This post originally appeared in Just Security

Is a driver speeding toward a military checkpoint launching a suicide attack or racing his pregnant wife to the hospital? Is a local man digging on a roadside at night planting an improvised explosive device (IED) or working his farm when the temperature is cooler? Is a resident who jumps up when troops burst into his home at 2am reaching for a gun or reacting in fear? In Afghanistan and Iraq, US troops have had to answer such questions repeatedly, often in split-second time. Civilian and military lives have depended on the accuracy of their determinations.

Under the US Standing Rules of Engagement (SROE), troops are allowed to fire in self-defense if they encounter someone demonstrating hostile intent, i.e., the “threat of imminent use of force.” Identifying such a threat presents challenges, however, especially when enemy forces blend in with the local population. Mistaken determinations of hostile intent were a major cause of civilian casualties attributable to the United States in Afghanistan and Iraq between 2001 and 2014. Tackling Tough Calls, a new report by the Harvard Law School International Human Rights Clinic, examines this problem. Drawing on interviews with combat veterans and current servicemembers as well as open source research, it shows how the US military could better protect civilians from such errors without jeopardizing the lives of its troops.

Civilian casualty data from Afghanistan illuminates the significant danger that hostile intent incidents pose to civilians. According to the United Nations Assistance Mission in Afghanistan (UNAMA), between 2008 and 2013, about 70 percent of civilian deaths from “pro-government forces,” including US and other international forces, were attributable to three types of operations: escalation of force (EOF) procedures, used primarily at checkpoints; search and seizure operations, especially night raids; and airstrikes, most notably unplanned “opportunity” strikes. All three frequently involve hostile intent determinations. A 2013 Defense Department study similarly found that mistaking civilians for the enemy, which commonly resulted from misperceptions of hostile intent, “was the primary cause” of civilian casualties in Afghanistan. Although the official “combat missions” in Afghanistan and Iraq have ended, the US military’s experiences there should inform its future operations and training of foreign forces. By focusing specifically on hostile intent, it could augment the civilian protection measures adopted during those recent conflicts.

The US military should begin by rectifying shortcomings in the rule of hostile intent, which permits troops to fire if they face an “imminent” threat. The latest SROE, which date to 2005, define “imminent” as “not necessarily … immediate or instantaneous.” This negative definition creates a vague standard that forces troops to rely on subjective decision making. The current definition of imminent is also excessively broad. It contrasts notably with NATO’s definition — “manifest, instant and overwhelming” — which is closer to the common understanding of the term. A clearer and narrower rule could reduce the risk that a civilian’s behavior is misconstrued as hostile.

The military should supplement such revisions to the SROE by providing troops with the means to interpret them. In Afghanistan, US commanders of the International Security Assistance Force (ISAF) issued tactical directives that decreased civilian casualties from EOF procedures, night raids, and airstrikes. These directives had positive humanitarian effects and advanced ISAF’s efforts to build support among the Afghan people for its counterinsurgency mission. They could serve as models for detailed guidelines that deal explicitly with hostile intent. Unlike tactical directives, which are conflict specific, however, these guidelines would have to be applicable across conflicts.

In practice, the US military should maximize use of three existing tools in implementing the rule of hostile intent: training, leadership, and local engagement. Veterans of Afghanistan and Iraq, ranging from specialists to generals, emphasized the need for improved training in this area. While pre-deployment training has grown more realistic over the past decade, interviewees said that training could better reflect the complexities of hostile intent situations on the ground. For example, a sergeant described being surprised when he arrived in Iraq by the number of unofficial checkpoints set up by armed men without uniforms. The men appeared hostile to him, and he only figured out that they were a common and accepted presence in the region because his comrades seemed unconcerned.

Leadership is another important tool for improving hostile intent determinations. Field leaders help troops understand the rules of engagement and apply them in specific circumstances. They also set a tone for how the rules should be interpreted. Discussing the power superiors have to influence their subordinates’ actions, a former UNAMA official recalled encountering one US field commander who insisted his unit conduct “proper warfare,” i.e., aggressive operations, and another who was willing to risk an IED explosion in order to avoid killing civilians. The US military should ensure that all leaders have a strong grasp on the rule of hostile intent and provide their troops with clear and consistent guidance on how to implement it.

Engagement with the local population, through cultural training before deployment and relationship building on the ground, can further increase the accuracy of hostile intent determinations. One Special Forces officer encouraged the military to focus on teaching troops about local behavior, not just social mores. An Iraq veteran said he was taught to look for “strange objects by the road,” but found the lesson to be oversimplified and unhelpful because “Iraq is full of strange objects.” When he observed a child carrying trash across a highway in Iraq, he was struck that what proved to be an ordinary and harmless activity would have signaled an IED planting during training scenarios. If troops are aware of what is normal in their operational environment, they can better identify when an individual’s intentions are hostile.

Finally, in addition to taking steps to prevent incorrect determinations of hostile intent, the military should seek to learn lessons from any mistakes that do occur. It could, for example, institutionalize a civilian casualty tracking cell, akin to what ISAF used in Afghanistan, to gather and analyze casualty data. Whatever mechanism the military develops should incorporate qualitative as well as quantitative analysis of hostile intent incidents. Its focus would not be to prosecute troops who honestly believed their lives were at risk. Rather it would strive to identify the reasons for erroneous decisions in order to minimize them in the future.

Tackling Tough Calls recognizes the challenges of accurately determining hostile intent and the right of troops to self-defense, while underscoring the moral, legal, and strategic imperative of civilian protection. If the US military revised the rule of hostile intent, took full advantage of the available tools to implement it, and adopted effective mechanisms to learn from its mistakes, it could save civilian lives without unduly endangering military ones.

This post and the report it discusses are written solely in the author’s Harvard Law School capacity.