The Safety of Rwanda (Asylum and Immigration) Bill will receive its second reading in the House of Lords on 29 January 2024, having cleared the House of Commons unamended.  There are a great many problems with the Rwanda Bill, any of which might weigh with the Lords, but this blog post focuses on just one: the likelihood that, if enacted, the Bill may well trigger a constitutional crisis between the courts and Parliament. It would be a crisis that is likely to endure beyond the life of the policy embodied in the Bill.  I argue here that one of the roles of the House of Lords is to act as a constitutional safeguard, a steam-valve, and, in exercise of this function under the rare circumstances that attend this Bill, it would be legitimate for the Lords to not only make and insist upon far-reaching changes to the Bill, but even to refuse to pass it altogether.  This post is not concerned with the realpolitik of whether peers would in fact vote the Bill down – though I come to the point in the conclusion. It rather seeks to refute the constitutional argument that it would be illegitimate to block or make potent amendments to it.

Constitutional Propriety and the House of Lords

The House of Lords is, by any standards, an odd institution.  A mixture of hereditary peers, bishops, and appointees, its democratic credentials are weak. I am not averse to a root and branch reform of it.  Nevertheless, at the present time, it remains part of the constitutional order, and has a role to play.  Since the Parliament Acts of 1911 and 1949, its assent is not necessary for a bill to become a law, letting it delay rather than block legislation.  It is mostly a revising chamber, but does disagree with the Commons where such disagreement has an important impact on the shape of the resulting legislation. The chamber is aware of this, and there has been a noteworthy uptick in the number of Government defeats in the chamber over the last years, most notably since reform in 1999. But as Professor Meg Russell observes in her peerless study of the chamber, ‘[a] defeat could mean anything between the (extremely rare) complete rejection of a government bill, to a minor tinkering with the technical detail.’ (The Contemporary House of Lords: Westminster Bicameralism Revived (OUP 2013), p.145).  She also observed that of the 406 defeats she tracked between 1999-2012, some twenty percent entailed a ‘major alteration to a major policy.’  Outright refusals to pass a bill are ‘extremely rare’ because of the generally acknowledged role of the ‘primacy of the Commons’ (see the Joint Committee on Conventions, Report of Session 2005-06 (HL 265-I, HC 1212-I), ch.2). So blocking a bill requires special justification.  

There is an established role for the House of Lords in identifying and safeguarding respect for constitutional norms. This is evident in the heavyweight role of its Constitution Committee, Delegated Powers and Regulatory Reform Committee, European Union Committee, International Agreements Committee and the important role of its joint committees, such as the Joint Committee on Human Rights, on constitutional questions and legislation. Russell’s analysis also identifies that when Labour was in power following the 1997 election, the two most common areas in which Lords defeats occurred concerned civil liberties and constitutional propriety (ch.6).  And during the early years of the Coalition Government commencing in 2010, constitutional propriety remained the key ground for Lords defeats. This constitutional watchdog and safeguarding role is a function that the Labour Party’s Commission on the Future of the Constitution (see pp.140-141) (Chaired by former Prime Minister Gordon Brown), not only recognises but proposes to expand – albeit as a part of plans for a revised and democratically revitalised second chamber.

At the same time, it cannot presently be the case that any bill raising a constitutional question should be fair game for the Lords to fundamentally amend or reject in its entirety.  Too many bills could meet that criterion to be consistent with the general primacy of the Commons.  A higher threshold would be required.  In my contention, the threshold would at least be met where the bill at issue would create potentially radical constitutional change, that change does not carry out a manifesto promise, the change cannot easily be reversed following a change of government, and there is good reason to think the issue can be tested in an upcoming election and the policy reaffirmed after it. All these circumstances are present in respect of the Safety of Rwanda Bill.

A constitutional crisis?

On policy grounds, the Bill has been regarded as a very serious mistake for several reasons, which have been well rehearsed by others.  In my view there is a significant risk that the Bill, if enacted in its current form, could prompt two constitutional shifts, one serious, the other radical. Both of these would be difficult to reverse by repealing the legislation following an election. The first is that the Bill would generate a systematic breach of the UK’s international law obligations.  These include the European Convention on Human Rights and the Refugee Convention – both being pillars of the post-war international legal framework.  Whatever the legal acrobatics arguing to the contrary, the Minister promoting the bill was obviously advised (correctly) by the Attorney General that he is ‘unable’ to issue a section 19 statement under the Human Rights Act 1998 that the Bill is compatible with the provisions of the ECHR.  The UN High Commissioner for Refugees gave decisive evidence in the Supreme Court’s judgment that Rwanda is not safe and cannot easily be made safe for asylum processing purposes, and has recently updated that analysis (published on 15 January 2024) in light of the new Treaty and reached the same conclusion.   

The rule of law is a constitutional principle. And the House of Lords Constitution Committee was unequivocal in its 2019 report on the Internal Market Bill that ‘respect for the rule of law requires respect for international law’ [176]. The requirement that the Government observe its international law obligations is recognised in the ‘codes of the constitution’, e.g. The Cabinet Manual (esp [3.46]); the Ministerial Code (as the Government accepted in a judgment of the Court of Appeal in 2019 ((see [19]-[22])); and the Civil Service Code that imposes a duty to ‘comply with the law.’  Of course, Parliament has the undoubted power to legislate contrary to the United Kingdom’s international obligations.  That is not the relevant point. Parliament, and the King, possess many legal powers whose unrestricted exercise would breach the conventions and principles of the constitution. The constitutionally appropriate way to exercise Parliament’s ultimate power to depart from international law is to act as it did during the Brexit process – withdraw from the relevant international obligations rather than systematically violate them.

So, the systematic violation of international law is a constitutional shift. But is it radical and potentially irreversible?  It is a very open question, to say the least. It is true that the policy on international law itself can be reversed by a successor Parliament. However, once this Government signals to the international community that the UK cannot be trusted to abide by its international treaty commitments, it will lose standing that is not easily regained, at a time when the country is seeking to revive bilateral trade relations and take action at the United Nations, in Ukraine, and in the Middle East to vindicate international law. There is a very real risk that the move will tarnish the nation, rather than just the political party that enacted the measure. 

Provoking the Judicial Nullification of Legislative Intent

The other and more significant constitutional shift is that the Bill has a high likelihood of radically exacerbating tensions between the courts and Parliament, by pushing the courts to adopt or strengthen a legal doctrine which would permit constitutional judicial review of statutes. In considering this claim, we can distinguish meaningfully between a judge (1) disapplying a provision of the legislation, (2) giving a linguistically strained reading to certain ambiguous words, and (3) giving a faithful textual interpretation but which is anyway at odds with parliamentary intentions. These are on a sliding scale for courting constitutional controversy and unseemly political backlash.

Disapplication

The Joint Committee on Human Rights reported (on 11 December 2023) that ‘a claim could be brought before the courts arguing that the Bill, if passed, should not be complied with because it is simply unconstitutional.’ [22]. It found that unlikely – and I agree that it would be unlikely for a finding to be so broad.  But neither is it entirely remote. The most likely candidate for disapplication would be part or all of clause 2, which requires ‘every decision-maker’ to ‘conclusively treat….Rwanda as a safe country,’ contrary to the comprehensive finding by the Supreme Court in R (AAA) v Secretary of State of the Home Department [2023] UKSC 42 that it is not.  Clause 4(2) is another potential candidate, which I examine further below. Both clauses 2 and 4(2) essentially force judges to disregard the core dangers confirmed in the judgment of the Supreme Court.

The former Attorney General Sir Geoffrey Cox MP wrote to the Daily Telegraph arguing that there was a real risk of the courts doing the unthinkable with the Bill and that ‘if the Supreme Court were to quash or disapply an Act of Parliament on domestic constitutional grounds for the first time, it will be impossible to put the constitutional genie back in the bottle.’  Professor Mark Elliott’s reply to this letter, and the indignant response to it by David Frost, observed that courts would and should ‘think long and hard before rejecting any Act of Parliament.’  I agree with that. But he added further that ‘it does not follow that we can be confident that a court would — or should — never do so.’  Other scholars, like Adam Tucker, explain very convincingly why a scheme such as the legal fiction enacted by clause 2 of the Bill ‘may yet provide the circumstances’ in which the courts take the unprecedented step of disapplying a provision of an Act of Parliament based on the theory that it offends the common law principle of the rule of law. 

There is dicta in several cases on this point but the leading and very recent authority is the case of R (Privacy International) v Investigatory Powers Tribunal [2019] UKSC 22. Lord Carnwath, with the agreement of Lady Hale and Lord Kerr, found that:

…although it is not necessary to decide the point, I see a strong case for holding that, consistently with the rule of law, binding effect cannot be given to a clause which purports wholly to exclude the supervisory jurisdiction of the High Court to review a decision of an inferior court or tribunal, whether for excess or abuse of jurisdiction, or error of law. In all cases, regardless of the words used, it should remain ultimately a matter for the court to determine the extent to which such a clause should be upheld, having regard to its purpose and statutory context, and the nature and importance of the legal issue in question; and to determine the level of scrutiny required by the rule of law.

[144]

Of seven judges, three subscribed to this dictum, and a fourth (Lord Lloyd-Jones) reserved judgment on the point (at [167]-[168]).  It was very close. Further, a key plank of the dissenting judgment of Lord Sumption focused on how ouster clauses tend to shield the decisions of other judicial or quasi-judicial bodies, not of executive officials. Lord Sumption held (Lord Reed concurring) at [182] that ‘[a] right of access to a court or similar judicial body to review the lawfulness of administrative or executive acts is an essential part of the rule of law. But the rule of law does not require a right of appeal from such a body or a right to call for a review of its decisions’ [emphasis added].  The holdings or dicta that would uphold or have upheld ouster clauses – and these remain in a minority of the relevant leading cases – have almost invariably concerned provisions that protect quasi-judicial determinations by tribunals. They can be distinguished on that basis from a case considering clause 2, whatever else such judgments had to say about the importance of the legislative supremacy of Parliament.

Now, it will and has been argued that the deeming provision in clause 2 is not strictly an ouster clause, because the provision tells immigration officers and courts alike what conclusion they have to come to, rather than insulates their legal determinations from judicial review. But as Tucker rightly observes, ‘the common law hostility to ouster clauses is not grounded in their form. It is grounded in their effects and the way they undermine the rule of law by purporting to exclude the executive from compliance with the law.’  The fact that this Bill reverses a recent Supreme Court judgment on the facts and enacts a legal fiction whose purpose is to exclude judicial review of factual questions makes it a notoriously plainer example of an affront to the rule of law and the separation of powers.  If the comparatively minor rule of law issue in Privacy International very nearly carried a majority of the Supreme Court in 2019, there is every reason to think the scheme in the Safety of Rwanda Bill poses a very significant risk of judicial disapplication, or another high-level finding that they would disapply it if they found no other way to read the provision down.

Reading Down

The commentary authored by Lord Sandhurst KC and Harry Gillow and published by the Society of Conservative Lawyers, argues that reading down is less offensive than a complete disapplication of the statute (pp.5-6). They argue, rightly, that a total exclusion of any review made more serious the risk that the courts would disapply the legislation.  That is why they prefer keeping the Bill as it is rather than adopting the proposed amendments pushed by a substantial minority of the Conservative Party in the Commons (see amendments 10-25 on the HC Amendment Paper of 15 January 2024 and HC Debates, Vol. 743, Col.703ff), which proposed extensive changes across the Bill to virtually close off any residual opportunity to legally challenge a removal decision. But as Sandhurst and Gillow candidly observe, ‘the two approaches blend substantially into one’ because it is ‘perfectly plausible that [the courts] will adopt an interpretation of that legislation entirely at odds with any conventional reading of the wording.’ (p.6).  Quite so, and that is why this option is, although more probable, bound to cause the same or quite similar kinds of constitutional tensions, just as the approach taken in the Privacy International case did.

The common law principle of legality is invoked routinely to justify at times quite strained linguistic readings of a statutes that offend the rule of law, human rights or similar constitutional principle (see e.g. R (Evans) v Attorney General [2015] UKSC 21; R (Privacy International) v Investigatory Powers Tribunal [2019] UKSC 22). The main avenues for reading down the Safety of Rwanda Act, would be found in clause 4.  Clause 4(1)(a) provides that clause 2 does not have the effect of preventing decisions by immigration officers that the Republic of Rwanda is not a safe country for the person in question if it is ‘based on compelling evidence relating specifically to the person’s particular individual circumstances (rather than on the grounds that …Rwanda is not a safe country in general).’  The clause unavoidably requires interpretation. Is systematic discrimination against gay and transgender persons evidence that relates to a person’s ‘particular individual circumstances’ as a gay person?  Or should the provision require that the state of Rwanda has a view on this particular person’s situation? It cannot be the latter, because in nearly every case Rwanda would have no detailed knowledge of the person being removed from the UK. Such an interpretation would render clause 4 nearly meaningless.  So if evidence supporting a real risk to persons in that class can be found, then representatives of claimants for whom sexual orientation forms part of their claim may have a plausible case for not being removed to Rwanda. Similar arguments could be made for victims of modern slavery or human trafficking who face a real risk of being re-trafficked if they are removed to Rwanda. Good lawyers would look for, and in many cases find, similar lines of argument and expand the categories to which it applies.

However, clause 4(2) provides that any determination under 4(1) cannot relate to the issue of whether Rwanda would remove the person to another state (e.g. refoulement), in contravention of any of its international obligations. So on its face one cannot argue that the asylum processing will be corrupted on account of discrimination based on sexual orientation and lead to refoulement.  That means that clause 4(2) specifically seeks to exclude any judicial consideration of exactly the main issue in the Supreme Court case.  It ensures that the Bill is still a palpable rejection of the legal ruling, based on the fiction – irrespective of its implementation in practice – that the subsequent treaty remedies the issue.  Most of the ingredients that would provoke strong judicial pushback are therefore still there.  If clause 4(2) stays in the Bill, there will be incredibly strong pressure to read it down for exactly the same reasons that there would be pressure to disapply the legislation if clause 4 stood apart from the Bill.

Similarly, clause 4(4) provides that a court can only issue an interim remedy (i.e. block removal to Rwanda pending resolution of the asylum claim) where ‘the person would, before the review or appeal is determined, face a real, imminent and foreseeable risk of serious and irreversible harm if removed to the Republic of Rwanda.’  The Government no doubt believes that imminent risk means evidence of a very proximate and unusual risk.  But it is likely that the requirement of ‘imminent risk’ could be given a soft reading, just as it is in the caselaw of the European Court of Human Rights in its interim measures jurisprudence (see e.g. Case of Paladi v. Moldova (Application no. 39806/05) (E.Ct.H.R., 10 March 2009, Grand Chamber).  How else would this provision have any meaning in a system where virtually no asylum applicants will have any connection to Rwanda (unlike in the classic refoulement scenarios at issue in the jurisprudence of the Strasbourg Court)?

Knowledge of these weaknesses is in part the cause of the backbench rebellion in the House of Commons and the basis of proposed amendments argued for by the Policy Exchange (at pp.20-22).  But they are on the horns of a dilemma – if the legislation excludes clause 4 the risk of disapplication is much higher, and if it includes it the risk of reading down will be quite strong.  Both options can normalise the judicial practice of candidly stopping the policy the Government sought to achieve in the legislature that passed the bill.

Conventional interpretation, where legislative intent is disregarded

A final interpretive avenue is for judges to apply the statute in its terms by making sense of the contradictory impulses embodied by the schemes in clauses 2 and 4.  It may be true that Parliament – or the Government – wishes clause 4 to be read very narrowly, but the courts must give the text a meaning that makes sense. Clause 4(1) explicitly qualifies clause 2; and clause 4(4) uses language applied liberally by the European Court of Human Rights.  These provisions must be read in a way that gives them some effect rather than no real effect.  Any assurances given by Ministers during passage of the Bill – and any external aids to construction of statutes, such as Explanatory Notes and the like – cannot alter the primacy of the text and the duty of the court to make coherent sense of it (R (on the application of O) v Secretary of State for the Home Department [2022] UKSC 3).  Neither, I suspect, could the unusual and juristically uncertain ‘introduction’ clause 1, which sets out the purpose of the statute.  So if lawyers offer the courts theories that make better textual sense of clause 4 than the theory that motivates the Government, then judges might adopt them, following conventional standards of statutory interpretation rather than any high-level references to the rule of law and principle of legality.  Yet even this unstrained and conventional form of statutory interpretation may put the courts into the position of implementing a legal meaning that the Government (and presumably Parliament) intends for it not to have.

The knock-on consequences

It seems unlikely that a court will straightforwardly disapply the legislation, but there is significant risk of it. There is an even higher risk of the legislation being read down, and clauses 4(1),(2) and (4) in particular.  And there is also a high risk that the litigation surrounding these provisions will provoke further judicial observations, whether binding or not, that judges would be prepared to disapply the statute if they did not have another way out in the judgment they are giving (just as Lord Carnwath did in his leading judgment in Privacy International).  The result is that the passage of the Bill in its present form stands a high chance of fortifying the theory that judges may have the common law power to disapply statutes, directly or indirectly.  These changes will have lasting effects on constitutional affairs in the UK. They cannot easily be legislatively reversed.  When these changes occur, and regardless of which of the interpretive strategies above is chosen, it will lead to further political discontent and backlash against the courts and lawyers. Pressure groups, some journalists, and political pundits will begin to argue publicly that if they cannot give effect to their policies by changing the law (half-heartedly in this case), they will need to change the judges.

Conclusion

All things considered, there is a very real and substantial risk that passage of the Bill in its current form could lead to a constitutional change and crisis.  The Lords would be acting within its constitutional role, in my view, if it passed far-reaching amendments or even refused to pass it and let the Conservative party make it a manifesto pledge in the upcoming election.  Of course, the fact that it would be constitutionally legitimate for peers to fundamentally amend or vote down the Bill down does not mean it will happen.  Success would require the support of the Labour peers.  That is doubtful, for two reasons. First, and as reported by the Financial Times, the Party would fear their own legislative programmes being caught up in the Lords on similar grounds.  Second, the result of the Lords not passing the Bill would be that the Rwanda scheme becomes an election issue and talking point during debates. It could become an electoral liability. So why do it when they could quietly repeal the legislation after taking power?

The point of this post is not to persuade the Lords to radically amend or vote the Bill down, but to refute the suggestion that it is inappropriate for them to do so.  The fact that it is appropriate can be taken as support for both refusing to pass the Bill, as well as for supporting amendments that plainly cut against the Government’s core purposes for it.

The author would like to thank Murray Hunt, Meg Russell, Gavin Phillipson, Tom Hickman, Mike Gordon, Se-shauna Wheatle, Jan van Zyl Smit and above all Nick Barber for very helpful discussions.

Jeff King is a Professor of Law at the Faculty of Laws, University College London, on secondment as Director of Research at the Bingham Centre for the Rule of Law. He was a Legal Adviser to the House of Lords Constitution Committee between 2019-21.

(Suggested citation: J. King, ‘The House of Lords, Constitutional Propriety, and the Safety of Rwanda Bill’, U.K. Const. L. Blog (26th January 2024) (available at https://ukconstitutionallaw.org/))