The recent groundbreaking investigation by the New York Times into civilian casualties in Iraq uncovered a shocking truth for the many Americans who believe the United States is engaged in the “cleanest” and “most precise” war of all time: the U.S.’s targeting system in Iraq is broken and, as a result, U.S. airstrikes there are killing far more civilians than the U.S. acknowledges. What is even more disturbing is that we can now safely assume that the number of uncounted civilian deaths extend far beyond Iraq into Syria, Afghanistan, and other countries the U.S. is bombing.

Now that we know the system is broken, the U.S. has a moral obligation to fix it.

There are numerous reasons why the system is broken, but the following four stand out in particular. When combined, these four problems create the perfect storm, as detailed by the Times:

• In recent years, the strategic military incentives that come with protecting civilians have decreased.

• There have been new shifts in targeting tactics, techniques, and procedures.

• Targeting practices and policies are taking a guilt by association approach.

• There has been a reduction in the military’s investigation resources that can help keep these problems and the civilian harm they cause hidden.

Problem 1:

The perceived strategic value of minimizing civilian casualties (known in the military as CIVCAS) and protecting civilians has diminished among military planners while their tolerance for such casualties has risen.

There are no longer thousands of troops in contact with the civilian populations they are supposed to be protecting. They no longer have strong bonds with the people nor are they endangered by IEDs and other risks.  As a result, there has been a significant loss of focus on civilian casualties bereft of this strategic imperative to protect troops by protecting the population. This is most evident in the huge cuts to CIVCAS analysis, even as the conflicts have expanded. Until recently, Operation Inherent Resolve (OIR) had only two analysts on their Coalition Civilian Casualty Assessment Team for both Iraq and Syria, compared with a team of more than a dozen in Afghanistan in 2011. Though the OIR team has recently grown to seven analysts, this is woefully small for the huge numbers of cases needing investigation.

At the same time, the U.S. military is more tolerant of killing civilians than it has a decade ago when tactical directives increased civilian protection. Where the goal had once been zero civilian deaths, the shift in rules allowed for a sliding scale, with cases of ten civilian deaths being acceptable, particularly in urban areas. The use of explosive weapons in populated areas has often led to massive numbers of civilian deaths, such as a May 2017 airstrike against a sniper that killed over 100 Iraqi civilians. This directly leads to the next problem.

Problem 2:

Without the strategic military incentives to minimize harm, it is less surprising that U.S. civilian casualty airstrike claims in Syria have eclipsed Russian civilian casualty airstrike claims there for the first time. This is despite the fact that even under its relaxed rules, the U.S. military takes far greater precautions against civilian casualties than the Russians. Every bomb the United States drops is a precision-guided weapon, the most accurate in the world, and many are designed specifically to minimize collateral effects on civilians. In contrast, the vast majority of Russian aerial bombs are unguided dumb bombs.

The shift in the war against ISIS to urban centers contributed to the increase in civilian casualties, as enemy targets were in closer proximity to civilians than in rural areas. Military experts often cite this reality for the rise in civilian casualties. Other wars have demonstrated that this factor might be overstated, so it needs to be interrogated. Careful targeting procedures and adherence to international humanitarian law, or the laws of war, have done a better job at sparing civilians in populated areas in the past. When I was working for the United Nations investigating civilian casualties from NATO airstrikes in Libya it was remarkable how few civilians perished – fewer than 100 from 7,700 weapons dropped, many in urban areas. And while protection of civilians was a stated goal of the intervention in Libya it was quickly overtaken by regime change, and as others have pointed out, there have been urban battles against ISIS before without such high CIVCAS. So why else are we seeing a rise in civilian casualties now?

In the past, those bombs would have required high level vetting, but new Tactics, Techniques, and Procedures (TTPs) have changed from the careful days of “tactical directives” – a set of rules that NATO’s International Security Assistance Force (ISAF) first put in place in 2007 to place restrictions on targeting to improve civilian protection.

These tactical directives worked. They were a major contributing factor to the reduction in civilian deaths by airstrikes in Afghanistan by nearly 70% from 2008 to 2012. In 2015, I was on the USS Theodore Roosevelt working on a project on post-strike analysis in Iraq and Syria for the Navy. At that time every strike directed against a building had to be authorized at flag rank – by an admiral or general. That was even for buildings that weren’t necessarily thought to be occupied.

In practice, this means that a strike approval coming from a flag officer sitting in an operations center has myriad checks and balances. There are lawyers, structural engineers, intelligence analysts, and a host of other specialists working together to ensure maximum military gain with minimum civilian harm. Relinquishing this authority to troops in the field potentially curtails many of the checks and balances built into the system to protect civilians.

In the spring of 2016, under the Obama administration, the requirement for flag officer approval was dropped. By the time civilian casualties began to spike in 2017 the targeting authority rested with the ground commander.

This isn’t a shock to those of us working civilian casualties on a daily basis – we tracked US CIVCAS claims as they skyrocketed in 2017.

Problem 3:

Another problem is the military’s longstanding guilt-by-association approach to targeting. When the United States kills people who are in close proximity to its intended target, it presumptively counts them as non-civilians, according to the New York Times investigation, as well as many other sources over the years. It seems to take this approach both when it decides whom it can target and when it conducts a post-strike review of who it killed.

Looking at the case of Bassim Razzo whose family was killed in an October 2015 coalition airstrike, we can see this in action. Razzo’s family was in their home in Mosul, Iraq. But because the U.S. targeteers were expecting to hit ISIS, they didn’t see a civilian presence when conducting their pattern of life assessment, the observation of the target before a strike, so the target was struck.  This prejudice to assume anyone in the vicinity of a target is a non-civilian flies in the face of IHL’s presumption of noncombatant status. Meaning the targeteers should have looked for an ISIS signature as reason to strike, not a civilian one not to strike. The United States admitted in their “CIVCAS Allegation Closure Report” for the Razzo case that their intelligence was faulty, and a more complete targeting assessment may have discerned a civilian presence. Similarly, in the case of a mosque bombed in Syria by U.S. F-15s, CENTCOM lumped all the dead together – save for one who was of “small stature” – and counted them all as terrorists who were lawful targets. This happened even as the United States admitted a failure in target characterization and admitted it had not spoken to anyone on the ground during its CIVCAS investigation. After an exhaustive investigation the UN accused the United States of violating IHL in this strike.

Problem 4:

The United States no longer prioritizes, properly staffs, or properly conducts CIVCAS investigations and has ceded that ground to the UN, human rights groups, and journalists like Azmat Khan and Anand Gopal who investigated the Razzo case. The tiny U.S. investigation team rarely speaks to people on the ground to gather valuable information about CIVCAS and has turned information sharing into a veritable one-way street where they receive reports but provide little feedback of value.

When I ran the UN’s CIVCAS team in Afghanistan in 2011, I met with ISAF officials on a daily basis; everyone from ISAF’s commanders Gens. Petraeus and Allen, to their staffs, but most importantly with the Civilian Casualty Mitigation Team (CCMT). This team of skilled professionals had numbers, expertise, and clout. Their large team of over a dozen full-time analysts for just Afghanistan worked at ISAF headquarters and drew from the many military disciplines needed to understand when mistakes were being made and how to rectify them. They deployed teams of field investigators when cases warranted and regularly engaged with the population. Allied nations also took CIVCAS seriously, placing many of their members on the CCMT. Today, sadly, some coalition allies including the UK do not even allow the US to report on CIVCAS caused by their forces. When ISAF left in 2014, so too did the CCMT.

How do we fix this?

The gloves need to be put back on. Civilian protection needs to be at the core of the U.S. mission against ISIS and other adversaries. To do this, the U.S. needs to lower the threshold of CIVCAS tolerance back to zero anticipated civilian deaths. In reality, the number of civilian deaths will never be zero as long as bombs are dropped, but that nevertheless needs to be the goal. It should never be tolerable to kill those we are trying to protect. The tactical directives used by NATO in Afghanistan should be reinstated by the U.S. and updated with civilian protection in mind. Targeteers must presume noncombatant status from the beginning of their analysis, and targets should be only engaged when it is clear those being targeted are actively participating in hostilities. Terms such as “suspected militants” only legitimize this weakening of the core principals of IHL and guilt-by association targeting must cease. Finally, there needs to be a focus on CIVCAS analysis and mitigation. Some have suggested an office in DOD should be created to give it the prominence it needs, but at a minimum a dedicated capability, with investigations, assessment, and coordination functions – working directly for the Commander – should be a requirement of all operations or commands working on CIVCAS in kinetic operations across the Middle East and Africa that can implement changes to TTPs rapidly when problems arise. They need either to be able to conduct field investigations as the UN and NGOs do or work with allies or the UN to help augment their investigations. Indeed, only working on the ground will provide ground-truth. Post-strike civilian casualty numbers must then be fed back into the targeting system because if anticipated numbers don’t match what is being seen on the ground they are useless.  Finally, the DOD CIVCAS teams need to be transparent and work with the UN, NGOs, and the press to improve their reporting. Civilian protection is too important to do alone.

“One civilian death is one too many” is a mantra I often hear from U.S. military commanders during my investigations. It is shocking that the U.S. is killing so many civilians and yet doesn’t know how many it has killed. We owe a lot to Azmat Khan and Anand Gopal for raising the issue of CIVCAS to such prominence. We owe even more to Basim Razzo and the families we have destroyed in the name of protecting them.

Image: Gokhan Sahin/Getty