At the end of a long review of my book Parliamentary Sovereignty, Contemporary Debates (CUP, 2010, hereafter PS), Vernon Bogdanor concludes that I have “suffered one of the worst fates that can befall a philosopher”: I have “become the prisoner of a doctrine” – that of parliamentary sovereignty (“Imprisoned by a Doctrine: The Modern Defence of Parliamentary Sovereignty” (2011) Oxford Journal of Legal Studies 179, hereinafter Bogdanor ). Apparently, self-imposed mental barriers prevent me from perceiving how the British constitution has changed.
What are Bogdanor’s reasons for this conclusion?
First, he puzzles over my claim that at present Parliament “cannot” limit its own substantive powers: is the impossibility a logical or contingent one? Since it cannot be a logical impossibility, he infers that I must mean that “it would be perfectly possible for Parliament to pass immutable laws but there are very good reasons why, as a matter of fact, it has never done so.” On this view, my “conception is a utilitarian one”, and is open to refutation “were it to be shown that the consequences would not necessarily be undesirable.” (Bogdanor,183) But this rests on a misunderstanding of my position. It is true that I believe there are good reasons of political morality for Parliament not being able to limit its own substantive powers, which I discuss in my book (PS, 7, 53-5, 116, 125-9). But I maintain that Parliament cannot lawfully limit its own substantive powers, because there is currently a fundamental legal rule (a rule of recognition) to that effect, whose existence is constituted by its being generally accepted by legal officialdom. That rule can be changed, but only by a change in official consensus, and not by Parliament acting unilaterally (116, 137-8). That is what I mean by “cannot”.
Secondly, Bogdanor believes that Parliament’s sovereignty has been limited by the European Communities Act 1972 (“EC Act”) (Bogdanor, 182). He acknowledges my suggested interpretation of the somewhat cryptic Factortame judgment, as construing the EC Act as imposing a kind of “manner and form” requirement that Parliament must use explicit language in order to legislate inconsistently with applicable EC law and with that Act (184). (He does not mention that I also offer two other, alternative interpretations of the judgment that are equally consistent with parliamentary sovereignty as I define it. (PS, 289-90, 296-8)) But he prefers the alternative view put forward by counsel in the Thoburn case, which he says “deserves more consideration than it has so far received.”(Bogdanor, 184) According to this view, the EC Act altered Britain’s fundamental rule of recognition with the effect that Parliament cannot unilaterally derogate from EC law. Consequently, “national courts would have to apply EC law in preference to inconsistent national law”(185). “If that is so”, he says, the EC Act limits Parliament’s substantive power. (186)
I freely acknowledge that this is a possible interpretation of the Factortame judgment, as I did in my book where I attributed it to Paul Craig (PS 287). There, I said that if Parliament can legislate inconsistently with applicable EC laws only if it first enacts legislation withdrawing Britain from the EC, “then the EC Act would have subjected Parliament’s lawmaking power to a limitation of substance”, which “could not be explained in terms of a mere requirement as to the form of British legislation.” (287-8) However, I argued that this interpretation was implausible because “it is the business of the government and Parliament, not the courts, to decide whether or not Britain should abide by its treaty commitments.” (287) In other words, if Parliament were to pass a law that explicitly contradicts applicable EC law, but without withdrawing Britain from the EC, and therefore violates Britain’s treaty commitments, that would be its business, not the courts’. They do not have legal authority to enforce treaty commitments in the face of a statute to the contrary. Nor is there much evidence that, on this point, there has been the requisite change in official consensus for the rule of recognition to have changed, as Bogdanor suggests it has (Bogdanor, 184). My argument then proceeded on that basis.
It is, of course, possible that the courts would take the opposite view, preferred by Bogdanor. But the fact that I offer reasons for disagreeing with a view that he thinks “deserves more consideration than it has so far received” (184) hardly suggests that I am imprisoned by a doctrine in the sense that I am unable to see beyond it. Nor does the fact that it is a possible view prove that it is the correct one, or that if Parliament’s substantive power have been limited in this respect they must be limitable in other respects as well (186). That is precisely what is in contention.
Thirdly, Bogdanor suggests that Parliament has also succeeded in limiting its substantive powers by enacting what he calls “binding” referendum requirements in s.1 of the Northern Ireland Act 1998 and ss.2-4 of the European Union Act (‘EU Act’) (187-8). But neither of these referendum requirements strikes me as “binding” Parliament in the requisite sense. This is because they are not self-entrenched: they can be repealed or amended by ordinary legislation, without any referendum being held (PS, 144). I emphasized the importance of self-entrenchment in PS (at 144), and in my written evidence to the House of Commons European Scrutiny Committee, which Bogdanor refers to (The EU Bill and Parliamentary Sovereignty, HC 633-II (7 December 2010), Ev 31, para 10; Bogdanor, n38). It might be argued that they are binding on Parliament until it repeals or amends them (Bogdanor, 189). But that argument would invite two responses. First, if Parliament were to legislate inconsistently with these requirements – for example, by legislating for the ratification of a treaty covered by s. 2 EU Act without any referendum first being held as required by that section – it is very likely that the referendum requirement would be regarded as repealed pro tanto by implication. That conclusion follows from the principle established in McCawley v R [1920] AC 691 (PC). But say I am wrong about this, and the courts were to hold that Parliament must follow a two-step rather than a one-step procedure, and expressly amend or repeal such requirements before legislating inconsistently with them. My second response is that this would amount to a requirement that Parliament must expressly amend or repeal such requirements, rather than doing so by implication (through the enactment of inconsistent legislation). According to my theory, a requirement that Parliament must expressly repeal or amend a provision is merely a requirement as to form, and not as to substance. This is not a minor, incidental aspect of my theory: it is central to it, and is discussed at length in my book (PS 179-82, 195, 289-90). It forms part of my argument that the EC Act, as construed in Factortame, is consistent with parliamentary sovereignty. Thus, a referendum requirement is perfectly consistent with my conception of parliamentary sovereignty, provided that it is not self-entrenched and can therefore be repealed or amended, whether impliedly or expressly, by ordinary legislation.
Bogdanor continues his discussion of referendum requirements by confusing (as many British writers do) two different arguments that attempt to justify them, which in my book I distinguish: the manner and form argument, and the reconstitution argument (Bogdanor, 189-190; see PS, 139, 155-60, 167, 171-73 and 198-99). The reconstitution argument construes referendum requirements as changing Parliament’s composition, by making the electorate part of Parliament for particular purposes. That is quite different from arguing that Parliament has subjected itself to a requirement governing the procedures or forms by which it must legislate. Bogdanor argues against the reconstitution argument (as I do), but then seems to assume that he has refuted the manner and form argument (Bogdanor, 189-90). Then follow sweeping claims: that Parliament in the EU Act has partially renounced its sovereign power to legislate; that since it has done so on that topic, it could do so on others as well; and therefore that it could gradually bind itself to a whole new constitution. But as I have shown, all this rests on a misunderstanding of the extent to which Parliament has already been able to bind itself. In statutes like the Northern Ireland Act and the EU Act it has not bound its substantive powers in any constitutionally significant sense.
Fourthly, Bogdanor argues that the judges have limited Parliament’s powers: it cannot protect Ministers or public authorities from judicial review, by conferring unfettered discretions or enacting ouster clauses (191). He asserts that Padfield and Anisminic make this “clear”, although he later says that “there is no case which allows one to confirm this interpretation.” (191, 192) I discuss Anisminic in my book, in a passage that Bogdanor does not mention (PS 285-6). All I can do is repeat what I said there: the House of Lords justified its decision on the orthodox ground of presumed legislative intention, and even if this was a “noble lie”, “the fact that a lie is felt to be required indicates that the judges themselves realise that their disobedience is, legally speaking, illicit” (PS, 286).
Fifthly, Bogdanor argues that if Parliament were to enact a statute subverting the foundations of the rule of law, it is possible that the judges would refuse to obey it. He refers to recent events and judicial statements (such as in the Jackson case) as possibly portending that outcome. It would shatter the official consensus on which (he and I both agree) the doctrine of parliamentary sovereignty has rested, and could lead to a constitutional crisis in which Parliament and the judges disagree about ultimate legal authority (Bogdanor, 193). His conclusion is that, since the question of what Parliament “can” do depends on the reaction of the courts, and we cannot predict what their reaction would be, it is impossible to say that Parliament “can” enact such a statute (193, 194).
Here again, Bogdanor is confusing what Parliament “can” or “cannot” do in a practical sense, with what it can or cannot do in a legal sense. I would say the following, to any judges who think that if Parliament were to enact a statute subverting the rule of law, they should hold the statute to be invalid.
“There is no basis in the constitution as it currently stands for you to do this: the long accepted rule of recognition gives Parliament sovereign lawmaking power. Therefore, you would be attempting to bring about constitutional change, so that what Parliament can legally do today, it could no longer legally do.
You can attempt such a change, but you cannot plausibly claim legal authority to do so by invoking the theory of “common law constitutionalism” – the theory that parliamentary sovereignty is a doctrine of the common law, which the judges created and can therefore unilaterally modify or repudiate. I have refuted that theory, and judicial support for it in Jackson’s case (such as in the judgment of Lord Steyn) is therefore based on demonstrable falsehoods. Obiter dicta that can be shown to be false should be accorded no authority.
There is no other basis in the constitution for your claiming authority unilaterally to change the constitution in this way. That does not mean you cannot successfully initiate such a change, but it does mean that the political branches of government must be persuaded, inveigled, bamboozled, or bluffed into acquiescing in it. But what if they are not? What if they resent and resist your efforts to change the constitutional rules that were previously accepted, and take strong action to defeat it, possibly including the impeachment of ‘over-mighty judges’? That might be regrettable, but if you tear up the consensus that currently supports the fundamental rules of the system, you are hardly well placed to complain if it is replaced by a power struggle you are ill-equipped to win. In the absence of consensus, your own legal authority as well as Parliament’s would be up for grabs. (This passage is adapted from PS, 55)
It might nevertheless be a good idea, before such a statute is enacted, for you to hint that you might refuse to obey it. The political branches may be just as concerned as you about the dangers of a constitutional crisis, and they might back down and decide not to enact it (as they did in 2003 when a sweeping ouster clause was withdrawn after protests by senior judges). But no matter how you couch such a threat, do not confuse in your own minds what would amount to judicial disobedience of the law, with the judicial exercise of constitutional authority. One of the practical restraints on Parliament’s exercise of lawmaking power is the possibility that it might not be obeyed. In an extraordinary case judges, like ordinary citizens, might be justified in disobeying a valid statute – and a fortiori, in hinting that they might disobey it. But it simply does not follow that it would be a good idea to abandon the doctrine of parliamentary sovereignty. What would replace it? A rule that the judges can subject Parliament’s authority to whatever limits they see fit (perhaps under the vague label “the rule of law”)? Remember that hard cases make bad laws. It would be better to retain the doctrine of parliamentary sovereignty, subject to the possibility of civil or even official disobedience in extraordinary situations, than to attempt to replace it with a rule of judicial supremacy capable of imposing fundamental constitutional changes on the nation.” (See also J. Goldsworthy, The Sovereignty of Parliament, History and Philosophy (OUP, 1999), 267-71)
Sixthly, Bogdanor argues that the concept of parliamentary sovereignty is of little value in analysing what Parliament can or cannot do. Indeed, the concept creates puzzles that would dissolve if it were abandoned (Bogdanor, 193-4). He recommends that we simply ask what rules govern Parliament’s composition, powers and procedures; whether they impose formal or substantive limits on legislation; how they are determined; and how they can be changed. The concept of sovereignty, he claims, plays no constructive role in answering these questions (194).
It was one of the purposes of my book to answer most of these questions. It is not clear to me whether Bogdanor disagrees with my answers, or with the way I use the concept of parliamentary sovereignty in arriving at or expressing them. For example, I reject the theory of “common law constitutionalism”, and in his book The New British Constitution (Hart, 2009) he seems to agree with me (at 82-3).
I am extremely sceptical about Bogdanor’s claim that the doctrine of parliamentary sovereignty is now useless for theoretical or practical purposes. For centuries, it has been generally understood that Parliament has sovereign lawmaking authority. This meant that there were no substantive limits to its authority, and it could not subject itself to such limits except by abdicating its authority with respect to territories capable of being excised from its jurisdiction. On the other hand, there was some uncertainty about its ability to subject itself to binding rules as to the procedure for or the form of legislation. A crucial question is: to what extent have recent developments changed these understandings?
A minimalist approach to answering that question, which I favour, construes somewhat cryptic developments, such as the effect of the EC Act as interpreted in Factortame, as altering previous understandings only to the minimum extent that is necessary to accommodate those developments. Factortame can be construed as using the EC Act as a very strong rule for interpreting later statutes, or alternatively, as imposing a binding rule as to the form of later statutes (PS 287-98). The minimalist approach leaves intact as much of the previous understanding as possible, for a number of reasons. An established rule of recognition is constituted by a consensus among legal officials, which it would be dangerous to construe as having radically changed unless there is reasonably clear evidence of the change. To the extent that an established rule of recognition remains unchanged, there is greater certainty about the allocation of constitutional authority, compared with a fluid situation in which it is regarded as having been repudiated but not yet replaced by any clear alternative. Uncertainty about such matters is more likely to cause conflicts between the branches of government. In addition, both of the major theoretical alternatives to orthodox (“continuing”) parliamentary sovereignty (“self-embracing” parliamentary sovereignty, and common law constitutionalism), which authorise either Parliament or the Supreme Court unilaterally to limit Parliament’s substantive powers, could in principle lead to very undemocratic constitutional change (116, 137-140).
Alternatively, a maximalist approach of recent developments can be taken, according to which the doctrine of parliamentary sovereignty is now dead, if not yet buried. But if so, what has replaced it? The problem is that if Parliament is no longer sovereign, almost everything is up for grabs. Bogdanor’s suggestion that we simply ask what Parliament can and cannot now do strikes me as naive. There is no strong evidence that orthodox parliamentary sovereignty has been supplanted by either of the usual alternative theories. As previously noted, Bogdanor does not seem to embrace common law constitutionalism, the theory that it is up to the courts in developing “the common law” to decide what limits Parliament’s powers. Does he, then, embrace the theory of self-embracing sovereignty, the idea that Parliament can limit its own powers in any way it chooses? He would not, of course, approve of the term “sovereignty”, which he does not find useful. But he might still endorse the idea that Parliament has the power to limit its own powers (Bogdanor, 183), notwithstanding the dangers of that idea (PS, 116, 137-138).
Can Parliament – in the legal sense of “can” – limit its substantive powers, such as by enacting a fully binding, self-entrenched, referendum requirement? I say it cannot – unless the rule of recognition is changed. This is a possibility I discuss: indeed, I suggest that if a binding referendum requirement were enacted only after being itself approved in a referendum, this would help to justify and fortify a change in the rule of recognition (PS 139-140). As I put the point in my written evidence to the House of Commons European Scrutiny Committee (to which Bogdanor refers),
“To make it more likely that, in this scenario, the judges would enforce the earlier statute prohibiting the future enactment of legislation without a referendum first being held, that statute should itself be put to a referendum. The support of a majority of voters for such a referendum requirement would greatly add to the strength of the case in favour of its future enforcement notwithstanding Parliament’s later change of mind, indicated by its attempt to legislate without complying with that requirement. This is because obtaining the support of the voters for a requirement that their support be required in the future overcomes a principled objection to the imposition of a referendum requirement by ordinary legislation. The objection is this: if an earlier Parliament can use ordinary legislation to implement its preferred policies, why should a future Parliament not have the same liberty? To put it another way, why should the later Parliament be bound by the expression of a will that has no higher authority than its own will? This is the main justification of the orthodox view that Parliament cannot bind itself. But if a referendum requirement is enacted with the support of a majority of voters in a referendum, the objection is overcome. A future Parliament could then be said to be bound, not by an earlier will of no higher authority than its own will, but by an earlier will that does have such a higher authority – the expressed will of the people.” (The EU Bill and Parliamentary Sovereignty, HC 633-II (7 December 2010), Ev 31, para 11; Bogdanor, n38.)
I am surprised to be accused of having argued myself into a blind alley – becoming imprisoned by the doctrine of parliamentary sovereignty – when I have suggested how, in this way, the doctrine could most legitimately and effectively be superseded. But, until a new constitutional settlement is clearly endorsed by the people, it would in my opinion be dangerously destabilizing to declare that parliamentary sovereignty is dead.
Jeffrey Goldsworthy is a Professor of Law at Monash University