In a hearing on April 6, 2022, Congresswoman Sara Jacobs (D-Calif.) asked Secretary Lloyd Austin whether the Department of Defense plans to “revisit cases that were likely prematurely dismissed as a result of faulty initial assessments” as part of its ongoing review of civilian harm policies and procedures. Secretary Austin responded, “At this point we don’t have an intent to re-litigate cases from before.”

Neglecting past cases would be a grave mistake that could damage the credibility of an otherwise promising process. Longstanding shortcomings in U.S. policies and practices have left two decades of civilian harm chronically under-investigated and undercounted, often unacknowledged, and almost never “litigated.” Any credible effort to reform the U.S. approach to civilian harm – as Secretary Austin has committed to do – must review and make amends for past harms even as it rightly looks to the future. Indeed, the Department cannot move forward without also learning from and reckoning with the past.

On Jan. 27, 2022, responding to congressional oversight, civil society advocacy, and numerous New York Times investigations, Secretary Austin directed the Department of Defense to develop a Civilian Harm Mitigation and Response Action Plan (CHMRAP) for improving how the Department prevents and responds to civilian harm. In particular, he directed the CHMRAP to provide for the establishment of a center of excellence on civilian harm mitigation and response; the development of standardized processes for collecting and learning from data related to civilian harm; improvements in how the Department responds to civilian harm, including through condolence payments and public acknowledgment; and the incorporation of guidance for addressing civilian harm in future doctrine and operational plans. According to the Secretary’s testimony, the Department is approximately 30 days into its 90-day timeline for developing the CHMRAP. (The Secretary also required that the long-awaited DOD Instruction on Civilian Harm be informed by the CHMRAP and finalized within another 90 days of the CHMRAP’s conclusion. We have written previously on our expectations for the DOD Instruction here.)

Secretary Austin’s directive, and the high-level focus on civilian harm it conveys, is a welcome step that could lay the foundation for much-needed structural reform after years of relative stasis. But to truly succeed in addressing systemic civilian harm issues, the Action Plan must include a commitment to reviewing and learning from past harms. And in that regard, the Secretary’s response to Rep. Jacobs suggests that the Department may be falling short.

Over the last twenty years, the U.S. military has regularly undercounted civilian casualties in its annual reporting, with a significant gap between the U.S. military’s numbers and those coming out of credible civil society observers as well as the United Nations. Persistent discrepancies partly stem from the kinds of sources upon which the DOD relies to assess civilian casualties. Research by our organization Center for Civilians in Conflict (CIVIC) and Columbia Law School Human Rights Institute reviewed over 220 administrative investigations into civilian harm and found that the U.S. military tends to rely solely on internal records and sources when assessing civilian harm and rarely seeks information from witnesses or survivors of attacks or visits the sites of the strikes. Our research also found that the U.S. military is often highly and unduly skeptical of external sources of information, such as reports from civil society and the media. This skepticism persists even though credible civil society organizations and media often undertake in-depth investigations using rigorous methodologies – including survivor interviews, physical site visits, and background documentation – and thus have access to information that the U.S. military lacks.

U.S. government-commissioned research corroborates our findings. The recently released RAND study, required by Congress and sponsored by the Defense Department itself, also acknowledged the large, problematic discrepancies between U.S. military and civil society numbers. The report found that “air campaigns have an inherent civilian-harm detection problem,” that military “data and records that support assessments of civilian harm can be incomplete,” and that “officials did not sufficiently engage external sources for information before concluding that reports of civilian casualties were not credible.” RAND recommended that the Department “expand the kinds of information available for assessments to make them more robust.”

Recent investigative reporting by the New York Times also found that the U.S. military often prematurely dismissed likely credible allegations of civilian harm in the initial assessment phase based on flawed reviews of basic facts and evidence. Among 1,311 civilian casualty credibility assessments that the New York Times reviewed in 2021, the vast majority (over 1,100) were dismissed as “noncredible.” Yet, in many cases that the Times reviewed, allegations were dismissed due to simple errors. Assessment officers misread allegations, confused similar town names, failed to conduct basic internet searches or consult public sources like Google Maps, relied on inaccurate strike log coordinates to corroborate allegations, and lacked necessary Arabic skills. And just as we discovered in our research, the New York Times also found that assessment officers often ignored relevant external evidence.

Because of these shortcomings, thousands of instances of civilian harm have likely gone unacknowledged. That means thousands of civilian survivors and grieving families are still waiting for recognition and amends. For them, these cases are not in the past but a daily source of grief, trauma, stigma, and hardship.

Reviewing past cases of harm is also critical to any attempts to learn for the future, as the Secretary’s directive rightly demands. Without a full picture of past civilian harm and the circumstances and causes of that harm, the Department cannot understand and learn from its mistakes. After too many years of failing to learn and implement lessons, the steps laid out in the Jan. 27 directive offer a valuable opportunity to finally do so – but only if the Department is willing to review, own, and address past mistakes.

Learning from past mistakes is a legal as well as an ethical imperative. Under the law of armed conflict as reflected in the Department’s own Law of War Manual, the DOD “must take feasible precautions to reduce the risk of harm to the civilian population and other protected persons and objects.” Yet relying on flawed past assessments of civilian casualties means that military operators will likely not understand how much civilian harm U.S. attacks produce, which is critical to informing precautionary measures. Indeed, as others have noted for Just Security, the failure to properly estimate civilian harm after a strike creates a “vicious cycle” that precludes proper pre-strike collateral damages assessments in the future. A persistent failure to learn may also amount to negligence.

In sum, it is essential that the forthcoming CHMRAP include a commitment to reviewing erroneously dismissed cases and issuing amends for credible cases, including with ex gratia payments and other offerings in consultation with victims. Such a commitment could ensure acknowledgement for victims and survivors waiting for recognition and justice and allow the U.S. military to learn for the future. Fail to do so, and the Department will undermine an otherwise promising effort to overhaul the U.S. approach to civilian harm.

IMAGE: Chairman of the Joint Chiefs of Staff General Mark Milley, Secretary of Defense Lloyd Austin and Under Secretary of Defense Comptroller Mike McCord testify before the House Armed Services Committee on Capitol Hill. (Photo by Kevin Dietsch/Getty Images)