What is the relationship between parliamentary sovereignty and the rule of law? The assumption that pervades the literature is that parliamentary sovereignty and the rule of law sometimes clash, so that it is sometimes necessary for the courts to ‘qualify’ the former so as to protect the latter. Qualifying parliamentary sovereignty is distinguished from abandoning it on the basis that Parliament retains the right to violate the rule of law so long as it uses ‘express language or necessary implication’. Thus the tension between the two principles is said to be managed.
Because this way of thinking treats parliamentary sovereignty and the rule of law as autonomous principles, I call it the ‘autonomous conception’. In a newly-published article, I argue that the autonomous conception is unsound. This blog post provides an overview of my argument. I start by presenting some problems the autonomous conception faces on its own terms. I then present my argument for the alternative understanding, in which sovereignty and the rule of law are essentially interlinked.
Problems with the Autonomous Conception
- A statute has no ‘prima facie meaning’ abstracted from constitutional principle
If parliamentary sovereignty and the rule of law are autonomous principles, then it must be possible to determine legislative intention without reference to what the rule of law requires. This attitude is evident in Privacy International, where Lord Carnwath declared that, when dealing with statutes that threaten the rule of law (here an ouster clause), the courts do not engage in ‘ordinary statutory interpretation’, and that ‘Parliament’s intentions [are] beside the point’.
This approach is not, however, consistent with the way in which statutory interpretation operates in ‘ordinary’ cases. Outside of public law, the presumption that the rule of law should be upheld is seen, not as a challenge to legislative intent, but a way of ascertaining what that intent is. A statute will not readily be interpreted as requiring the impossible, having retroactive effect, creating a strict liability offence, or violating the maxim that like cases be treated alike. These presumptions are not treated as ‘qualifying’ parliamentary sovereignty, but rather as an important way of determining legislative intent.
There is good reason why this should be so. As an institution, Parliament is defined not simply by its membership, but by the role it exists to fulfil; as Lord Steyn memorably put it, ‘legislat[ing] for a European liberal democracy founded on the principles and traditions of the common law’. We are interpreting, not acts of parliamentarians, but Acts of Parliament, and we must attribute to Parliament a concern for the basic values of liberal democracy.
- The ‘clear statement principle’ discloses no principled interpretive standard
As noted above, the autonomous conception allows Parliament to violate the rule of law so long as it uses ‘express language or necessary implication’. The question here is not what Parliament’s intention is, but whether it has expressed its intention with sufficient clarity. However, once we drop the search for legislative intent, there is no sensible way of determining whether language is sufficiently ‘express’.
The idea that we can identify ‘express language’ without considering the intention of the speaker treats language as a kind of code. But this is not how language generally, nor legislative language in particular, works. Semantic meaning always underdetermines content: faced with a general rule, it will always be open to a party to argue that some fact about which the statute is silent entails that that rule does not apply to their case. Semantics alone cannot resolve such a dispute. Instead, we must view the enactment as a form of rational action, and ask what interpretation of the statutory wording can best explain Parliament’s decision to pass the enactment in those particular terms. In other words: we attribute an intention to Parliament.
It would, of course, be possible to devise a code, by stipulating that a particular form of words must be used to meet the clear statement principle. As Adam Tucker has neatly put it, this approach involves treating the clear statement principle as a quest for the ‘magic formula’. But this would not be an appropriate line for the courts to take. Their role is to adjudicate concrete disputes: they have no constitutional warrant to dictate, prospectively, requirements that Parliament must use certain words to achieve its ends. And this leads us on to the third problem with the autonomous conception.
- The autonomous conception engages the courts in a legislative role
The flipside of legislative intention being detached from the rule of law is that the rule of law is detached from legislative intention. To work out what the rule of law requires becomes a matter of curial discretion. The courts’ role is depicted as essentially legislative rather than adjudicative.
This way of understanding things underlies Varuhas’ critical discussion of recent public law caselaw. The use of the clear statement principle in novel scenarios, Varuhas argues, places hurdles on Parliament’s ability to legislate as it wishes. These hurdles are imposed ex post, but with an ex ante justification: ‘Parliament should, in passing this legislation, have squarely confronted interference with constitutional norms’.
The autonomous conception leaves the courts open to Varuhas’ attack. It presents the courts as a counterweight to Parliament, and likens statutory interpretation to a game of ‘cat and mouse’ (another apt choice of words from Tucker). This places the courts on constitutionally shaky ground: not only does it unreasonably expect Parliament to second guess how courts will respond to its enactments, it also presents the courts as engaged in actively creating new norms where they deem parliamentary sovereignty to be in need of ‘qualification’ – essentially a legislative function.
However, Varuhas’ criticisms may be met if we understand public law adjudication as simply the ordinary interpretation of legislation in politically-charged circumstances. For what Varuhas sees as a new hurdle (e.g. the right to access to the courts in Unison) might actually be nothing more than an articulation of how a deeper principle, already embedded within the law, applies to a novel set of facts.
Overcoming Lingering Austinianism
The shift from an ‘absolute’ to a ‘qualified’ understanding of parliamentary sovereignty is usually considered to be a liberal and progressive development. But underlying the apparently modern autonomous conception lies the lingering legacy of John Austin’s discredited conception of legislation as command.
Austin argued that every law is a command: a request that something be done, coupled with a potential sanction for disobedience. Faithfully complying with a command means doing what the person who issued the command intend one to do. If law is a command, then when interpreting a statute we should ask: ‘What does Parliament want done?’ The autonomous conception, with its empiricist, value-neutral understanding of Parliament’s intention, makes sense on this basis.
This understanding of legislation was, however, rebuffed by Hart in The Concept of Law. Law consists not of commands, but of ‘general standards of behaviour to be followed by the society’. To legislate is not to order someone to do something, it is to create a new legal norm. The statute “No vehicles are permitted in the park” in itself creates a rule of law whereby no vehicles are permitted in the park, without anyone having to do anything. The proper approach to statutory interpretation is thus not ‘What does Parliament want done?’ but ‘What has Parliament done?’ – i.e. ‘What legal norm has Parliament created?’
This understanding of legislation imports the rule of law into legislative intention. Legislation can only be understood in the context of a legal system, i.e. one in which political power is exercised in accordance with general rules. This justifies a host of assumptions about the way in which Parliament intends public officials to use their legal powers: they should be exercised for proper purposes, in a broadly consistent way, respecting the principles recognised by the legal system, without undermining citizens’ access to the courts, and so on. These assumptions are not – as the autonomous conception would have it – principles the courts ought to develop to protect ‘the rule of law’ as an abstract value. Rather, they follow from a full understanding of what it means to legislate, and are thus necessary for us to reach a considered view of Parliament’s intentions. Parliamentary sovereignty does not need to be ‘qualified’ by the rule of law. Parliament rules through law; its enactments must be interpreted as law.
Note that we are presuming, not that Parliament intends to act morally, but that it intends to act rationally, given the enterprise it is engaged in. Remember again that Parliament is an institution, not an aggregation of individuals. It is defined by the function it exists to serve, namely to enact generally binding standards of conduct so as to promote justice and/or the common good. Whatever ignoble intentions individual parliamentarians might have, it makes no sense to attribute ulterior motives to Parliament. We must assume that it carries out its function rationally, and so we cannot impute to it intentions that would undermine its purpose.
Conclusion
The assumption that the rule of law ‘qualifies’ parliamentary sovereignty is unwarranted. It misrepresents how statutory interpretation works in ‘ordinary’ cases, it drives us towards an ultimately incoherent ‘clear statement principle’, and it casts the courts’ role in upholding the rule of law as legislative in nature. Most fundamentally, it wrongly reduces legislation to a kind of command. Since to legislate is to make law, Parliament must be understood to intend that the exercise of public powers does not undermine the system of government under law. This principle of statutory interpretation might not always endear itself to parliamentarians, but it does not undermine the sovereignty of Parliament.