The Inadequacy of the US State Department Report on Arms Exports Assurances: Part II

[William Worster has taught public international law, the law of international organizations and international migration and refugee law at The Hague University of Applied Sciences for more than fifteen years]

This is the second part of a post analyzing the US State Department Report on assurances it has received that exported arms will not be misused. In the first part, this post critiqued the analytical framework that the US State Department applied to assess the assurances it received. It concluded that the framework was insufficient because it asked only a few questions and largely avoided obvious and necessary questions that would clearly assist in making a good faith determination of risk. And by omitting those questions, the analysis may already be a violation of the US’ international obligations. In this second part, this post will change gears to examine the evidence relied on by the State Department in completing its analysis within the – poorly constructed – framework. It concludes that, just as the framework was inadequate, the selection and analysis of evidence was also inadequate. As a result, if there were any doubts that the US was not in compliance with its obligations due to the analytical framework, it surely is in violation due to a lack of engagement with relevant evidence. The obviously insufficient report suggests that the US did not complete a good faith analysis of risk and may, in turn, not be in compliance with its international obligations.

The US requested assurances from states receiving arms exports in order to comply with National Security Memorandum 20 (“NSM-20”). NSM-20 seeks to satisfy the US’ international and domestic legal obligations by requesting assurances from states when there is risk that exported arms will be misused. Among others, the international legal obligations come from the Geneva Conventions and Genocide Convention and require the US to prevent serious violations  of international law by other states using arms exported from the US. The specifics of these obligations have been provided in an earlier post and by many other authorities. In short, those obligations oblige the US to make a prediction of risk. The obligation to prevent is only partially backwards-looking. Past practices may be relevant but only insofar as they bear on making a good faith attempt to predict the future behavior of another state. As a part of this prediction, NSM-20 gives the US the option to request assurances – essentially a promise – that exported arms will not be misused. If the assurances are credible and reliable, then they convince the US that misuse is unlikely to happen. As such, any international obligations to prevent violations by prohibited export would not be implicated and the US would not violate its international obligations.

Limitations on the Scope of the Inquiry

Having issued NSM-20, the US State Department identified several states for which it had credible indications that they might be misusing exported arms. It requested and received assurances that the arms would not be misused to violate international law. It completed its analysis of those assurances and reported to Congress on its assessment. In the prior part, this post noted that the State Department Report only proposed considering whether the assurances were provided by a senior official with authority to make commitments, whether the individual was credible, and the likelihood that the partner government will comply. The first part of this post criticized the US for an inadequate framework for making its prediction of risk based on factors drawn from international legal practice. Of course, had the Report not omitted those other factors, then it would have considered a range of evidence. 

However, the State Department then omitted any specific comment on any of the very criteria it suggested were relevant. There is no evidence in the Report of whether the assurances were provided by a senior official or Minister-level official, what kind of authority this person has, or whether the individual was credible. The US specifically stated that it would take this evidence into consideration. There are some reports that, in the case of Israel, the assurance were signed by Yoav Gallant, the Minister of Defense, though the Report does not confirm this, nor does it engage in any analysis on this point. While it demands the assurances in writing, it does not make them public. Instead, the State Department essentially summarized past practice of recipient states complying with international law. This past practice focused only on credible reports that certain arms were used inappropriately, whether there was any accountability, whether there were any measures for mitigating civilian harm, whether arms reached their intended recipient, and whether humanitarian aid was successfully delivered. This is a serious oversight and makes it almost impossible for Congress, other states, and any third parties to reach independent assessments of the assurances. The fact that this is easily remedied and yet no effort is made for transparency, suggests that the assurances are perhaps not very persuasive.

Analysis of the Evidence in the Report 

Turning to the evidence for the factors that the Report considered, the State Department largely omitted a great deal of relevant evidence. Where it did provide evidence, it was entirely focused on past practice complying with international law. The Report appears to assume that the “likelihood that the partner government will comply with … assurances based on past practice” will be proved only on evidence of past practice and not any other considerations. It clarifies that the criterion of past practice does not concern past practices regarding assurances generally, nor past practice of compliance with international (humanitarian) law generally, but only past practice of misuse of arms provided by the US. This is important because the request for assurances pursuant to NSM-20 could be seen as evidence that a state did not comply with the assurances given prior to NSM-20. 

However, the Report gives no insight into how past practice is evaluated. This is critical because the Report places the most reliance on this evidence. Yet, while the Report clearly acknowledges the instances of violations of international law in the past, it does not explain whether or how these instances contribute to risk or not. Because the Report concludes that the parties will comply with their assurances, we can only presume that the level of past violations is not persuasive of future risk in the view of the State Department. The Report also limits any inquiry into the existence of independent verification or investigation mechanisms by examining past practice in investigating misuse.

In addition, focusing specifically on the Israel section, the Report took a slightly different approach to evidence in this section compared to the other states. It asked the same four questions with similar evidence, but included several introductory pages that framed the context of the conflict. We can speculate that this approach was taken because of the significant public interest in these assurances in particular. None of the sections concerning other states – including Ukraine – contained such introductory remarks. The Report first took note of the context of the conflict, which it dated as beginning on October 7, 2023, a decision not completely free of political choices. It reached several conclusions, such as “Hamas does not follow any portion of and consistently violates IHL” which are most likely correct, though potentially lacking nuance. But they do not submit any evidence for such conclusions in the Report. The Report also took note of the “stated objectives” of the military operation “of destroying Hamas’s military capabilities and dismantling its infrastructure.” It observed, without deciding, that there are conflicting numbers of civilian casualties from the “Hamas-controlled” Gaza Ministry of Health and Government of Israel. However, the State Department omits any mention of how these additional facts and conclusions have any bearing on whether Israel is likely to comply with its assurances in the future. In addition to being largely irrelevant, the fact that this information is even mentioned in the Report suggests that the State Department may have considered irrelevant facts in reaching its assessment on the risk of misuse. 

For all of the reasons in the two parts of this post, the Report is woefully shallow and incomplete. It focused almost exclusively on evidence of past practices violating international law to conclude that the partner government was likely to comply with assurances. It seems difficult to justify this Report as a good faith, due diligence examination of the risk of arms misuse. At some point prior to issuing NSM-20, the US had sufficiently clear evidence to predict that there was a risk of misuse or else it would not have requested the (second round) assurances at all. To confirm or refute this concern, the Report should have examined all foreseeable and relevant factors that would affect that determination. And given the gravity of the decision, and the failure to pursue various foreseeable lines of inquiry, the Report is indefensible. In fact, we can recall that some reporting claimed that the internal State Department cables between the non-political departments reached a drastically different outcome on the assurances than the Report given to Congress does. We know that a State Department employee who was a contributing author on the first draft of the Report resigned and stated that the Report was significantly altered by senior officials and is now incorrect.

In turn, the poor quality of this Report opens the door to further concerns that the US may be in violation of its international obligations, as well as US domestic law. The US is, of course, not party to the Arms Trade Treaty and not bound by the EU Common Position, which both bear on arms export restrictions, so any due diligence considerations arising from those sources would not be directly applicable. Nonetheless, the standards in the ATT and Common Position, as well as standards arising in other contexts, are relevant for assessing assurances insofar as they operationalize existing standards for implementing obligations contained in the Geneva Conventions and other human rights instruments. In addition, even if we exclude those sources, the implementation of the Geneva Conventions requires states to undertake a good faith assessment of risk when they export arms in order to ensure respect for those laws. The Genocide Convention, as interpreted by the International Court of Justice in the Bosnia v Serbia Genocide case, also requires third states to take steps to ensure respect for the Convention. Quite simply, the State Department needs to make a good faith prediction of whether the assurances reduce the risk that the arms will be misused. There are considerable diplomatic statements and judicial decisions providing guidance on how to make such predictions from assurances. In this situation, we can only guess whether the State Department has willfully ignored these international standards and/or various sources of information that would be relevant to making a good faith prediction. By disregarding important predictive factors and not producing a good faith prediction, the US State Department may now have placed the US at risk of incurring state responsibility for failing to ensure respect for the Geneva Conventions and Genocide Convention.