Today marks the fifth anniversary of the horrific attack by the U.S. military on the Médicins Sans Frontières (MSF) trauma center in Kunduz, Afghanistan. Our global public consciousness should never forget the unfathomable tragedy that was inflicted on the patients, families, staff, and medical workers on that fateful day.
The discourse involving the attack on this functioning hospital engages with broader themes of enduring significance such as protections for medical providers and humanitarian workers in armed conflict and notions of justice and accountability following incidents of civilian casualties in war. This anniversary presents an opportunity to reflect both on the appalling tragedy of the Kunduz airstrike and on the events that followed.
One central issue following the Kunduz airstrike, and attacks that result in civilian casualties in general, is whether military accountability processes following such attacks are adequate. The apparent discrepancies between the findings of the official U.S. military investigation and the resulting dispositions of the servicemembers involved provide a striking example: How is it, in the case of this high-profile incident, that the official U.S. military investigation can find what appear to amount to serious violations of the law of armed conflict (LOAC) – namely, that the personnel involved “arbitrarily” chose a target, failed to “distinguish between combatants or civilians,” and engaged in a “facially disproportionate” attack – and still result in not one court-martial?
Because internal disciplinary processes are often utilized to address incidents such as the Kunduz strike, the accountability measures that are implemented can be difficult for the public to discern, let alone to evaluate. In this piece, I aim to provide a degree of clarity to the measures that were implemented following the Kunduz airstrike that is absent from the official narrative. In doing so, I also endeavor to address matters of broader and enduring concern involving military justice and accountability by examining the apparent disconnect between the findings of the official investigation and the disciplinary measures ultimately imposed.
Public discourse following attacks that result in civilian casualties is often focused on notions of retributive justice (for any potentially culpable military personnel) and on the prevention of similar incidents in the future. Both of these appeals are discernible, for example, in the questions posed by MSF to the U.S. military after the official investigation was released to the public. Meinie Nicolai, who was the MSF Association President when the U.S. military investigation was released, similarly appealed to both concerns in MSF’s “initial reaction” to the investigation. This piece engages with the retributive justice aspect of the Kunduz airstrike by exploring the administrative measures and processes that were implemented following the attack and situating them within the broader context of the investigation that preceded their imposition.
Describing the Administrative Measures Implemented
Little is publicly known about the specific measures that were imposed by the U.S. military in relation to personnel that were directly involved in or otherwise responsible for the tragic outcome in Kunduz. General Joseph Votel, commander of U.S. Central Command when the investigation was released to the public, revealed that the official investigation identified 16 personnel “whose conduct warranted consideration for appropriate administrative or disciplinary action, including a general officer.” According to General Votel’s remarks, the administrative measures adopted ranged from “suspension and removal from command, letters of reprimand, formal counseling and extensive retraining.”
I examine the specific administrative measures likely imposed in greater detail in a forthcoming critical assessment of the official investigation and the lessons “learned” therefrom. For present purposes, it is sufficient to point to General Votel’s observation that such adverse administrative measures “can carry severe repercussions on the careers and professional qualification of these individuals, that could include denial of promotion or advancement” and “possible separation from the service.”
While General Votel is correct to assert that administrative measures “can carry severe repercussions,” the potential consequences are considerably less severe than a criminal sanction such as fines, imprisonment, or an unfavorable discharge that can be imposed pursuant to a court-martial proceeding. The horrific unintended outcome of the attack and the seemingly incriminatory assessment of the airstrike that is reflected in the official investigation have led to quite reasonable questions regarding why administrative measures rather than court-martial proceedings were initiated in response to the attack.
Accounting for the Disconnect Between Investigation and Adjudication
The decision not to initiate court-martial proceedings has been the subject of extensive criticism in public discourse (see here, here, here for a few of many examples). The decision seems particularly troubling considering the extraordinarily damning findings of the military investigation.
The official explanation for declining prosecution is sparse on details. As General Votel explains, because the military investigation concluded “that the errors committed [by the personnel involved in the attack] were unintentional, and after considering other mitigating factors,” the relevant senior commanders determined that the administrative measures “taken against these individuals were appropriate to address the errors they made.”
It is true that the investigation repeatedly concludes that the personnel involved in the attack believed the airstrike would be against a Taliban-controlled compound and that they were not aware that the compound was in fact a hospital. For that reason, the characterization that the horrific outcome was “unintentional” is supportable. However, the investigation also repeatedly offers findings that seem to condemn the personnel involved in the attack for committing any number of ostensibly serious violations of applicable international law and U.S. military rules of engagement.
If it is true, for example, that neither the pilot of the aircraft that conducted the attack nor the ground force commander “distinguished between combatants and civilians nor a military objective and protected property,” or that the attack was “facially disproportional,” or that the aircrew “arbitrarily chose the building they engaged,” or that the attack “did not comply with either the governing” NATO or U.S. rules of engagement, the decision not to initiate criminal prosecution for anyone involved in what are ostensibly serious LOAC violations seems rather inexplicable. How is it, then, as General Votel asserts, that “from a senior commander’s perspective, the [administrative] measures taken against these individuals were appropriate to address the errors they made”?
The answer to that riddle is in two parts. First, the conclusions provided by an investigating officer in a report are just as they are described in doctrine and no more: findings and recommendations. A finding is defined doctrinally simply as a “clear and concise statement of a fact that can be readily deduced from evidence in the record,” while a recommendation is required to be “consistent with and logically based on the findings.”
As doctrine establishes, a commander responsible for considering the findings and recommendations of an investigation (described as the “approval authority”) “is neither bound nor limited by the findings or recommendations” of the investigating officer and “may approve, disapprove, modify, or add to the findings and recommendations, consistent with the evidence included in the report.” The second part of the answer is that the findings and recommendations of the Kunduz investigation are irreparably flawed.
Commander’s Focus: Factual Account Rather Than Investigation Conclusions
In the hundreds of investigations in which I have been involved in advising an approval authority as an Army judge advocate, one consistent trend is that the approval authority finds the factual record to be the most persuasive aspect of an investigation report. The investigating officer and, when assigned, investigating team, is appointed, above all, to answer one question: what happened? The approval authority will always consider the findings and recommendations reflected in the investigation report, but the doctrinal responsibility of an investigating officer “to ascertain facts, document and preserve evidence, and then report the facts and evidence” to the approval authority is what commanders care most about.
Investigating officers and approval authorities alike are advised by military lawyers, but ultimately a commander makes an independent disposition decision based on her own judgement and assessment of the factual record ascertained in the investigation. While this degree of authority that is vested in military commanders may be disquieting to some observers, such authority is advisable and rational since it is the commander – not her advisers or staff or investigating officers – that bears responsibility for mission accomplishment. Investigating officers, military lawyers, and, at times, members of the commander’s staff all provide valuable input, but disposition decisions following investigations are for commanders alone to make.
In the case of the Kunduz investigation, the factual record indicates that the aircrew observed the target compound, which was in fact the MSF trauma center, for 68 minutes before initiating the attack. During this time, they engaged in an extensive dialogue in an attempt to verify the identity of the proposed target. Although it is clear after the attack that the target identification was horribly wrong, at the time the personnel involved in the attack believed “with 100% certainty” (to borrow directly from a statement provided during the investigation) that they were attacking the correct target.
The approval authority’s emphasis on the factual record rather than the findings and recommendations of the investigation is evident from the official narratives provided by the senior commanders involved. For example, while General Votel observes that the “investigation concluded that certain personnel failed to comply with the rules of engagement in the law of armed conflict,” he later asserts that the personnel “were absolutely trying to do the right thing” and that there “was no intention on any of their parts to take a short cut, or to violate any rules that were laid out for them.” The findings of the investigating officer indicating that provisions of applicable law and policy were violated are defective because they are not supported by the record of facts ascertained during the course of the investigation, and General Votel’s reflection that the personnel “were absolutely trying to do the right thing” reflects his own independent assessment of the factual record.
The first part of the answer to the mystery of the decision not to prosecute, then, is that commanders are primarily interested in the factual record ascertained by an investigating officer and are far less persuaded by the findings and recommendations reflected in the report. This explains how it is that the Kunduz investigation report can contain a finding that the personnel involved in the attack failed to distinguish “between combatants and civilians” and “arbitrarily chose the building they engaged” while the authority deciding dispositions of the personnel involved can assert that those personnel “were absolutely trying to do the right thing” and therefore should not be subjected to criminal prosecution. The findings and recommendations of the investigation are not supported by the factual record derived by the investigating officer, and commanders are much more likely to rely on their own independent assessment of the factual record when making decisions regarding imposition of appropriate corrective or punitive measures.
This is predictable based on the doctrinal relationship between the investigating officer, whose role is simply to gather and interpret facts and make recommendations based on that interpretation, and the commander, whose role is to draw on the findings and recommendations presented and make an independent decision regarding what corrective or disciplinary measures are appropriate. The second part of the answer to the question of why there was no criminal prosecution is rather less predictable: extensive inaccuracies in the official investigation render the findings and recommendations reflected in the report unreliable and deficient.
Extensive Deficiencies of the Kunduz Investigation
A full assessment of the litany of failings of the official investigation is beyond the scope of this post (it is a primary subject of the project I am completing involving the assessment of the investigation and lessons “learned”), but a careful evaluation of the official investigation reveals that essentially the only thing the report gets right is the factual narrative.
The prevalence of the deficiencies of the official investigation can be attributed to three main factors. First, the investigation fails to differentiate the factual activities and doctrinal responsibilities of the personnel involved in the attack. Instead of considering the activities and responsibilities of each individual person, the report routinely attributes all the activities of the ground element that requested the airstrike, for example, to the ground force commander and all those of the aircraft to the pilot.
A second central shortcoming of the investigation is that it adopts a retrospective assessment rather than evaluating the process that led to the attack based on the facts that were known by the personnel involved in the airstrike at the time. It is true, for example, that the MSF trauma center did not in fact qualify as a military objective in retrospect, but the investigation ascertains overwhelming evidence to support the conclusion that the personnel prospectively believed at the time that the compound was controlled by opposing insurgents. Moreover, this belief was formed after extensive deliberation, which included observing the target area for over an hour in an attempt to ensure that the characterization of the compound as a military objective was correct.
The LOAC requirement not to “second-guess military decisions with the benefit of hindsight” is well established in U.S. military doctrine. A similar sentiment is also reflected in the statements, declarations, and reservations provided by states such as the U.K, Australia, Canada, Italy, among others, when ratifying Additional Protocol I to the 1949 Geneva Conventions. That the Kunduz investigation, without exception, assesses compliance with applicable law and policy requirements from an ex post rather than ex ante perspective renders the official findings deficient.
The third central shortcoming of the official investigation that I will address here is that the report routinely fails to articulate formulations of applicable law and policy that most informed observers would recognize as legitimate. For example, the finding that “any engagement of a target that is not a lawful military objective is facially disproportional” because the proportionality rule ostensibly “assumes that the target to be engaged is a lawful military objective” represents an unintelligible formulation of the rule purportedly being assessed.
The fundamental LOAC proportionality rule, which prohibits an attack in which the incidental damage expected is “excessive in relation to the concrete and direct military advantage” anticipated does not support the assertion that the rule is based on a retrospective assumption that the target is a “lawful military objective.” This is but one example in a litany of dubious formulations of relevant use of force legal and policy rules articulated by the investigation report against which the investigating officer applies the known facts to derive findings and conclusions.
Even this brief assessment is sufficient to elucidate how the findings of the official investigation report do not present a reliable narrative against which to measure the disposition decisions made by the various senior commanders following the completion of the investigation. The commanders were likely much more persuaded by the factual account than the findings and recommendations in the first instance, and the findings and recommendations reflected in the report are irreparably flawed anyway.
A full accounting of the deficiencies of the Kunduz investigation is not possible here. However, in my forthcoming comprehensive critical assessment of the investigation I demonstrate the assortment of official findings indicating that the personnel involved in the attack committed serious violations of LOAC rules to be legally deficient. Such findings are not supported by the factual record and are founded upon inaccurate formulations of the LOAC rules purportedly evaluated. These latent deficiencies account for the disparity between the seemingly damning characterizations of the conduct of the personnel involved in the attack that are reflected in the investigation and the decisions of the respective commanders who conducted an independent assessment of the factual record and decided to refrain from initiating court-martial proceedings.
Accountability, Justice, and the Kunduz Airstrike
With this depiction of the decision process and the failings of the official investigation in focus, the question of whether the administrative measures ultimately imposed were sufficient remains. Is it “justice” for military members responsible for killing 42 people, injuring scores more, and destroying a functioning hospital that never should have been attacked to face only administrative measures that “can carry severe repercussions on” their careers rather than criminal charges? Were the measures truly “appropriate to address the errors” made by the personnel directly involved in or otherwise responsible for the horrific outcome of the attack?
The answers to these unresolved questions are, of course, a matter of perspective. Perhaps the most challenging aspect of engaging with these issues, though, is disengaging the outcome of the attack from the process that led to the decision to strike. If assessing the process truly requires a prospective approach and does not permit the disposition authority to “second-guess military decisions with the benefit of hindsight,” the outcome of the attack is of little consequence – no matter how tragic, horrific, and utterly unimaginable it may be – when deciding what accountability and corrective measures should be implemented afterward.
This challenge with decoupling the process from the outcome is perhaps best captured by Ban Ki-moon’s comments to the Security Council in 2016. In the meeting addressing healthcare in armed conflict that led to adoption of the landmark Resolution 2286, the Secretary-General observed in his opening remarks, “When so-called surgical strikes end up hitting surgical wards, something is deeply wrong. Explanations ring hollow to parents burying their children and communities pushed closer to collapse.”
Ban’s remarks are, of course, correct. Explanations do ring hollow following such tragedies. As MSF’s Dr. Joanne Liu observed days after the Kunduz airstrike, such an attack “fundamentally undermines the core principles of humanitarian action” and therefore “cannot be brushed aside as a mere mistake or an inevitable consequence of war.”
While nothing about the process that led to an attack such as the Kunduz airstrike can change the tragic outcome, a full and complete account of the factors that led to the decision to engage in the attack is vitally important in the search for both justice and accountability. As unsatisfying as it may seem, this endeavor must be decoupled from the tragedy of the outcome in order to be effective.