RWA0011 Joseph Crampin

1. I am a Lecturer in International Law with expertise that includes the law of treaties and the relation between international and domestic law. I address questions 1, 2, and 7 of the Call for Submissions. The main object of my submission is to assist the Committee in connection with the international law principles governing agreements of this kind.

Summary of Submission:

2. The main points I make in answer to Qs 1, 2 and 7 are as follows:

•              The effect of para 1.6 is most likely that the MoU is not a binding treaty under international law.

•              The 5 consequences of this as a matter of international law are: (1) no publicity requirements; (2) it is not governed by the law of treaties; (3) breach does not directly give rise to international responsibility; (4) disputes arising directly from the MoU are non-justiciable before international courts; (5) other rules of international law continue to apply unaffected by the MoU.

•              The main consequence as a matter of domestic law is that the constitutional requirements for treaty scrutiny and ratification do not apply.

•              The use of non-binding instruments for this kind of agreement is inappropriate because (1) it is antidemocratic and involves an avoidance of constitutional rules designed to protect Parliamentary Sovereignty; (2) it undermines the institution of treaty-making which is fundamental to upholding the rule of law in international society; (3) non-binding agreements are insufficient to provide proper legal frameworks necessary to govern  transfers of large numbers of individuals.

•              The lack of a period between signature and entry into force is not objectionable per se either as a matter of international or domestic law, but for a significant agreement it is undesirable that Parliament should have no opportunity for scrutiny or debate before the agreement comes into effect.

•              The fact the MoU is non-binding means it does not give rise to legal obligations directly. It may still give rise to legal obligations indirectly through the doctrine of estoppel. It still creates meaningful and sincere commitments which must be complied with in good faith. The commitments may still have legal effects.

•              Individuals may not enforce the agreement directly but they may protect their rights through normal public law/human rights law processes; and compliance with the agreement may be considered in that context, notwithstanding para 2.2.

Q1 What are the implications of signing an agreement that asserts that it is not binding on either Party in international law? Is the MoU an appropriate vehicle for this agreement?

3. An agreement that asserts that it is not binding on either Party in international law is very likely to be not binding under international law. The legal rules governing the binding force of agreements in international law are found in the Vienna Convention on the Law of Treaties 1969 (‘VCLT’). It defines a treaty as: ‘an international agreement concluded between States in written form and governed by international law…whatever its particular designation’.[1] The key consideration is whether the agreement is ‘governed by international law’. This is undefined in the VCLT, but it is generally understood to include consideration of whether the agreement is intended to be binding under international law.[2] One of the simplest ways to establish that an agreement is not intended to be binding is to state so expressly (the use of ‘MoU’ is in itself irrelevant, except as a usual convention typically adopted by the FCDO). An assertion of non-bindingness is as close to conclusive as possible—though it may not be absolutely determinative.
4. The effect of the assertion should be understood to be that the MoU is not a binding treaty under international law. The assertion denies that the MoU is ‘binding under international law’, but this is limited to the question whether the MoU is a binding agreement, ie a treaty. It does not and cannot preclude the possibility that the MoU has other legal effects under international law (or domestic law).
5. For the sake of completeness, the denial of any binding force under international law does not mean the instrument is subject to the law of agreements in some other legal system, but rather remains within the sphere of international law and international relations as a form of ‘gentleman’s agreement’. It is an international agreement that is not intended to be a binding treaty.
6. Under international law, this has five main implications:

(i)     First, the UK-Rwanda MoU does not need to comply with registration requirements under art 102, UN Charter.  They may, therefore, be kept confidential.

(ii)  Second, the MoU is not subject to the law of treaties. This means it is not subject to formality requirements or rules on validity or withdrawal.

(iii) Third, since it is not a treaty, the agreement itself is not a direct source of obligation, and thus breach of the MoU does not directly give rise to international responsibility.[3] If the MoU is breached, no reparations are due as of right, nor may countermeasures be adopted.

(iv) Fourth, neither the UK nor Rwanda can bring an international claim before an international court in respect of the MoU, as confirmed by paragraph 22.

(v)  Fifth, the UK-Rwanda MoU does not modify or depart from the UK or Rwanda’s other international obligations. Nonetheless, non-binding agreements may be important for the practical realisation of or compliance with other rules, and it is possible, in some circumstances, that they may influence the interpretation or application of other rules.

7. As a matter of UK law, the fact the MoU is not a treaty means that the constitutional requirements for Parliamentary scrutiny and approval are inapplicable. Under section 20 of the Constitutional Reform and Governance Act 2010 (‘CRAG’) all treaties must be laid before Parliament prior to ratification, and may be disapproved of by Parliament. It only applies to a ‘treaty’, which is defined, in line with the VCLT definition,[4] as a (1) written agreement (2) between States or between States and International Organizations (3) that is binding under international law. As a result of para 1.6, CRAG does not apply, and there is thus no legal basis to secure Parliamentary scrutiny and approval before the UK pledges its good faith and reputation under a non-binding agreement.
8. The use of a non-binding MoU for this agreement is undesirable for at least three reasons.  Before examining them, however, it should be noted that there are certain well-recognised advantages to non-binding agreements—speed, flexibility and confidentiality—and their use is not necessarily undesirable. To determine the propriety of their use, it is also worth distinguishing between three different kinds of non-binding agreement: (1) political agreements that are aspirational in nature (eg the Atlantic Charter), (2) supplementary agreements that implement or otherwise co-exist with legally binding agreements (e.g. air services agreements, side letters to trade agreements, and decisions of conferences of the parties), and (3) ‘normative agreements’ that involve ‘binding’ commitments that are not intended to be legally binding under international law.[5] This last category is the most difficult since such agreements imitate and may be almost indistinguishable from standalone treaties, yet they avoid the designation and implications associated with actually being a treaty. Nevertheless, they are fairly common and deemed useful for certain kinds of agreements (e.g. defence, commercial, technical or administrative) where the advantages of a non-binding agreement are great and the costs (lack of scrutiny etc) are deemed low. They have also come to be used, though more controversially, for extradition. This MoU is a kind of ‘normative agreement’.
9. In my view, a non-binding agreement is not an appropriate vehicle for this agreement and agreements of this kind. My view is based on three points, the first domestic, the second and third international.

(i)                First, the use of non-binding agreements for significant matters, such as the mass deportation of, what the MoU itself recognises to be, ‘vulnerable’ refugees, is antidemocratic. It enables the avoidance of constitutional control mechanisms—CRAG—that are intended to give effect to the principle of democratic consent and sovereignty of Parliament. Such avoidance is inappropriate, because it undermines the purpose of those constitutional requirements and restricts their effectiveness to a narrower range of international agreements. Indeed, if the use of non-binding agreements continues to be extended into areas that are the traditional preserve of treaties, then the executive will be able to bypass democratic scrutiny whenever it wishes to implement a controversial or unpopular international policy.

(ii)              Second, the use of non-binding agreements for significant matters undermines (a) the integrity of the institution of treaty-making, and (b) the requirements of publicity and transparency that are intended to protect the purposes of the modern international legal order (art 102, UN Charter). These are of systemic importance to the international legal order, which has come under increasing attack in recent years, and measures by the UK that further undermine it ought to be discouraged.

(iii)            Third, non-binding agreements are inappropriate as a tool to govern the transfer of large numbers of individuals. Non-binding agreements are sometimes used for individual cases of extradition—usually to address specific aspects of the individuals’ transfer against a background treaty governing general issues of extradition between two states. In such cases, there are good reasons of administrative convenience and proportionality in using a non-binding agreement to address an identified individual’s situation, and there is a recognised interest in upholding and enabling the institution of extradition in order to protect the rule of law.[6] In this MoU, the aim is to establish an ongoing framework, which provides a rules-based set of arrangements, that govern the human rights of a whole class of persons. The MoU does not simply make arrangements of an administrative character to manage the processing of asylum claims, but instead seeks to alter the allocation of legal responsibilities and jurisdiction over people seeking asylum. The advantages of administrative convenience (speed, confidentiality etc), which apply to other kinds of MoU, should not outweigh the disadvantages of lack of legal stability, minimisation of due process, and avoidance of public scrutiny arising in matters affecting the rights of whole classes of vulnerable people. Here, the rule of law considerations support the use of binding instruments to regulate the process.

10. It would be better, in my view, if a policy were adopted that all significant ‘normative’  (ie not political or supplementary) agreements (perhaps excluding those that need to be confidential for national security reasons or for reasons of commercial sensitivity) are adopted in a legally binding form. For this purpose, ‘significant’ agreements should be distinguished from merely technical or administrative agreements and should be understood to include agreements that involve some or some combination of: (a) state power over individuals, (b) affect individual rights, (c) are ‘public’ ‘normative’ agreements (eg that purport to affect other public authorities, as para 2.2 does here), or (d) cover matters that are not already within the administrative competence of the relevant department (ie would require new powers or funds).

Q2 What are the implications of a significant MoU, such as this, becoming operational on signature? In what circumstances should there be a gap between signature and “entry into force”?

11. There is nothing improper as such, from an international law perspective, in an international agreement, of any kind, entering into force upon signature. The VCLT expressly provides for the possibility of treaties that become binding on signature, and so the fact that the MoU is binding upon signature does not involve any avoidance of requirements under international law.
12. Since, as noted above, there is no constitutional requirement for ratification of a non-binding agreement under CRAG, there is no legal difficulty with a non-binding agreement entering into force upon signature. Nor is there any additional avoidance of legal rules on scrutiny, since such avoidance is already achieved as a result of the MoU being non-binding.
13. It might, nevertheless, be constitutionally improper, or at least undesirable, for an international agreement to be entered into immediately upon signature and without any opportunity for debate where the agreement envisages or may likely require public expenditure or legislation, may affect human rights, and/or where the agreement or policy it implements may result in breaches of other internationally binding commitments of the UK. The question of constitutional propriety may further depend upon whether the agreement falls within existing powers of the relevant minister or requires new legislation and powers for its implementation.
14. If, failing my suggestion above, non-binding agreements continue to be used for arrangements of this sort, it would be highly desirable for practice on entry into force to follow the approach used for treaties. There should be a gap between signature and entry into force to allow for scrutiny and, where necessary, for legislation to be passed before the MoU is signed.

Q7 Does the agreement impose any binding or enforceable obligations on either Party? Given the arrangement asserts it is non-legally binding, and the wording of Article 2.2, what are the consequences if either Party were to breach any of their assurances under the arrangement, and what recourse would be available to those affected?

15. The picture of ‘binding’ obligations is more complicated than para 1.6 implies. First, there may still be legally binding obligations that arise indirectly from the MoU. Second, the pledge of good faith involved in a ‘non-binding’ commitment is solemn and the undertakings made are serious; courts will expect states to comply with them and they are, in effect, little different in terms of their normativity (ie moral/political bindingness) from legally binding obligations.  They are meaningful commitments from which it would be improper to resile. Third, the commitments made in non-binding agreements may be legally effective for certain purposes, such as the grant of a waiver[7] or the giving of assurances that satisfy the requirements of art 3, ECHR.
16. Notwithstanding para 1.6, the MoU can give rise to binding and enforceable obligations under international law—though the mechanism by which it does so is rather limited. Even though it is not binding under international law, the MoU may give rise to estoppels.[8] The fact the MoU is not a binding treaty means that it does not ground rights and obligations directly.[9] In the course of performance, acts of reliance by one party may result in the other party being estopped from resiling from a commitment, which is practically analogous to the commitment being indirectly binding. It is fundamental to understand that the MoU does not cease to be regulated by international law as a result of para 1.6. It is precisely and only because of the rules of international law that para 1.6 is able to produce a non-binding agreement, and such an assertion does not release states into an entirely different, non-legal realm. The MoU remains subject to international law—especially the principles of good faith and reasonableness.[10]
17. The reality, however, is that, although an estoppel can produce legally binding obligations, an estoppel is unlikely to be useful for individuals affected by non-compliance. It is most likely to be helpful in situations where, eg, Rwanda takes its share of individuals but the UK refuses to pay, or where the UK pays and Rwanda refuses to take its share of individuals.
18. In terms of the ability to bring claims, each party would be able to raise objections to breach. Even though these would not be ‘legal’ claims, they would be able to be pursued in a fairly similar way through diplomatic channels. The MoU itself provides for a Joint Committee to monitor implementation and compliance (para 21), and requires parties to pursue negotiations (para 22). Estoppel claims would be legal claims, but without the existence of an international court or tribunal with jurisdiction, these would be pursued through diplomatic channels in much the same way.
19. Individuals may be able to bring complaints to the Monitoring Committee. Individuals would not, however, be able to bring legal proceedings before courts directly to enforce the MoU, because it is not binding (and unincorporated for purposes of UK litigation) and thus does not directly create any rights for individuals. Nor will individuals be able to use estoppel under international law to enforce the MoU indirectly—though similar principles of good faith, reasonableness and consistency might be relevant within other litigation and UK courts will treat the commitments in an MoU as formal and solemn undertakings even if not strictly binding. (I do not know what legal avenues are open to individuals within Rwanda, and offer no comment on that.)
20. Instead, individuals are still able to bring ordinary public law claims (including, possibly, habeas corpus)[11] or human rights claims. To the extent that they are not precluded from doing so by limits to jurisdiction or the Act of State doctrine, they may raise issues of compliance with the MoU in support of their claims. I make no comment on the intricacies or likely success of such claims, merely that these are the sort of claims that individuals will be able to pursue.
21. It is worth distinguishing between issues of non-compliance affecting challenges to deportation by individuals based in the UK and claims brought by individuals already deported to Rwanda. Few issues arise in the former. Individuals would be able to challenge their deportation and to cite instances of non-compliance or other mistreatment as reasons not to give weight to assurances within the MoU. In the latter, however, two issues arise. The first issue is whether the UK has continuing obligations (eg human rights) in respect of individuals based in Rwanda. This may turn on whether the UK’s control, including under para 11 of the MoU, is sufficient to establish jurisdiction.[12] The second issue is the extent to which the UK can be pursued by individuals in Rwanda for mistreatment occurring within Rwanda. It is not clear to me that individual instances of mistreatment would be sufficient to ground a claim against the UK government. Instead, such claims might require the demonstration of a breach of a positive obligation of the UK to protect individuals in Rwanda, including (eg) a failure of the UK to monitor compliance properly.
22. Paragraph 2.2 does not prevent individuals from bringing claims concerning their treatment nor from using the MoU in such proceedings. The provision appears to attempt to achieve two things: (1) to restrict judicial review applications by individuals, e.g. by undermining ‘legitimate expectations’ claims; (2) to prevent domestic or ECtHR judges from evaluating compliance with the MoU in other litigation—for instance, in order to determine whether removal of a person present within the UK to Rwanda is compatible with the individual’s human rights. But, since the agreement is not binding at all and certainly does not bind third parties, it is hard to see how para 2.2 can bind UK courts and impossible to see how it could bind an international court to not examine the MoU. The existence, content, and compliance of the UK/Rwanda with the MoU could be relevant to (eg) a human rights claim, and courts would, where necessary, have regard to the MoU’s assurances and problems of compliance.[13]

              7 of 7