Grading the Pompeo Certification on Yemen War and Civilian Protection: Time for Serious Reconsideration

Last week, Secretary of State Michael Pompeo provided certification to Congress that Saudi Arabia and the United Arab Emirates are undertaking “demonstrable actions to reduce the risk of harm to civilians and civilian infrastructure” in Yemen. In the absence of a waiver by the State Department, this certification was required under the McCain National Defense Authorization Act (Section 1290, to be exact)  to avoid a congressionally imposed cut off of US military assistance to those countries. Humanitarian concerns resulting from the three and a half year military campaign in Yemen brought Congress to this near boiling point. These concerns are, in significant respects, undisputed: even the seven-page memorandum accompanying the certification states, “The Administration recognizes that civilian casualties have occurred at rates that are far too high in the Saudi-led Coalition’s campaign in Yemen.” The memo then goes on to say, “The Administration assesses that Saudi Arabia and the UAE are undertaking some actions to reduce the risk of harm to civilians and civilian infrastructure resulting from Saudi-led Coalition military operations.” Now that this certification has been provided, Congress will need to answer: how adequate are these actions?

There are four specific actions listed in the memo: (1) use of a No Strike List (NSL), (2) a change to Rules of Engagement (ROE), (3) formation of the Joint Incident Assessment Team (JIAT), and (3) the Saudi government’s commitment to paying for a course on the Law of Armed Conflict (LOAC) and targeting practices. Since these are the four actions pointed to as justification for the certification decision with regard to civilian protection, it is critical to examine each of them in more detail. I speak to these from the perspective of a subject matter expert in civilian casualties as well as my role in the State Department until May 2017, both advising the Saudi-led Coalition regarding how to reduce civilian casualties and participating in interagency discussions on the conduct of the Yemen campaign.

● The No Strike List. The Saudi-led Coalition has been using a NSL since the very early days of the campaign. The US military provided elements of a NSL, while the United Nations also provided a separate list. Over time, USAID worked to add additional elements to the NSL; Saudi Arabia also added additional elements to the NSL over time. The challenge has been that this list is not always used in so-called “dynamic strikes,” which are not preplanned but instead may involve a rapid reaction to an immediate threat or unplanned target of opportunity. The result is a loophole in civilian protection big enough to allow airstrikes on hospitals and other protected entities.

● Changes in Rules of Engagement. The U.S. military provided LOAC training and some targeting training near the end of 2015. As a result of this training and separate civilian casualty mitigation advising from the State Department, the Saudi-led Coalition offered to have its ROE reviewed, and the US military provided, on the whole, relatively minor suggestions that the Saudi forces adopted.

● Establishment of the JIAT. One of the recommendations from the State Department advisory effort was the establishment of an official body to examine suspected civilian casualty incidents and identify refinements to improve the conduct of operations. This approach was modeled after a similar construct used by NATO in Afghanistan. The Saudi-led Coalition heeded this recommendation and launched the JIAT in early May 2016.

● Commitment to pay for a training course. A $750 million air-to-ground training package, announced in June 2017, includes provision for the US to provide a course on LOAC and targeting. Such courses were provided before that date. It is notable that this course was a requirement added to the June package by the State Department in an example of interagency involvement in military assistance. The originally proposed training package did not include this course. In a meeting with Deputy Secretary of State Blinken, he asked for my thoughts on the merits of this package. I replied that, since it was the same as the previous training package, if we were happy with the training outcome apparent in Yemen, then continuing the same approach made sense. The new course was added as a result.

Given these four actions, can they be considered adequate for reducing the risk of harm for civilians and civilian infrastructure, as Congress requires? One observation is that most of these actions occurred early in the campaign. The NSL has been used since the earliest part of the conflict in 2015, the changes in ROE occurred in late 2015, and the JIAT was established in May 2016. Those three actions occurred well before the resumption of hostilities in August 2016, which was seen by many to be a return to a higher level of civilian harm and damage to infrastructure observed in 2015, compared to a relative improvement in early 2016. The period after this resumption of hostilities was then punctuated by the notorious Funeral Hall Strike, which involved the highest total of civilian casualties seen in a single incident throughout the campaign.

The latest action cited in the Pompeo memo is the training course provided before June 2017. While it is laudable to provide improved training to Saudi Arabia in light of the chronic deficiencies seen during the course of the Yemen campaign, general training in LOAC and targeting is also unlikely to have significant benefits to the Coalition’s ability to reduce civilian harm. Such training has been provided since 2015, so this course is simply not new. In addition, it is entirely possible for a party to a conflict to cause significant numbers of civilian casualties and still comply with LOAC. This was seen in the NATO campaign in Afghanistan, and this is what I – and a UK military Judge Advocate General – observed in my work with the JIAT in 2016.

The only action of the four listed that really holds potential for improving the Coalition’s ability to better protect civilians and civilian infrastructure is the JIAT. The intent behind the formation of this body is frequently misunderstood. Rather than a body to perform independent legal investigations, the original purpose was to enable more effective operational learning so that the execution of strikes would improve over time. The biggest challenge facing the JIAT is that ultimate success does not rest in its hands alone. The JIAT was responsible for identifying deficiencies – and ideally, patterns that increased risk to civilians – but these insights needed to be accepted by the Coalition, which needed to then make operational adjustments to reduce the identified risks. In my early work with the JIAT, they were not perfect by any means, but they were able to identify some risk factors. However, they lamented the Coalition’s lack of receptivity to operationalize solutions to these deficiencies. And the US government did not press for specific actions in these identified areas, sticking instead to more general statements calling for improved protection of civilians. And, indeed, Pompeo’s memo itself echoes those abstract statements, “The United States has called on Saudi Arabia and the UAE to conduct thorough, transparent, and expedient investigations in cases where airstrikes have caused harm to civilians and civilian objects.” What good has that done thus far?

These, then, are the factors that the US government has relied upon for its certification: a Coalition that set up a NSL that exists but is not used in some of the most consequential operations; modest changes in Rules of Engagement early in the campaign; creation of a body that has been insufficient in itself to drive operational improvements; and the Saudis’ commitment to pay for a training course that provides material already provided to the coalition earlier in the campaign with no apparent improvements. The Pompeo memo states “recent civilian casualty incidents indicate insufficient implementation of reforms and targeting practices.” Considering that three of the four actions occurred more than two years ago, and the fourth action is merely a repeat of earlier training that did not seem to help, this begs the question: if the US government is not happy with the outcome, then does relying on actions shown over time to be ineffective make sense?  While it seems late to have a serious conversation about alternatives so late in the campaign, late is better than never. Given that working with partners seems to be the new American way of war, this is a conversation we need to start having more often. 

About the Author(s)

Larry Lewis

Director of the Center for Autonomy and Artificial Intelligence at Center for Naval Analyses. Lewis spent a decade analyzing real world operations as the project lead and primary author for many of the Department of Defense 's Joint Lessons Learned studies.