The British constitution is famously unentrenched: constitutional laws are not intrinsically more difficult to override than ordinary laws. However, in the 2002 case of Thoburn v Sunderland City Council,Footnote 1 the Administrative Court suggested that constitutional statutes are more difficult to repeal than ordinary statutes. The Court said that constitutional statutes are susceptible to implied repeal in a narrower range of circumstances than ordinary statutes. Initially there was intense academic interest in Thoburn. As time went on, however, and no higher court gave its approval, Thoburn began to seem like an outlier, not a forerunner.
That is what makes the largely overlooked 2012 case of H v Lord Advocate Footnote 2 important. In H, the Supreme Court repeatedly said that the Scotland Act 1998 cannot be impliedly repealed, under any circumstances, due to its “fundamental constitutional” status.Footnote 3 While these remarks were obiter dicta, they suggest the path the law will take. Unless judicial thinking changes, courts in the future are likely to treat constitutional statutes, including the Scotland Act, as capable only of express repeal. That would make constitutional statutes “quasi-entrenched”, to coin a term, with potentially significant consequences for Parliament's powers and the role of courts.
In addition to showing that H deserves more attention from constitutional scholars than it has received thus far, our aim in this article is to demonstrate that, as a judicial innovation, the quasi-entrenchment of statutes lacks a sound legal basis. We shall argue that Parliament is capable of making its intention to repeal a constitutional statute clear without making it express, and that judges cannot, on their own initiative, lawfully ignore Parliament's clear decision to repeal even a constitutional statute.
Our argument is relevant so long as there are constitutional statutes, however defined, so we shall not consider what makes a statute “constitutional”.Footnote 4 We shall rather assume, consistently with Thoburn, that a constitutional statute is one that conditions our relationship as citizens with the state, or alters the scope of basic rights.Footnote 5 This definition includes the Scotland Act, as well as the Human Rights Act 1998, the European Communities Act 1972, the Union with Scotland Act 1706, the Bill of Rights 1689, and Magna Carta 1215.
I. The Path to H
A.V. Dicey wrote in 1885 that “fundamental or so-called constitutional laws are under our constitution changed by the same body and in the same manner as other laws”.Footnote 6 Perhaps nothing Dicey said about the constitution has entirely escaped criticism, but at least until recently it was generally accepted that constitutional laws are not per se more difficult to change than ordinary laws.Footnote 7 In 1998, for example, Eric Barendt was able to say with confidence that “fundamental laws … can be as easily repealed as, say, the Animals Act 1971 or the Estate Agents Act 1979”.Footnote 8
The one exception, according to Barendt, was the European Communities Act 1972 (hereafter “ECA”). By the combined operation of s. 2(1) and s. 2(4) of the ECA, subsequent Acts of Parliament only “have effect” subject to directly enforceable European Community (now European Union) law. The operation of these sections was, of course, the issue in the Factortame litigation.Footnote 9 Simplifying greatly, the Merchant Shipping Act 1988 imposed nationality-based restrictions on the registration of fishing vessels. These restrictions were inconsistent with directly enforceable EU law, and hence inconsistent with the ECA. The doctrine of implied repeal says that a later statute brings about the repeal of an earlier statute to the extent of their inconsistency, provided that the later statute is not more “general” than the earlier statute.Footnote 10 According to that doctrine, the Merchant Shipping Act, as the later statute, should have taken priority over the ECA. Instead, in Factortame (No. 2), the House of Lords “disapplied” the Merchant Shipping Act.Footnote 11 Contrary to the doctrine of implied repeal, “the Merchant Shipping Act … yielded to the superior force of an earlier statute”.Footnote 12 The justification for and exact significance of Factortame are contested, but it is now generally recognised that the ECA cannot be repealed except by express wordsFootnote 13 (or, some would add, by necessary implicationFootnote 14).
In the 2002 case of Thoburn, the Administrative Court sought to justify similar protection from repeal for a much larger class of statutes. As with Factortame, the facts in Thoburn are so well documented that they do not need to be described here.Footnote 15 In essence, the issue was whether the Weights and Measurements Act 1985 had impliedly repealed s. 2(2) of the ECA (which deals with subordinate legislation). Laws L.J., with whom Crane J. agreed, held that the statutes were consistent.Footnote 16 Thus, no issue of implied repeal arose. In case he was wrong on that point, Laws L.J. considered whether the ECA could be impliedly repealed; in light of Factortame, he held it could not.Footnote 17
Laws L.J. then went a step further. It is not only the ECA that is protected from implied repeal, he said; every constitutional statute is, to some degree, protected from implied repeal. At first, Laws L.J. put the point categorically: “Ordinary statutes may be impliedly repealed. Constitutional statutes may not.”Footnote 18 Some commentators, perhaps reading this statement in isolation, took it to reflect Laws L.J.'s considered view.Footnote 19 But he immediately explained that, under some conditions, even constitutional statutes can be repealed by implication:
For the repeal of a constitutional Act … the court would apply this test: is it shown that the legislature's actual – not imputed, constructive, or presumed – intention was to effect the repeal or abrogation? I think the test could only be met by express words in the later statute, or by words so specific that the inference of an actual determination to effect the result contended for was irresistible. The ordinary rule of implied repeal does not satisfy this test.Footnote 20
Laws L.J. also said that a constitutional statute can be repealed by “unambiguous words”, which may not be express.Footnote 21 So, according to Laws L.J., the test of whether a constitutional statute is repealed is whether there are express words or words that are “unambiguous” or so “specific” that the inference of an intention to repeal is “irresistible”.Footnote 22 The second branch of this test imposes a more exacting standard than the traditional doctrine of implied repeal, which requires a “mere”Footnote 23 implication. Thus, Laws L.J. concluded, the traditional doctrine has “no application to constitutional statutes”.Footnote 24
In addition to Factortame, Laws L.J. argued that the “principle of legality” favoured giving a special status to constitutional statutes.Footnote 25 According to the principle of legality, Parliament is presumed not to legislate contrary to common law constitutional rights and principles.Footnote 26 There was at one time a debate about whether this principle could be overridden impliedly as well as expressly, but that debate has been settled.Footnote 27 In R v Secretary of State for the Home Department, ex p Simms, which Laws L.J. cited in Thoburn, Lord Hoffman said that the principle of legality could be overridden by “express words or necessary implication”.Footnote 28 In Mohammed Jabar Ahmed v Her Majesty's Treasury,Footnote 29 the most recent Supreme Court case on the principle of legality, four out of the five Justices to write an opinion held that common law constitutional rights could be overridden by necessary implication.Footnote 30 A “necessary implication” in this context is an implication that is especially obvious – or what could be described, in Laws L.J.'s terminology, as an “irresistible” or “unambiguous” implication.Footnote 31 Unsurprisingly, then, Laws L.J. claimed (extrajudicially) that Thoburn does “for statutory constitutional guarantees what the law already does for common law constitutional guarantees”.Footnote 32
After Thoburn was decided, there was a surge of academic interest in implied repeal and constitutional statutes, in part due to the perceived novelty of Laws L.J.'s remarks.Footnote 33 Ultimately, though, Thoburn was a decision of the Administrative Court and leave to appeal had been denied. Laws L.J.'s remarks were obiter. The cases that cited Thoburn relied on it to show that there are constitutional statutes or that EU law is supreme, or for guidance as to the interpretation of constitutional statutes.Footnote 34 When the Supreme Court cited Thoburn in its judgment in Watkins v Home Office, it was in relation to the idea of a “constitutional right”.Footnote 35 What no case did was rely on, or develop, Laws L.J.'s remarks on the implied repeal of constitutional statutes, leaving the status of his remarks uncertain.Footnote 36
That brings us to H, a decision of the Supreme Court, and the first clear judicial statement about the implied repeal of constitutional statutes since Thoburn. H has not yet been discussed in the constitutional context, so we shall describe it in some detail here.Footnote 37 The proceedings began when the US sought the extradition of a husband and wife, H and BH, on charges relating to the importation of chemicals normally used in the manufacture of methamphetamine. H and BH, who were resident in Scotland and had children there, argued before Sheriff McColl that their extradition would be incompatible with their rights under the European Convention on Human Rights within the meaning of the Human Rights Act 1998, specifically their Article 8 right to respect for a family life. The Sheriff rejected this argument and, in accordance with the Extradition Act 2003, sent the case to the Scottish Ministers to determine whether H and BH ought to be extradited. The Ministers ordered their extradition. H and BH appealed to the High Court of Justiciary, where their appeal was dismissed.
H and BH then sought to appeal to the Supreme Court. At that point, the issue with which we are concerned arose. There were two sets of provisions relevant to the competency of the appeal. Under the Extradition Act, a decision of the Scottish Ministers made under that Act can only be appealed against under that Act, and the Extradition Act does not provide a right of appeal to the Supreme Court from the High Court of Justiciary. So, on first sight, the Extradition Act prevented H and BH from appealing to the Supreme Court. However, the Scotland Act does provide a right of appeal to the Supreme Court from the High Court on a “devolution issue”.Footnote 38 Section 57(2) of the Scotland Act prohibits the Scottish Ministers from acting inconsistently with any of the Convention rights, and whether the Ministers have violated s. 57(2) is a devolution issue. There was therefore at least the possibility of a conflict between the Extradition Act and the Scotland Act. Under the doctrine of implied repeal, the Extradition Act, as the later statute, would take priority. Although none of the parties contended that the Supreme Court lacked jurisdiction to hear the case, the Court decided to consider the issue due to its “general public importance”, and did so with the assistance of written submissions from the counsel for the Scottish Ministers.Footnote 39
Lord Hope, with whom the other judges agreed on the issue of competency, concluded that the Court had jurisdiction to hear the appeal because, properly interpreted, the provisions of the Extradition Act and the Scotland Act were consistent. The Extradition Act only prevented appeals from decisions of the High Court of Justiciary that were based on the Extradition Act. The system of appeal under the Scotland Act, meanwhile, “lies outside the contemplation of the sections of the … [Extradition] Act”.Footnote 40 It provides a “parallel remedy”.Footnote 41 Such is the ratio of H on this issue.
What interests us here is the obiter dictum. The crucial passage comes when Lord Hope comments on what would have happened had the Extradition Act conflicted with the Scotland Act. He says:
It would perhaps have been open to Parliament [in the Extradition Act] to override the provisions of s 57(2) so as to confer on … [the Scottish Ministers] more ample powers than that section would permit in the exercise of their functions under the … [Extradition] Act. But in my opinion only an express provision to that effect could be held to lead to such a result. This is because of the fundamental constitutional nature of the settlement that was achieved by the Scotland Act. This in itself must be held to render it incapable of being altered otherwise than by an express enactment. Its provisions cannot be regarded as vulnerable to alteration by implication from some other enactment in which an intention to alter the Scotland Act is not set forth expressly on the face of the statute.Footnote 42
Lord Hope adds that the provisions of the Extradition Act cannot be understood to preclude resort to the appeal procedure in the Scotland Act “because they do not exclude resort to it expressly”.Footnote 43 It is difficult to think how Lord Hope could have been clearer: the Scotland Act can only be expressly repealed; it cannot be impliedly repealed; that is because of its “fundamental constitutional nature”. Unlike Laws L.J. in Thoburn, Lord Hope in H never qualifies these claims or suggests that there are conditions under which the Scotland Act can be impliedly repealed.
Ultimately, the Court in H went on to dismiss the appeals, and to uphold the extradition order against H and BH.
The dictum in H is significant for several reasons. First, whereas Thoburn was a decision of the Administrative Court, H is a Supreme Court decision and, on the issue of competency, it was unanimous.Footnote 44 Second, whereas Thoburn said that a constitutional statute can be impliedly repealed by a particularly clear implication, and the principle of legality says that a common law constitutional right can be overridden by necessary implication, H says that the Scotland Act cannot be impliedly repealed – no exceptions.
There is a third reason why H matters. Lord Hope says that the Scotland Act cannot be impliedly repealed due to its “fundamental constitutional nature”. The Scotland Act is undeniably important, but it is not more fundamental or constitutional in nature than, for example, the ECA, the Human Rights Act, or the Government of Wales Act 2006.Footnote 45 Lord Hope's reasoning therefore seems to support a general exemption for constitutional and fundamental statutes from implied repeal. H has the likely scope of Thoburn but, as we have indicated, greater stringency. That makes H not just significant, but quite radical.Footnote 46
Why, if H is so important, has it not been discussed before? Of course it is a recent decision, but two other factors suggest themselves. One is that the implied repeal of the Scotland Act was a preliminary issue in H. The main issue was one of extradition law. Also, although Thoburn was extensively cited by counsel for the Scottish Ministers, it was not cited in the judgment itself.
Whatever the reason, H raises the strong possibility (or the real danger, we shall argue) that courts in the future will take a new approach to constitutional statutes. Unless there is a change in judicial thinking, courts will not treat constitutional statutes as exempt from express repeal, but they will treat them as exempt from implied repeal. Constitutional statutes will thus not be fully entrenched, but they will be quasi-entrenched.
II. Legal Reasons
Significant changes to the law need strong reasons in their support. Unfortunately, Lord Hope did not identify a legal reason for why the Scotland Act cannot be impliedly repealed. But there are only two possibilities. In this section, we shall explain what they are, starting with a brief discussion about statutory meaning and Parliament's powers.
A. Meaning and Power
The meaning of a statute is, roughly, the meaning it is reasonable to infer that Parliament intended the statute to have.Footnote 47 What it is reasonable to infer as to Parliament's intention will depend on the available evidence, including the words that Parliament chose, along with relevant rules of syntax and semantics, the statutory context, the reasons for which the statute was enacted, and so on.Footnote 48 Courts piece together this evidence using interpretive presumptions, like the principle of legality. Interpretive presumptions are guidelines or rules of thumb as to Parliament's intentions, and by nature they can be rebutted by sufficient contrary evidence.
The meaning of a statute may be express or implied. An express meaning is one that is spelt out on the face of the statute. A statute expressly says that another statute is repealed when it names or describes that other statute and says it is “repealed” or of no “force or effect” or it uses words understood to be synonymous.Footnote 49 We shall have much to say about implied meaning in the next section. For now, it is enough to say that an implied meaning of a statute is one that is not express but that is nonetheless reasonable to infer that Parliament intended. In the typical case, a statute implies that another statute is repealed when it is later than, and inconsistent with, that other statute. (Courts seem to assume that this typical case is the only case in which a repeal is implied – an opinion we do not share, and one we shall criticise shortly.) This claim about implied meaning is one half of the traditional doctrine of implied repeal (the other half is set out below).
Parliament clearly has the power to expressly repeal any statute. That is to say, Parliament's use of express words to communicate its intention to repeal a statute brings about the change in the law it intended. Until Factortame, most commentators believed that Parliament also had the power to impliedly repeal any statute. That idea formed the other half of the traditional doctrine of implied repeal. After Factortame, many believed that Parliament no longer had the power to impliedly repeal one statute, namely, the ECA.Footnote 50
B. Two Reasons
It seems obvious that the Scotland Act can be impliedly repealed if two conditions are satisfied: it is possible for there to be a statute the implied meaning of which is that the Scotland Act is repealed, and such a statute would in fact bring about the repeal of that Act. So, conversely, if the Scotland Act cannot be impliedly repealed, it must be for one of two reasons. Either:
(1) Parliament cannot convey its intention to repeal the Scotland Act by implication (i.e. there cannot be a statute the implied meaning of which is that the Scotland Act is repealed); or
(2) Parliament may be able to convey its intention to repeal the Scotland Act by implication but, even were it to do so, it would not thereby bring about the repeal of that Act.
Here is the argument for (1) in outline. Courts rely on presumptions to guide their inferences as to Parliament's intentions. The more unlikely Parliament is to have intended some change in the law, the stronger the presumption it did not have that intention. Also, meanings which are express are generally easier to identify and understand than ones which are implied. Putting these points together, one might think there are some meanings that are so unlikely to be Parliament's intended meaning that, absent express words, it could never be reasonable to suppose that is what Parliament actually intended. Such a meaning could only be express; it could never be implied. One might then think that the repeal of the Scotland Act, or another constitutional statute, is a sufficiently unlikely thing for Parliament to intend that it cannot be a matter of implication. In short, the Scotland Act cannot be repealed by implication because such an implication is impossible.
The alternative is to argue for (2). If (2) were true, then, to bring about the repeal of the Scotland Act, it would not be enough for Parliament to convey its intention to bring about that Act's repeal. It would have to convey its intention in a particular form, i.e. through express words. In that case, it would be possible for Parliament to intend to repeal the Scotland Act, to communicate that intention, and yet fail to bring about the repeal of that Act because it did not use the “right” form of words. Parliament's will, in that case, would be frustrated. That Parliament's powers are so limited is not a conceptual impossibility. Some scholars think that Parliament limited its power to repeal the ECA in a similar way.Footnote 51 But each new limit on Parliament's powers must have a legal basis. The starting point of any argument for (2) must, therefore, be an explanation of how Parliament's power to repeal the Scotland Act came to be limited.
In short, defending (1) is a matter of showing that Parliament cannot be understood to intend the repeal of the Scotland Act absent express words. Defending (2) is a matter of showing that Parliament lacks the power to repeal the Scotland Act absent express words.
We shall consider these alternatives in the next two sections. Rejecting both, we shall conclude that neither the Scotland Act nor constitutional statutes generally are quasi-entrenched. To be clear, our focus at this point is on the potential legal reasons for quasi-entrenchment. We are deliberately setting aside broader normative considerations until section V.
III. Implications, Meaning, and Consistency
Is it true that Parliament cannot convey its intention to repeal the Scotland Act by implication? That is to say, is Parliament unable to make its intention to repeal a constitutional statute clear without making it express?
A. Repeals from Defective Expressions
Let us start by distinguishing two kinds of implications.Footnote 52Deliberate implications arise when there is a gap between what someone – a “speaker” – says expressly, and what he or she obviously intends, such that it is reasonable to infer that the speaker intends his or her audience to “read between the lines”. For example, if, when asked whether someone is a fast runner, you reply “Perhaps Usain Bolt could beat him”, you imply without saying expressly that the runner in question is indeed fast. Deliberate implications are common in ordinary life, but rare in law.Footnote 53
More common in law, and more important for present purposes, are implications from defective expressions. These implications are related to deliberate implications, except that here the gap between what is express and what is obviously intended makes it reasonable to infer a mistake on the speaker's part, or an incompleteness in what he or she has said. For example, if at a pub you order a “pint of crisps and a packet of lager”, the bartender will know that you “really meant” a pint of lager and a packet of crisps. That is not what you expressly said, but it is implied in what you said, because it is obvious it is what you intended.
In the statutory context, an implication from a defective expression typically arises when Parliament's intended meaning is obvious but it has not been made express due to a drafting error. Courts in such cases will give effect to the statute's implied meaning – which is also its intended meaning – rather than to its defective, express meaning.Footnote 54 This general point holds true when Parliament intends to repeal one statute or part of a statute but, as a result of a drafting error, expressly says that it repeals another. Here is a real example. The Repeal Schedule of the Interpretation Act 1978 says that part of “paragraph 48” of “Schedule 5” of the Medical Act 1978 is repealed. In fact, there is no paragraph 48 in Schedule 5; that paragraph is in Schedule 6. There is no doubt as to Parliament's intended meaning. There is also no doubt that courts would read the reference in the Interpretation Act in its “intended form”, as Francis Bennion says in his discussion of this example.Footnote 55
The kind of mistake that occurred in the Interpretation Act and that occurs in other statutesFootnote 56 could occur in a statute designed to repeal a part of the Scotland Act. Parliament could attempt to expressly repeal a part of the Scotland Act, but fail to do so because of a drafting error. It might intend to repeal s. 84(5), for instance, but due to a typo say that s. 85(4) is repealed. Parliament's intended meaning could still be clear, taking into account all the available evidence. In that case, Parliament's intention to repeal part of the Scotland Act would be conveyed by implication.
B. Inconsistency and Implicit Assumptions
So far, we have been taking Lord Hope's remarks at face value and assuming he meant to rule out any kind of implied repeal of the Scotland Act. Perhaps, though, he meant merely that the Scotland Act cannot be repealed by operation of the doctrine of implied repeal. If that is true, and it is true because of the meaning of statutes, then either there cannot be a later statute that is inconsistent with the Scotland Act, or even a later and inconsistent statute does not imply that the Scotland Act is repealed.
Starting with the first possibility, there is a well-established presumption that Parliament intends to legislate consistently with the existing law.Footnote 57 That presumption favours interpreting a later statute in a way that is consistent with an earlier statute. It is a strong presumption, and it is even stronger when the earlier statute is important.Footnote 58 Plausibly, because constitutional statutes are very important, there is a very strong presumption that Parliament did not intend to legislate inconsistently with them.Footnote 59 With this very strong presumption at their disposal, could courts always find a way to interpret later statutes to be consistent with constitutional statutes?Footnote 60
The difficulty is that even a very strong interpretive presumption can be rebutted by very strong evidence to the contrary (as courts have acknowledged with respect to s. 3(1) of the Human Rights ActFootnote 61). And there can be very strong evidence indeed that Parliament intended to legislate contrary to a constitutional statute. Consider s. 45 of the Scotland Act, which states: “The First Minister shall be Keeper of the Scottish Seal.” Suppose Parliament enacts the Keeper of the Scottish Seal Reform Act 2014. Its only provision says: “Henceforth the Lord Advocate of Scotland shall be sole Keeper of the Scottish Seal.” (This is not an express repeal of s. 45 of the Scotland Act because neither that section nor the Scotland Act is named or described, nor is that section said to be “repealed” or of no “force or effect” or anything similar.) There is overwhelming evidence in the form of statutory text and context that Parliament's intended enactment is inconsistent with the Scotland Act. No matter how strong an interpretive presumption is at the disposal of courts, there is no plausible, consistent interpretation of these statutes.
Even if later statutes can be inconsistent with the Scotland Act, would the implication be that the Scotland Act is repealed to the extent of the inconsistency? Could both statutes remain “on the books”? It helps to see why such an implication arises when ordinary statutes are inconsistent. That requires some explanation of a third kind of implication, in addition to the two we have already discussed. An implicit assumption is what a speaker reasonably takes for granted as likely to also be taken for granted by his or her audience.Footnote 62 It is what goes without saying – what is too obvious to be worth making express. For example, if you order a hamburger in a restaurant, you “would not think to add that the hamburger should not be encased in a cube of solid lucite plastic that can only be broken by a jackhammer”.Footnote 63 Your order implicitly requires a hamburger that can be eaten without great difficulty, so this is part of the meaning of what you have said.
Courts regularly interpret statutes to include what Parliament would reasonably have taken for granted. As Richard Ekins says, the “legislature may safely leave various points unsaid, say that the offence its enactment creates or regulates is limited to acts within its jurisdiction, does not apply retrospectively and does not oust the standard criminal law defences”.Footnote 64
The kind of implication at work in the doctrine of implied repeal is also an implicit assumption. In the most basic terms, Parliament enacts a statute to change the law, and it aims to change the law so as to change the way that we, the law's subjects, act. Its aim is to guide our conduct, in other words.Footnote 65 But normally the law cannot guide our conduct to the extent it is inconsistent.Footnote 66 To take a simple example, if the law tells you to drive on the left of the road, and also to drive on the right, you have not just received poor guidance; you have not received any guidance at all. It would be self-defeating, therefore, for Parliament to enact a statute without at the same time repealing earlier and inconsistent statutes. That this is obvious is precisely why Parliament can safely leave it unsaid – as implicit – that when it enacts a statute, it also intends to repeal inconsistent provisions of earlier statutes.
So we grant that it is unlikely that Parliament intends to repeal a constitutional statute, and that this is a strong reason for thinking that the implied meaning of a later statute is not that a constitutional statute is repealed. It is even less likely, however, that Parliament would act in a transparently irrational and self-defeating way, which it would do were it to enact a statute while leaving inconsistent statutes intact.
In summary, Parliament can make its intention to repeal a constitutional statute clear without making it express. It can do so by attempting to expressly repeal a constitutional statute and failing because of a drafting error. It can also do so by enacting a statute that is obviously inconsistent with a constitutional statute.
One final point. We have said that there needs to be very strong evidence to infer that Parliament intended to repeal a constitutional statute, but that such evidence need not come in the form of express words. That is consistent with the test articulated in Thoburn, because Thoburn allows for the repeal of a constitutional statute so long as the evidence of an intention to repeal is “irresistible”. Laws L.J. did not justify his decision on interpretive grounds (as we shall explain shortly). Our point is simply that he could have.Footnote 67 We shall return to this aspect of Thoburn in the final section of this paper.
IV. Judges, Powers, and Recognition
If Parliament were to communicate its decision to repeal a constitutional statute by implication, but the effect would not be the repeal of that statute, then Parliament's will would be frustrated. Parliament would lack the power to bring about the repeal of that statute by implication. Does Parliament actually lack that power?
Let us be clear about what is not at issue here. Had Parliament said that the Scotland Act could only be expressly repealed, we would need to know whether, and to what extent, Parliament's powers include imposing requirements as to “form” on itself.Footnote 68 There would be a parallel with s. 2(4) of the ECA, which might be understood as imposing a requirement as to the form of future legislation, to the effect that Parliament must use express words to effectively repeal the ECA.Footnote 69 The question would be whether H could be justified by analogy with a prominent interpretation of Factortame. But of course Parliament has not said that the Scotland Act can only be expressly repealed, or that later statutes only take effect insofar as they are consistent with it. It has not said that expressly. Nor has Parliament said as much by implication: clearly not by deliberate implication or by implication arising from a defective expression; and not by implicit assumption, either.Footnote 70 We know there was no such implicit assumption because Parliament could not have reasonably taken it for granted that the Scotland Act cannot be impliedly repealed. That is a novel, or at the very least controversial, proposition. It is not a proposition that is too obvious to be worth mentioning in the way it is too obvious to mention that a hamburger should not come encased in lucite.
What is at issue, then, is whether Parliament can be bound in a way not of its own choosing and, if so, how. Here we come to the theory that the common law is the basis of parliamentary sovereignty. We shall refer to this as “common law constitutionalism”, although we recognise that some other ideas attract the same label.Footnote 71 Supposing for the sake of argument that common law constitutionalism is correct, and supposing also that judges are able to unilaterally change the common law, it follows that judges can unilaterally limit Parliament's powers. That suggests a possible legal basis for H: the Supreme Court in H did not recognise an existing constraint on Parliament's powers; it created one. Put another way, H might be legally justified because the Supreme Court had the authority to impose on Parliament what is in effect a requirement as to form.
Both Laws L.J. and Lord Hope appear to favour this kind of common law constitutionalism. Laws L.J., for example, said in Thoburn that, although Parliament cannot impose manner and form requirements on itself, judges can impose equivalent requirements.Footnote 72 Lord Hope said in Jackson v Attorney General that “[t]he principle of parliamentary sovereignty … in the absence of higher authority, has been created by the common law”.Footnote 73 He recently expanded on his remarks extrajudicially:
[A] law does not simply exist. It has to come from somewhere. It is either enacted law, for which Parliament is the source, or it is a product of the common law by the judges. There is, as Lord Bingham says, no statute to which the principle [of parliamentary sovereignty] can be ascribed.Footnote 74
Lord Hope also said that judges are able to change the principle of parliamentary sovereignty by themselves.Footnote 75
This type of common law constitutionalism has come under stinging criticism. The main objection is that the above analysis – i.e. parliamentary sovereignty derives from either statute or the common law; it does not derive from statute; therefore it derives from the common law – is misleading. The reason for thinking that parliamentary sovereignty cannot be based on a statute is that Parliament cannot confer power on itself. By the same reasoning, the common law, if it is judge-made law, cannot be the source of judicial power. If we then accept that all legal power derives from either statute or the common law, we must conclude that Parliament was given its powers by judges, and judges their powers by Parliament. In that case, judges and Parliament are not pulling themselves up by their own bootstraps; they are pulling each other up by each other's bootstraps, which is to say the argument is circular. The point has been well made by others and we shall not labour it here.Footnote 76
Not every common law constitutionalist believes that judges can change the common law. Trevor Allan, for example, adopts a Dworkinian conception of the common law according to which the common law is a set of norms based on the fundamental principles of political morality that are best able to justify the legal system as a whole. These principles include (Allan says) liberty and democracy and, although they may change over time, judges cannot deliberately alter them. Judges are limited, therefore, to identifying and expounding the common law. Because on this theory the common law is not judge-made law, it is not vulnerable to the bootstrapping objection above. Does this theory lend any support to H? In outline, an argument for a positive answer would presumably go like this: following Factortame, the law on implied repeal is unsettled; because the law is unsettled, we should make sense of the rival positions in light of constitutional principle; and the best interpretation of constitutional principle is that constitutional statutes cannot be impliedly repealed.
To this line of thinking (which, to be clear, we are not attributing to Allan, and which he may well not supportFootnote 77) we have two responses. First, we do not accept that, before H, the law on implied repeal was unsettled in any general sense (more about why in a moment). As a result, the suggested limit on Parliament's power does not “fit” official practice in the way that Allan (following Dworkin) would require.Footnote 78 Second, it is far from clear that constitutional principle favours a general, absolute exemption from implied repeal for the Scotland Act or constitutional statutes generally. Such an exemption is unnecessary for the prevention of grave injustice, say, or the violation of basic rights.Footnote 79 Also, while an exemption would help ensure that Parliament accepts the “political cost”Footnote 80 of interfering with fundamental statutes, an exemption from implied repeal would run counter to the weighty democratic principle that favours giving effect to the clear “will of an elected assembly”.Footnote 81
Let us now set aside common law constitutionalism. In truth, most constitutional scholars and judges accept that the foundation of legal authority, including Parliament's sovereignty, is not a common law rule but rather what Sir William Wade called a “political fact”Footnote 82 and what H.L.A. Hart termed an “ultimate rule of recognition”. For our purposes, there are two crucial points to make about an ultimate rule of recognition. First, an ultimate rule of recognition sets out basic criteria of legal validity. Depending on what those criteria are, a constitutional statute may or may not be “recognised” as valid. Second, an ultimate rule of recognition is a social rule that exists amongst law-applying officials, including, but not limited to, judges. That is to say, it is a rule that exists because law-applying officials generally accept it and conform to it.Footnote 83
From this second point, it follows that the Supreme Court, and even judges as a whole, cannot unilaterally alter the rule of recognition.Footnote 84 They need the cooperation, or at least the acquiescence, of other law-applying officials. We know, as a result, that the Court in H did not change the ultimate rule of recognition. It could not have. (Lord Hope seems to accept that legal authority is based on an ultimate rule of recognition.Footnote 85 That is hard to square with his belief that legal authority is also based on the common law.Footnote 86 Perhaps Lord Hope believes that the rule of recognition is a common law rule. That would clearly be a mistake if, as Lord Hope seems to think, the common law is capable of being changed by judges alone, because, as we have now explained, an ultimate rule of recognition cannot be changed by judges alone.)
But did the Court need to change that rule? Perhaps, before H was decided, the ultimate rule of recognition already recognised the quasi-entrenched status of constitutional statutes, and the Court in H simply acknowledged that fact. Whether this is true depends on what law-applying officials accepted and how they acted prior to H. Was their practice consistent with a rule that recognised constitutional statutes as quasi-entrenched, or as susceptible to implied repeal?
It seems clear that, traditionally, officials accepted that Parliament was able to impliedly repeal any statute. True, there were dicta in several cases, most notably MacCormick v Lord Advocate,Footnote 87 to the effect that certain “Articles” of the Acts of Union 1706 and 1707 could not be repealed.Footnote 88 But these dicta were narrow. They do not support an exemption from implied repeal for other Articles, let alone for the Scotland Act, let alone for constitutional statutes generally. More importantly, they were never widely adopted. Indeed, traditionally, almost all judges presented the doctrine of implied repeal as applicable to all statutes.Footnote 89
Judges did not just say that they accepted that constitutional statutes could be impliedly repealed. They acted on that basis, too, as did other law-applying officials. The Anglo-Scottish union legislation is, in fact, a good example. Consider Article XXI, which says that the “Rights and Privileges” of Scotland's “Royal Burroughs” are to “remain entire” notwithstanding the union with England. Article XXI has never been expressly repealed. However, the Local Government (Scotland) Act 1973 stripped the royal burghs of their powers, and s. 1 of that Act provided that “all local government areas existing immediately before” the Act came into force “cease to exist” at that point. No court has been asked to determine whether Article XXI remained valid after the Local Government (Scotland) Act came into force. But judges and other law-applying officials have clearly proceeded on the basis that the royal burghs have been abolished, that local government business is now to be conducted by the bodies established by the Local Government (Scotland) Act (i.e. regions, districts, and so on) instead, and that Article XXI is no longer of force or effect. Other Articles of Union have been treated as impliedly repealed, too. For example, Article XX, regarding heritable jurisdictions, was never expressly repealed, but heritable jurisdictions were abolished by the Heritable Jurisdictions (Scotland) Act 1746.Footnote 90
The rule of recognition could have changed recently, so that what officials accept and act upon is different than it was in even the 1970s, but there is no real evidence of that change having occurred. Since Factortame, it may have been accepted that Parliament cannot impliedly repeal the ECA, but that is for reasons specific to that statute.Footnote 91 In Thoburn, Laws L.J. and Crane J may have accepted that Parliament's powers are limited more widely, but, as we explained, that view was not widely adopted. Overall, therefore, the practice of law-applying officials pre-H was consistent with a rule that allows Parliament to impliedly repeal even a constitutional statute (other than, perhaps, the ECA). In other words, pre-H, Parliament possessed a power that, on one reading of his dictum, Lord Hope claimed that it lacked.
V. Constitutional Change and Judicial Disobedience
So we do not accept that, as a matter of law, the Scotland Act or other constitutional statutes can only be expressly repealed. Factortame can potentially be justified on the basis that Parliament chose to exempt the ECA from implied repeal, but Parliament did not choose to exempt the Scotland Act, or constitutional statutes generally, from implied repeal, and judges cannot give them a quasi-entrenched status on their own. The narrow test for implied repeal set out in Thoburn can be justified on the basis that it is unlikely that Parliament intends to repeal a constitutional statute, but H cannot be justified on that basis, because Parliament can make its intention to repeal a constitutional statute “irresistibly” clear without making it express. H lacks any legal justification, therefore.
Could H still be justified on broader normative grounds? Let us outline an argument for a positive answer. The argument has two parts: (1) the quasi-entrenchment of constitutional statutes is, on the whole, desirable; and (2) H helps make it the case that constitutional statutes are quasi-entrenched. Starting with (1), it may be that Parliament would tend to decide whether to repeal a constitutional statute more carefully were it forced to make its decision to do so express. It may also be that Parliament would be less likely to repeal constitutional statutes, including statutes that protect basic rights, were the “political cost” of doing so higher. Now, we are unsure whether these good consequences would actually come about and, if they did, whether they would outweigh the drawbacks of quasi-entrenchment, which could be considerable. Given the quantity of legislation that might be considered “constitutional”, and the difficulty of predicting conflicts between statutes in the abstract, quasi-entrenchment could make it very difficult for Parliament to reliably achieve its legislative aims. However, we shall set aside our reservations, and assume for the sake of argument that quasi-entrenchment is desirable.
We turn now to (2). If our analysis in the previous two sections is correct, then judges have no legal authority to make constitutional statutes quasi-entrenched. Quasi-entrenchment requires a change in the ultimate rule of recognition, and judges cannot change that rule by themselves. However, judges can influence that rule's development. The ultimate rule of recognition is, after all, a consensus among law-applying officials as to the content of that rule. One way to shift that consensus is to foster the impression that it has already shifted. To make quasi-entrenchment a reality, judges might declare that constitutional statutes are quasi-entrenched, and generally treat the issue as settled and obvious. If they did so over a long enough period of time (in a series of obiter remarks, say), then other law-applying officials might come to take it for granted that constitutional statutes are quasi-entrenched – at which point, they actually would be. The constitution would have been changed, and a decision like H could have played a part.Footnote 92 (To be clear, we do not think that the Supreme Court was insincere in this way in H. We are merely suggesting a possible justification for H.)
So, quasi-entrenchment may be desirable, and H could help to make quasi-entrenchment a reality. There may be something to be said for H, therefore. H would still not be justified in an overall or all-things-considered sense, however, because there are powerful objections to the strategy outlined in the last paragraph. For one thing, it is risky. As Goldsworthy says, any attempt by judges to deliberately change the content of the ultimate rule of recognition is “fraught with danger”.Footnote 93 Other officials might be “persuaded, inveigled, bamboozled, or bluffed into acquiescing in the change”.Footnote 94 But they might, instead, “resent and resist the judicial attempt to change the rules that had previously been accepted, and take strong action to defeat it, possibly including the impeachment of ‘over-mighty judges’”.Footnote 95 Judges are poorly equipped to win this kind of power struggle, and quasi-entrenchment is not important enough to warrant the risk of losing.
Moreover, if judges were to refuse to hold that a constitutional statute is repealed absent express words, they would be refusing to apply the law. That ought to trouble us on principle. Judges are legally obligated to apply the law, of course, so their refusal would amount to disobedience. But judges are also under a moral obligation to apply the law, except when it would cause injustice.Footnote 96 The reasons for this moral obligation are familiar.Footnote 97 Judicial fidelity to the law preserves the coherence of the legal system, encourages the resolution of disputes according to the law, and sets an example for citizens. Most importantly, judicial fidelity to the law upholds the rule of law itself. Quasi-entrenchment may be desirable, but that does not entitle judges to disregard the law, or to imperil the rule of law, in order to achieve it.
VI. Conclusion
Let us conclude, then, with two guidelines we think are appropriate for the implied repeal of constitutional statutes. The first is implicit in what we have already said. An ordinary statute repeals an earlier constitutional statute if, and only if, either: (1) it says expressly that the constitutional statute is repealed; (2) it would have said that expressly, had it not been for a drafting error; or (3) the two statutes are clearly inconsistent. What is the meaning of “clearly” in (3)? We gave an example of a clear inconsistency earlier, involving the Keeper of the Scottish Seal. The inconsistency between the Acts of Union and the Local Government (Scotland) Act is another. The standard we have in mind is, we think, the same standard that Laws L.J. articulated in Thoburn, when he said that an “irresistible” implication could bring about the repeal of a constitutional statute.Footnote 98 This is also the standard applied to statutes that endanger common law constitutional rights and principles, as articulated in Simms and subsequent cases.Footnote 99 What we favour, therefore, is allowing for the implied repeal of a constitutional statute, but in narrower circumstances than ordinary statutes. We favour a departure from the traditional doctrine of implied repeal (according to which a repeal is effected by a “mere” implication), but not the radical change proposed in H.
As we explained, if two statutes are inconsistent, and both remain “on the books”, then the law fails to guide our conduct to the extent of the inconsistency. In other words, the law fails to “rule” to that extent. The rule of law, in this formal sense, is part of the British constitution. When a court chooses to uphold an ordinary statute over an inconsistent constitutional statute, it protects the rule of law, and thus a constitutional principle, albeit at the expense of a constitutional statute. Once we reflect on this tension within the constitution, it is easy to see the possibility of another. For it is possible that two constitutional statutes will conflict (e.g. the Scotland Act 1998 and an earlier constitutional statuteFootnote 100). In that case, the presumption that Parliament does not intend to legislate contrary to constitutional statutes is less obviously relevant. So, more tentatively, we propose a second guideline, namely that one constitutional statute repeals another constitutional statute if the two are (merely) inconsistent. The issue is complex, however, and deserves a longer treatment than we can provide here.
These guidelines are designed to give due weight to parliamentary intent while acknowledging the existence of constitutional statutes and the absence of a fundamental shift in the rule of recognition. They are meant as an alternative to the dictum in H, and as such do not give constitutional statutes quasi-entrenched status.
