Ronan Cormacain: What’s wrong with the Safety of Rwanda Bill?

The Safety of Rwanda (Asylum and Immigration) Bill was introduced into Parliament on 7 December 2023.  According to its long title, its purpose is to facilitate “the removal of certain migrants” to Rwanda.  Without even looking at the international law dimensions, the Bill contains constitutionally troubling provisions.  Firstly, it purports to legislatively state “facts”.  Secondly, it usurps the judicial function by purporting to make a judgement about the facts.  Thirdly, it ousts the jurisdiction of the courts.  And fourthly, it contains non-law-bearing law.

Legislating facts – the Humpty Dumpty clause

Clause 2(1) is probably the key provision in the Bill.  It states that “Every decision maker must conclusively treat the Republic of Rwanda as a safe country”.  When in Liversidge v Anderson it was argued that words in a statute had a particular meaning,  Lord Atkin’s famous dissenting response was: 

I know of only one authority which might justify the suggested method of construction: ‘When I use a word,’ Humpty Dumpty said in rather a scornful tone, ‘it means just what I choose it to mean, neither more nor less.’

Clause 2(1) is this Parliament’s Humpty Dumpty clause.  It provides that Rwanda is a safe country because Parliament says it is safe, as if the simple action of putting an assertion into a law makes it a fact.  But legislation is not meant to set out facts, it is meant to set out law.  Putting an assertion into legislation does not, by some mysterious legal alchemy, turn it into a fact.  Conservative MP Damian Green has already said that he is “worried about legislation by assertion”.  Adam Tucker on this blog describes this as a “legal fiction” which would be “an abuse of Parliament’s legislative authority”.

Legislation could say “Migrants may be sent to a country which is not safe” – this is a legal rule.  Or it could say “Migrants may be sent to (a) a safe country, or (b) Rwanda” – this is another legal rule.  But legislation cannot of itself turn an unsafe country into a safe one, anymore than it can make the earth flat.  Declaring Rwanda to be safe is a legislative attempt to spare the Government’s blushes, by removing migrants to Rwanda whilst at the same time claiming Rwanda to be safe. A more honest approach would be to state that they could be sent to Rwanda, regardless of whether or not it is safe.  This is where politics meets law – using raw parliamentary power, the Government could send migrants to unsafe countries, but only if it expressly states that it is sending migrants to unsafe countries.  What it cannot do is, by legislative force of will, make Rwanda safe.

Usurping the judicial function

The Bill doesn’t only presume to turn assertions into fact by including them in legislative form, it also presumes to make a judgement of the facts.  Clause 1(2)(b) states that “this Act gives effect to the judgement of Parliament that the Republic of Rwanda is a safe country”.

This fundamentally misunderstands the nature of legislation.  Parliament makes law, courts determine facts.  “The judgement of Parliament” is an odd phrase, as it is judges who make judgements.  The phrase does not appear at all on a search of legislation.gov.uk. As Joshua Rozenberg put it “The message seems to be that the judgement of parliament displaces the judgement of the Supreme Court”.  At the risk of stating the obvious, when there are contested facts, then: there is a hearing in a court, with evidence introduced by both sides, arguments made and evidence introduced, adjudicated upon by an independent judge, who then makes a ruling which is a finding of fact.  This is what happened in the original Supreme Court decision, and the Supreme Court ruled, after listening to all the arguments and all the evidence, that Rwanda was not a safe country for the purposes of removing migrants to it.  But in this Bill, the Government substitutes its own judgement for the judgement of the Court.  It might be regarded as efficient if Parliament can make the law and also determine what the facts are, but it negates the entire point of having courts.  A legislative finding of fact is constitutionally unsound.

Ousting the courts

Not content with usurping the judicial function by making a finding of fact, the Bill also ousts the jurisdiction of the courts.  Clause 2 goes into great detail on how courts cannot consider the question of whether or not Rwanda is safe.  Firstly, clause 2(2)(b) states that decision maker includes courts and tribunals, the rest of the clause goes on to say that they:

  • “must conclusively treat” Rwanda as a safe country – clause 2(1)
  • “must not consider a review of, or an appeal against” a decision on these grounds – clause 2(3)
  • “in particular, a court or tribunal must not consider” a series of specific related arguments to the contrary – clause 2(4)

If this wasn’t clear enough, this applies, “notwithstanding”: the Immigration Acts, the Human Rights Act, any other provision or rule of domestic law (including any common law), and international law (all in clause 2(5)).  Furthermore, clause 7(1) defines “decision” to also include “a purported decision” (addressing the Anisminic loophole).  So even if a decision of the Secretary of State is wrong in law, or wrong in fact, the courts are barred from even considering that decision.  

According to the official statistics published by the Ministry of Justice on the First Tier Tribunal Immigration and Asylum Chamber for the year 2022/2023:

Around half (50%) of the 7,700 cases determined at a hearing or on the papers were allowed/granted, although this varied by case type (51% of Asylum/Protection, 52% of Human Rights and 48% of EEA Free Movement appeals were allowed/granted).

The Home Office is clearly not infallible.  Ousting the jurisdiction of the courts means that inevitably wrong decisions will be made, but there will be no venue for correcting them, except in the highly specific circumstances set out in clause 4.

In addition, the given reason for Rwanda being a safe country is the Treaty the Government has just signed with it.  But what if Rwanda breaks this Treaty (even in a “limited and specific way”)?  The Secretary of State said this on the face of this Bill “I am unable to make a statement that, in my view, the provisions of the Safety of Rwanda (Asylum and Immigration) Bill are compatible with the Convention rights”.  Given that the Government admits this Bill may break its treaty obligations, it isn’t inconceivable that Rwanda will break its treaty obligations.  For this reason Mark Elliott claims this Bill reveals “an astounding level of hypocrisy”.   The Supreme Court (paragraph 78 onwards) questioned Rwanda’s practical compliance with its international obligations.  Despite all this, another specific aspect of the ouster of the courts is that courts in the UK are prohibited from considering “any claim or complaint that the Republic of Rwanda will not act in accordance with the Rwanda Treaty”.  So even if the justification for Rwanda being safe is demonstrably untrue, the domestic courts are still not allowed to question it.

Non-law-bearing law

As I previously argued in The Form of Legislation and the Rule of Law (page 11), to legislate is to make law: to create a legal norm, to confer a legal right, to impose a legal duty, or to confer a legal function.  But part of clause 1 of the Bill is what David Feldman has previously described as non-law-bearing law, in that it doesn’t actually do anything legally.  Feldman classifies non-law-bearing law as promissory, declaratory, aspirational or statements of political support.  Clause 1 demands that we recognise an emerging category of non-law-bearing law – legislation as a narrative.

Clause 1 begins with a purpose clause – that the purpose of the Bill is to prevent and deter unlawful migration.  The Renton Committee suggested that purpose clauses be used sparingly, and see the House of Commons Political and Constitutional Reform Committee (1st Report of Session 2013-14 at para 68) for arguments for and against them.  In itself, a purpose clause can be helpful, as it sets out the point of the Bill, making things easier for the reader.  However, clause 1 continues with what can only be described as a narrative – setting out things that have happened.  These are that the UK has signed Rwanda Treaty (clause 1(2)), and the content of that Treaty (clause 1(3)).  This isn’t really law.  This is simply description, narrative masquerading as legislation. 

 Clause 1(4) then recognises that Parliament is sovereign and that the validity of an Act is unaffected by international law.  This falls under Feldman’s “declaratory” category – declaring a thing to be true which is already recognised to be true.  What is the legal effect of “recognising” sovereignty of Parliament in a piece of legislation?  If this provision is repealed, does it mean that Parliament is no longer sovereign, or rather that it is no longer recognised that Parliament is sovereign?  Is it a reminder to the courts, lawyers or the public?  Demanding that Parliament is sovereign rather reminds one of Tywin Lannister’s phrase in Game of Thrones “Any man who must say ‘I am the king’ is no true king”.

Legislation is there to make law, not to remind people of what the law is, or to make aspirational statements.  Every word in legislation should be designed to have legal effect.  The underlying assumption of Bennion on Statutory Interpretationis precision, and this is echoed by the courts view of legislation: “Precision drafting rather than disorganised composition is presumed” (Gillen J at paragraph 47).  Precision means that the exact right word is used, with no prolixity or unnecessary hyperbole.  But narrative and declaratory statements are non-law-bearing and have no legal effect.   To quote former First Parliamentary Counsel Sir Geoffrey Bowman (at page 77) referring to legally inert material in legislation – “excess matter in Bills, as in people, tends to go septic.”  Half of clause 1 isn’t law – it is just writing.

Retrospective legislation?

One final point.  Under clause 9(1), the Bill comes into force on the day on which the Rwanda Treaty enters into force.  Under Article 24 of the Treaty, the Treaty enters into force once all internal procedures have been completed by the UK and Rwanda.  We do not know when that will be.  If the Treaty enters into force before this Bill is enacted, then the Bill will have retrospective effect, that is to say, it will apply before it is made.

Conclusion

To enact this Bill would be to enact something which is constitutionally unsound.  It is not for a Bill to determine what “the facts” are.  For Parliament to make “a judgement” is to ignore the constitutional function of the courts.  The Bill ousts the jurisdiction of the courts almost entirely, even if there is a clear legal or factual error in a decision.  And finally, the Bill contains justifications and narrative statements masquerading as law.  Tucker argued that the courts could interpret this Bill in a strained way in order to call out the legal fictions it contains.  The step prior to this is even more important: Parliament could refuse to enact such a Bill.

Dr Ronan Cormacain, Consultant Legislative Counsel.

(Suggested citation: R. Cormacain, ‘What’s wrong with the Safety of Rwanda Bill?’, U.K. Const. L. Blog (11th December 2023) (available at https://ukconstitutionallaw.org/))