROE) and the law of armed conflict (LOAC).
Asked what changes to targeting practices the Pentagon has made because of OIR, the Pentagon replied in an email that “among the many lessons reflected in CJCSI 3160.01C, No-Strike and the Collateral Damage Estimation Methodology, is the elimination of the use of Non-Combatant Cutoff Values in the CDE methodology.”
Regarding NCVs, “we stopped using them in 2018,” Mike Howard, the DoD spokesman, wrote in an email. “We removed them from our doctrine because we did not find them to be either effective or useful.”
But Chris Woods, the director of the United Kingdom-based Airwars, a non-profit that investigates civilian harm, suggested that NCVs “make commanders think through the implications of civilians on the ground because too often, particularly with airpower, those taking the actions are too divorced from populations on the ground.”
“It’s different when you’re on the ground with troops, and you’re interacting with populations,” he said. “When you’re not, when you’re disconnected, when your own ability to understand civilian harm is so civilian impeded, which I would argue the U.S. military still is, particularly in urban conflict, you have to have checks, and, if you don’t have checks, you’re going to end up with catastrophic civilian harm very quickly, as we found during the Battle of Mosul in certain areas.”
“There is strong evidence to suggest that the lower the NCV, the better the outcome for civilians,” Woods said. “Where the message from the top is clear, there’s an NCV of 1 or 5 here, civilian harm comes down. The higher the NCV goes, the more latitude you get, the greater the risk and the higher the risk of civilian harm getting out of control. The experience of communities on the ground often doesn’t match the understanding of those in the military who think they understand the implications of their actions.”
CDE Calibration
Woods suggested that no data is available that would indicate that CDE is properly calibrated to give an accurate measurement of the likely harm to civilians.
“If I felt that the Air Force had a robust system of properly understanding civilian harm from its actions, that the CDE was properly calibrated, that commanders had a proper feedback process, that they learned from where they were getting things wrong, I would be possibly more relaxed about the removal of NCVs,” Woods said. “But I don’t see that. I don’t see the U.S. military properly being able to determine civilian harm from its actions.”
The establishment of numerical casualty limits, which evolved into the term NCVs by 2011, came after 9/11, in response to the failed attempt to kill Taliban leader Mullah Omar, on Oct. 7, 2001–the first day of Operation Enduring Freedom in Afghanistan.
“Prior to 9/11, [a non-combatant casualty limit] wasn’t per se discussed,” said retired Air Force Col. Peter “Gunz” Gersten, the former deputy commander for operations and intelligence for Combined Joint Task Force-OIR. “It was inherent in the LOAC…in the concept of proportionality. It was always up to the combatant commander, sometimes down to the tank or the fighter pilot, to make that call. It was largely a judgment call. One of the biggest changes that drove proportionality to become a discussion of NCV was, as technology started to improve and the advent of the RPA [remotely piloted aircraft] in the battlespace, that technology allowed for others beside the combatant commander to be part of that proportionality discussion, given the [greater amount of] time [to target and fire].”
While front line forces and combatant commanders were to make such targeting decisions based on their LOAC training, the introduction of the armed General Atomics MQ-1 Predator drone after 9/11 to eliminate “high value targets” (HVTs), such as Omar and Osama Bin Laden, brought others into the proportionality discussion.
“Now we have an RPA, which presents [a picture] on the screen for everyone to see, and everyone feels compelled to be able to have a discussion–specifically the lawyers–and they want to quantify the morality and ethical dilemma that is often presented to commanders during warfare against an identified combatant,” Gersten said. “They were trying to quantify that morality, proportionality, and the ethical dilemmas that are faced every single time with commanders because they can. That attempt to objectively quantify proportionality in a potentially chaotic environment wrapped in ‘fog and friction’ of warfare generates NCV. That was enabled primarily by the RPA.”
On Oct. 7, 2001, the first armed Predator, the CIA-owned Predator 334, nicknamed “Tooth,” found Omar in Kandahar. At the time, Gersten headed the Predator operations cell under the command of then-Lt. Gen. Chuck Wald, the commander of U.S. Central Command Air Forces, and then-Brig. Gen. David Deptula, the director of the CAOC at Prince Sultan Air Base, Saudi Arabia for Operation Enduring Freedom.
“There was 100 percent identification of Mullah Omar walking out of his house getting into his car, and Gen. Wald attempted to make the call, and the lawyer said, ‘There’s four people around him. No. Those are civilians,’ and Gen. Wald had to stop because it was a CENTCOM lawyer [at CENTCOM headquarters in Tampa, Fla.] who said, ‘You cannot shoot him. We don’t know who those other people are,’ and I was standing there going, ‘Sir, those are his bodyguards. We know this,’ and they said, ‘You can’t prove that, major,’ and I’m like, ‘We’ve been watching him for three days. We have intelligence that’s him,'” Gersten said.
“And Gen. Wald was told directly by the lawyer, ‘You cannot shoot,’ and Mullah Omar got in his vehicle and drove away, and we tried to follow him, but we lost him,” he said. “We ended up chasing him for years. He ran the cells. He ran the Taliban. He ran the insurgency and the terrorist attacks.”
Afghan officials said in 2015 that Omar had died of tuberculosis in 2013.
CENTCOM Conference
After the failure to kill Mullah Omar on Oct. 7, 2001, CENTCOM held a conference several months later to create a “predetermined set of numbers, primarily to use against those that were questioning the lawful authorities of the combatant commander, primarily the lawyers and other second guessers,” Gersten said.
“The conversation that followed on was [from] the commander of CENTCOM, ‘Mr. President, we need to know before so that Gen. Wald can say, ‘NCV value of 6. There’s 4. I’m in,’ he said. “That would somehow settle the lawyer because they knew the NCV. Therefore, no matter what they said, NCV could be a definitive, objective, quantifiable number that the lawyers could be told to stand down on, therefore enabling the commander to make a proportionality judgment. The fallacy of NCV in that equation is you’re not accounting for the fog, friction, and timing of warfare–split second, life and death decisions. The combatant commander is taught his entire life to make those decisions and deal with the repercussions, if the decision’s bad. That’s why we have commanders do that. We don’t have lawyers do that, but NCV was a tool given to the commanders, as well as guidance from the senior commanders–the CENTCOM commander and the president–to say this is where we have a tolerance for it [civilian casualties].”
That effort led to a civilian casualty limit of 30 during Operation Iraqi Freedom in 2003 and to the creation of NCV less than a decade later. During much of the Obama administration, the NCV value for most targets was zero until ROE was relaxed at the end of 2016.
“Different HVTs ran with different NCVs,” Gersten said. “With Osama Bin Laden, you’d have an NCV value of 30, but if you had a low-level commander, his NCV was typically zero. We ran zero for the longest time.”
In 2015-16, Army Lt. Gen. Sean McFarland, the then commander of Combined Joint Task Force-OIR and Austin, then CENTCOM commander, entrusted subordinate commanders to order strikes based on the commanders’ LOAC and proportionality training and experience, Gersten said.
“If I was on the operations floor and executing a mission that was coming up against a high NCV number, I would go to Gen. McFarland, my boss, and say, ‘Sir. I’m about to run an operation. I think my NCV is going to be high. I think I’m going to need something like a 14 or 15 [NCV], but I can put a request in,'” Gersten said. “He got to the point where he was like, ‘Gunz, stop calling me. Call CENTCOM.” I’ve woken Gen. Austin up in the middle of the night and said, ‘Sir, I need authority for an NCV of 15. I need it now.’ And, I would tell you, 100 times out of 100, Gen. Austin said, ‘Gunz, do what you see best.’ What he was saying was, ‘Execute your commander’s proportionality [judgment].’ Every single time, Gen. Austin never questioned me, nor did my commander, Gen. MacFarland. So when the tool [NCV] had to be exceeded, they defaulted to the basic laws of armed conflict and proportionality because of the [need for rapid strike] time. When there was time to have a [targeting] discussion, it would take hours.”
Deptula, now the dean of the Mitchell Institute for Aerospace Power Studies, said that the zero NCV standard during much of the Obama administration’s operations against ISIS “far exceeds the standards of international law” in the Geneva Conventions. Deptula, a retired Air Force lieutenant general, is an architect of the 1991 air campaign against Iraq and commanded Operation Northern Watch, among a host of assignments.
“The danger of attempting to conduct immaculate warfare…gives the gift of time to the enemy, which is not what we want to do,” Deptula said. “We can and we must minimize unintended casualties, and every military planner rolls that into his or her calculus.”
“The [zero NCV] policy backfires, and the reason it backfires is it extends the time to secure U.S. or coalition military objectives and that, therefore, allows more time for an adversary, like the Islamic State, to commit atrocities,” Deptula said. “Under international law, you can have near certainty that civilians will be killed and still launch the strike. If civilian casualties do occur in an attack, the strike is not necessarily illegal under international law, even if the casualties are not the results of mistakes or errors. If it wasn’t legal to conduct an attack even when we’re certain that civilians will be killed, bad guys would be incentivized to surround themselves with civilians to create a legal sanctuary from attack, and, in fact, that’s what a lot of them have done.”
The Geneva Conventions’ Protocol I, adopted in 1977, suggests that an attack is unlawful if “it may be expected to cause incidental loss of civilian life, injury to civilians, damage to civilian objects, or a combination thereof, which would be excessive in relation to the concrete and direct military advantage anticipated.”
Inclement weather had prevented the use of full-motion video monitoring of the building for two days before the strike, and information provided to CENTCOM before the strike conflicts on whether there was, or was not, a high-risk of civilian casualties (CIVCAS).
Documents released last year by CENTCOM under the Freedom of Information Act said that then-Army Brig. Gen. John Richardson, the Target Engagement Authority (TEA) at the time at Camp Erbil, Iraq “was unaware that over 100 civilians were in the structure.”
“Had the TEA known of the large number of civilians, his proportionality assessment would have been drastically different, and he would not have approved the engagement,” per U.S. Air Force-Brig. Gen. Matthew Isler, the investigating officer for the March 17, 2017 incident, who said that he advised a change in tactics, techniques and procedures to reduce possible “CIVCAS entrapment” casualties after the March 17 airstrike.
Deptula suggested that effects-based targeting, using inert weapons, may help prevent loss of civilian lives.
Deptula said that he used inert weapons in Mosul and northern Iraq during Operation Northern Watch.
“Iraqi forces attempted to prevent attacks on their surface-to-air missile systems by [putting] them next to mosques,” he said. “I had my aircraft load up inert, cement 500 pound bombs. It surprised the hell out of the crew members, and my response to them was, ‘Look. 500 pounds of concrete going 500 knots will ruin your whole day, if it hits you.'”