Two recent developments have brought to the fore the issue of involuntary transfer of detainees to countries where they fear severe mistreatment such as torture or death. The first development involved the administration’s decision earlier this month to repatriate two Guantanamo detainees (Djamel Saiid Ali Ameziane and Bensayah Belkecem) to Algeria against their will. The second is that Congress is now poised to provide greater authority for the President to transfer detainees from Guantanamo to foreign countries. The types of concerns raised by the Algerians will likely arise again (and again) in Guantanamo, Parwan and elsewhere. A key question then is whether international law permits the administration to transfer individuals by relying on “diplomatic assurances” from the foreign state that reduce the risk of torture or ill-treatment, and, if so, under what circumstances? In this post, I argue that the administration and its critics both get the answer to that question wrong.

Two UN special rapporteurs —  the Special Rapporteur on Torture (Juan E. Méndez) and the Special Rapporteur on Human Rights and Counter-Terrorism (Ben Emmerson) — criticized the use of diplomatic assurances arising out of the recent Algerian transfer. Their statement suggested that there is a categorical bar against the use of assurances in certain cases: “[D]iplomatic assurances are unreliable and ineffective in protecting against torture and ill-treatment, and States should not resort to them.”

In contrast, the US government’s position, recently articulated in its report to the UN Human Rights Committee, states that:

“The United States will not transfer any person to a country where it determines it is more likely than not that the person will be tortured.

In every Guantanamo transfer case in which security measures or detention are foreseen, the U.S. Government seeks assurance of humane treatment, including treatment in accordance with the international obligations of the destination country, in particular under the Convention Against Torture. In every decision to transfer such a detainee, the U.S. Government takes into account the totality of relevant factors relating to the individual and the government in question, including but not limited to any diplomatic assurances that have been provided.”

[For a more detailed description of US practices involving assurances and post-transfer monitoring, see also the 2013 US Reply to the Human Rights Committee (paras 52-56).]

In this post, I argue that neither the legal position of the two Special Rapporteurs—nor the legal position of the administration—is correct.

Proposition One: The Special Rapporteurs position with regard to the permissibilty of diplomatic assurances is flawed.

In general, many international human rights bodies accept greater use of diplomatic assurances against torture and other inhumane treatment than the Rapporteurs suggest. Indeed, in its most recent review of US practices, the Human Rights Committee articulated a sort of sliding scale approach (my emphasis added):

“The State party should exercise the utmost care in the use of diplomatic assurances, and adopt clear and transparent procedures with adequate judicial mechanisms for review before individuals are deported, as well as effective mechanisms to monitor scrupulously and rigorously the fate of the affected individuals. The State party should further recognize that the more systematic the practice of torture or cruel, inhuman or degrading treatment or punishment, the less likely it will be that a real risk of such treatment can be avoided by such assurances, however stringent any agreed follow-up procedures may be.”

Truth be told, there are a range of approaches that human rights bodies have adopted on this topic. In Philip Alston and my textbook, we divide these approaches into five categories and detail the various authorities for each type. Here are the five types:

Level 1: Categorical and comprehensive approach: prohibits any use of diplomatic assurances to address torture
Level 2: Categorical approach based on individual risk: prohibits use of diplomatic assurances when the person is likely to be subject to torture
Level 3: Categorical approach based on country conditions: prohibits use of diplomatic assurances when torture is practiced systematically in the receiving state
Level 4: Valid but not central factor: diplomatic assurances are a valid factor in considering, and reducing, the risk of torture but should never be a central factor
Level 5: Unqualified, valid factor: diplomatic assurances are a valid factor in considering, and reducing, the risk of torture in all cases

Given the importance of this topic, I have included an excerpt of the textbook that discusses these categories at length.

Proposition Two: The administration’s position on the standard of proof is flawed.

On the other hand, the standard of proof used by the United States is unsustainable as a matter of international law (and principled policy). The US position (admittedly a long-held one) maintains, as I quote above, that the risk of harm to the detainee must be “more likely than not.” Believe it or not, the administration’s official legal position and policy choice is that states can transfer a detainee when there’s a 49% percent likelihood that the individual will be killed or tortured. That standard appears to be largely out of sync with most international legal authorities, which adopt a much lower threshold termed “real risk” (for a good exposition of the standards, see here).

By way of illustration, in the previous review of US practices by the UN Human Rights Committee, one of the Committee’s experts (Walter Kälin) raised the following criticism of the Bush administration:

“He had taken note of the delegation’s statement that there was no rendition to a place where it was ‘more likely than not’ that a person would be tortured. It must therefore unfortunately be inferred that persons could be rendered to a place where the risk of torture was as great as 49 per cent.”

Proposition Three: The administration cannot avail itself of the more permissive standards for diplomatic assurances.

Here’s the rub. At first blush, it might appear that the US position is consistent with the more permissive standards described in Proposition One. Those standards, however, generally include the “real risk” threshold. The Human Rights Committee’s sliding-scale approach, for example, includes “real risk” as an element in its framework (in accord with Kälin’s concerns about the US position). [See also Human Rights Committee, Alzery v. Sweden (Communication No. 1416/2005 (2006) para. 11.3]. Indeed, it is fair to say that the more permissive approaches do not simply include the far lower threshold as the legally acceptable quantum of risk, those approaches, properly conceived, are predicated on it.