CJ Chivers and Eric Schmitt have a lengthy and important front-page story in this past Sunday’s New York Times detailing scores of instances of civilian collateral damage from NATO air operations in the Libya fighting. Anyone dealing with laws of war, humanitarian intervention, targeting standards, and technology and war needs to read this article closely. Although I have no special factual knowledge beyond the public and media accounts, I incline to believe that these accounts are correct and that the real numbers are considerably higher. (This is apart from a separate question of violations of the laws of war committed by the Libyan rebels as NATO co-belligerents and whether NATO has any attributional responsibility for the actions of these forces, or for negligence in failing to take reasonable steps to restrain these forces. That legal issue is not considered in this post or in the original NYT article.)
I am an uncompromising advocate of precision technologies in warfare, including drones and other remotely controlled machines that, by removing the issue of soldiers defending themselves, allow for more careful targeting decisions. I’m also a strong advocate of the Obama administration’s targeting killing programs, both by the military and the CIA. That said, proponents of precision targeting through better technology have long risked raising expectations well beyond what technology supports at any given moment. Gains in civilian protection come incrementally, much of it through trial and error, a combination of technology and experience gained about operations using it in the field. Either over-promising too much, too soon or demands that any “improvement” meet a zero damage, strict liability standard has the effect of stopping improvement in its tracks because improvements come incrementally, one grinding step at a time. Magic bullets come after a long series of unmagic ones.
Unsurprising but nonetheless disappointing was the initial reaction of NATO officials to these Times reporters in their inquiries. It amounted to a shrug of the shoulders:
By NATO’s telling during the war, and in statements since sorties ended on Oct. 31, the alliance-led operation was nearly flawless — a model air war that used high technology, meticulous planning and restraint to protect civilians from Colonel Qaddafi’s troops, which was the alliance’s mandate. “We have carried out this operation very carefully, without confirmed civilian casualties,” the secretary general of NATO, Anders Fogh Rasmussen, said in November.
“Without confirmed civilian casualties”? Not confirmed only because NATO refused to make inquiries. Only after the Times reporters gave NATO officials a 27 page memo documenting collateral damage deaths in particular sites in Libya did NATO change its stance:
Two weeks after being provided a 27-page memorandum from The Times containing extensive details of nine separate attacks in which evidence indicated that allied planes had killed or wounded unintended victims, NATO modified its stance. “From what you have gathered on the ground, it appears that innocent civilians may have been killed or injured, despite all the care and precision,” said Oana Lungescu, a spokeswoman for NATO headquarters in Brussels. “We deeply regret any loss of life.”
That is not an offer by NATO to conduct an investigation; it’s a dismissal of the issue. NATO will conduct an after-action review in the form of “lessons learned” – a quite essential and useful exercise for guiding future operations – but apparently not a formal legal review of targeting. This seems to me a grave mistake, for reasons I’ll detail below – and particularly a lost opportunity for the United States. NATO should conduct a legal review of targeting in the Libya campaign, and the United States military should likewise undertake its own review of all NATO targeting in Libya. I do not find anything in the Chivers and Schmitt report that looks to be negligent under the proportionality standard for collateral damage in the laws of war. Liability is not the issue or the reason for a formal legal review (I suppose it might arise in a very extreme case of gross and wanton recklessness, but the legal presumption in favor of the commander’s good faith judgment on issues of proportionality is profound, and that does not appear to be an issue in what the Times writes).
The real reason for doing this is, instead, that the Libya hostilities offer an opportunity for the United States to examine targeting standards as state practice. State practice defended as lawful, as undertaken by the United Kingdom, France, and other allies, in an operation undertaken on their initiative, and with their own equipment, intelligence, legal officers, targeting standards and operational law.
This is an opportunity for the US to enunciate a legal baseline of state practice in targeting regarded as lawful in particular factual circumstances by its closest allies. And to do so in circumstances that are not Afghanistan after ten years of war, but instead a new situation, with vastly less intelligence and other capabilities than are available today in the Afghanistan conflict. Circumstances, however, in which the targeting, and collateral damage, was regarded by our closest allies and friends as being fully compliant with the laws of war. And a conflict in which the usual outside human rights groups were lobbying for it as humanitarian intervention, and so had far less incentive than usual to try and use the situation to raise the bar on what constitutes lawful targeting.
Thus, the fundamental reason US DOD should insist on an event-by-event legal review of targeting state practice is implicitly to reiterate that the standards applied in Afghanistan today are proper and virtuous – and also that they far, far exceed any legal requirement. And that the US’s NATO allies demonstrated important state practice which they regarded as lawful in the conduct of the Libya operations. Because the next time the US goes to war, it will likely be a lot closer in its initial intelligence and information about targeting to what just took place in Libya than what goes on today in Afghanistan. The US should be able to draw upon the detailed, incident-by-incident state practice of its closest allies, and their defense of those practices as lawful, as support for its own conduct.
That’s my basic policy take-away – a plea to DOD to insist on a serious analysis of what was regarded in Libya as lawful targeting, including acknowledgment of its limitations and regrettable yet lawful collateral damage. Ideally there would be an internal version. But also, importantly, a public version aimed at emphasizing that the “commander’s discretion” under the law of war is real and that the laws of war provide a firm safe harbor for commander’s decisions on targeting – and doing so on the basis of evaluating the particular targeting practices in the Libya operation by allied NATO forces.
Below the fold, a little more discussion of the Times article itself and the crucial role of ground level intelligence in the conduct of drone warfare and targeted killing.
I have pointed out in earlier blog posts at both Volokh and Opinio Juris that Libya is an important marker of operational targeting law. Chivers and Schmitt reach more or less the same conclusion:
The failure to thoroughly assess the civilian toll reduces the chances that allied forces, which are relying ever more heavily on air power rather than risking ground troops in overseas conflicts, will examine their Libyan experience to minimize collateral deaths elsewhere. Allied commanders have been ordered to submit a lessons-learned report to NATO headquarters in February. NATO’s incuriosity about the many lethal accidents raises questions about how thorough that review will be.
The lasting significance of NATO and US Defense Department review of targeting rules in Libya is different from what Chivers and Schmitt argue that it is, however. They argue that the failure to investigate collateral damage incidents mirrors the early period of the Afghanistan war, and that it was on account of investigations of collateral damage that led in Afghanistan to tightening of the rules on targeting, thus reducing collateral deaths:
NATO’s experience in Libya also reveals an attitude that initially prevailed in Afghanistan. There, NATO forces, led by the United States, tightened the rules of engagement for airstrikes and insisted on better targeting to reduce civilian deaths only after repeatedly ignoring or disputing accounts of airstrikes that left many civilians dead.
One can debate whether NATO – i.e., the US – repeatedly ignored or disputed these accounts. Far more important – and the real lesson of the debate over Libyan collateral damage – is that the tightening of targeting rules for airstrikes became possible over the course of that decade-long war mostly on account of two developments: better technology, such as drones and more precise weaponry, and crucially better intelligence. Or, more exactly, the accumulation of far greater intelligence and means of gathering intelligence about Afghanistan. One intelligence officer remarked to me a year ago that parts of Afghanistan are the most mapped areas on earth. The intelligence gathering runs from physical, political, and social geography to human intellligence gathered on the ground – which is, by many accounts, the key reason for the US success with its counterterrorism targeted killing campaign – but the accumulation over ten years is a, if not “the,” crucial reason why the US has been able to shift to narrower and more demanding targeting standards.
I would recommend to any journalists reading this post that they read closely Gregory McNeal’s new article on the extraordinarily fine-tuned targeting and damage mitigation standards in Afghanistan today. They have reached the point where “law” is not really an issue – the standards are so exacting that even one anticipated civilian death in a pre-planned operation requires the sign-off of the theatre commander or higher (at least as of that article’s final date). It is obvious, however, that no such level of planning, review, collateral damage mitigation exercise, etc., could meaningfully take place for most of the Libya operation – not from lack of desire to minimize collateral damage, but from lack of accumulated intelligence of all kinds by which such review could have some real effect.
One can insist in a targeting review on any standards one likes, in other words, but if you don’t know a vast amount about the target environment, it won’t actually improve your targeting. In Afghanistan at this point, that information includes detailed knowledge about the load-bearing of the earthen walls of compounds, anthropological surveys about likelihood of civilians in the streets after dark, and lots of other things that have never been imagined to be asked about Libya, let alone answered.
Although I don’t doubt that NATO/US officials wanted to avoid awkward questions about collateral damage early on in the Afghanistan war – who wouldn’t? – and although I don’t doubt that awkward questions helped raise the profile of the issue, the actual ability to do anything about it arose from technology, experience, and the accumulation of intelligence. The technology can be transferred from one place to another, and so can part of the operational experience. But intelligence accumulation, especially local geography in all its senses and ground-level human intelligence, as well as the operational experience that those convey with respect to the new precision technologies – no, those are not transferrable. And note that the US ability to engage in its remarkably low collateral damage operations has required many years of patient, unsung work by both military intelligence and the oft-reviled CIA.
So the ability to make good on demands for lower collateral damage is in considerable part a side benefit of an entire decade of war. If you want those kinds of possibilities in Libya, you need to have those kinds of intelligence capabilities on the ground. If you don’t have them, but you think – on grounds of imperative national security, imperative demands of humanitarian intervention, or any other reason – that you must use force, you can set your standards as high as you like, you can conduct as many after-action reviews, you can bring in the ICC prosecutor. Either you won’t attack at all, or else your targeting will continue to be less precise – by orders of magnitude less precise – than Afghanistan today. Fixing the problem requires not just insistent demands for it, and not just technology – but detailed intelligence, which takes time to develop. The key point is that targeting in Libya resembled Afghanistan in the early days because the intelligence framework was not there. As for drones – they are an important element in gathering intelligence, but when it comes to firing missiles, the drone is the last kinetic step in a long process of intelligence accumulation. The less intelligence, the less ability to use the drone precisely.
I worry that the US might be passively or inadvertently allowing the implicit legal bar to be raised regarding the quality of targeting. In some of my conversations with JAG lawyers who have served as legal advisors on targeting, they sometimes take for granted everything that is now available for the Afghanistan theater – and assume that it will be there for every other situation. It’s almost inevitable after a decade of fighting there – for many, their entire operational law careers have been framed by Afghanistan and Iraq. It is hard to keep in mind that there will be other wars in which there is not a decade of intelligence built up over years and on the backs of many unfortunate and regrettable – but lawful – mistakes.
I’ve raised this occasionally in guest lectures at the JAG school, and I find there is often a palpable double-take when students think about what it would be like to arrive in a new conflict. I think as well that there is much greater awareness of this issue precisely on account of the Libya fighting. But I also hope there is long term awareness in DOD and other agencies of the need to remind the public that the legal threshold is much, much lower than what we are able to do in Afghanistan. The casuistical, incident-by-incident analysis of state practice in Libya helps keep that in mind. And a public release of a version of that analysis, in order to emphasize precisely those limits to the US and NATO’s interlocutors in the public and among the advocacy groups.