„Counter-law“, as theorised by Richard Ericson, involves „using law against law“: the use of legal resources „to erode or eliminate traditional principles, standards, and procedures of criminal law“ and facilitate pre-emptive policing interventions in areas such as counter‑terrorism. This, Ericson argues, threatens the rule of law, defined as the principle that „[p]olice and citizens alike should know what is and is not legally authorized […] to ensure a predictable environment in which to make rational choices about rule‑governed behavior“.
Should the image of „law against law“ be discounted as a polemical gesture to add weight to a liberal critique — or can it be grounded in a defensible theoretical model of the Rule of Law? Such a model will be outlined, and its features contrasted with twenty-first-century developments in British counter-terrorist legislation — developments which, it will be argued, exemplify “counter-law” tendencies and make the Rule of Law more relevant than ever.
1. Rule of Law: Minimal, Maximal or Neither?
Ericson’s definition of the Rule of Law (hereafter RoL) echoes Hayek’s formulation: „government in all its actions is bound by rules fixed and announced beforehand“. The content of those rules is entirely undefined: a state under the RoL may impose any combination of duties and freedoms on its citizens, or on different groups of its citizens. All that is required is that everything the state does is first enacted into law.
Spelt out, this minimal definition of the RoL does not seem particularly desirable. It is highly permissive, but at the same time highly restrictive — so much so as to make deviations and trade‑offs inevitable. The question would not be whether government conduct falls short of the RoL but how often, in what ways and with what costs and benefits.
Other definitions of the RoL go considerably further than Hayek’s. Bingham’s widely-cited definition of the RoL includes provision for the protection of fundamental human rights and state compliance with international law, as well as stipulating that laws should be intelligible, apply equally to all and bind the government as well as its citizens. This has been criticised as excessively wide-ranging; Raz criticises Bingham’s definition as “an assembly of diverse principles, with diverse rationales behind them”, arguing that “the law, to be just or legitimate, or fundamentally good, should conform to more than one moral principle or doctrine”.
Raz for his part offered two alternative extended definitions of the RoL. His 1979 definition combines a formal definition in terms of the properties of laws themselves (which should be clear, stable, publicised, prospective and general) with a set of principles of procedural justice. In 2019, Raz again cited the necessary formal properties of laws, then added a set of provisions relating to the reasonableness of government decisions. Ironically, both of Raz’s own definitions are vulnerable to a critique similar to his own. Each of the two definitions is, arguably, not a singular doctrine but a pair of principles, with the first group of stipulations defining good laws in formal terms and the second defining the proper administration of justice and good governance respectively.
The multi-dimensionality which Raz decries in Bingham’s definition of the RoL — and which, to a lesser extent, can also be identified in Raz’s own definitions — makes compliance with the RoL difficult to achieve: the more internal complexity the concept is understood as having, the more likely it is that government action will depart from one or the other aspect of it. Once again, the RoL is defined in such a way as to make trade-offs inevitable, reducing the force of any critique which would make departure from the RoL an evil in itself.
In short, the RoL represents either a procedural restriction on government, or a set of principles whose partial or total realisation is conducive to greater government legitimacy. In both cases departures from the RoL are easy to justify on other grounds (e.g. for reasons of public safety), meaning that the RoL cannot be invoked as a stand-alone ground for critique.
2. A Third Alternative: The Formal Rule of Law
It may be possible to rescue the RoL as a ground for critique by using a purely formal definition, akin, though not identical, to the first half of Raz’s two pairs of principles. The definition proposed here draws (like Raz’s definitions) on Lon Fuller’s eight „principles of legality“. Fuller argues that laws should be general, be publicised and be prospective in effect; should not be unintelligible, mutually contradictory, impossible to obey or so changeable as to be impossible to identify; and should exhibit „congruence between the rules as announced and their actual administration“, to be secured primarily through excellence in legal drafting, interpretation and administration.
I have argued that Fuller’s eight principles can be summed up as the principles that laws should be universal, knowable and followable, and that the RoL also requires a fourth principle of justifiability. In fact, the eight principles reduce to two. Generality, publicity, intelligibility, stability over time and congruence of official action are factors of knowability: only if all these requirements are met is it reliably possible for any individual, in any situation, to ascertain what laws effectively apply to them. Knowability thus entails comprehensiveness: there can be in principle no social situation, and no group of people, not covered by any law.
Secondly, the requirements of prospectivity, non-contradiction, and possible obedience are factors of followability. Followability entails freedom of choice to follow a law (or not to do so): if the requirement of followability is to be met, an individual’s social existence may not be structured to the point where no margin of choice remains, nor may obedience to law generally be secured oppressively, through actual or threatened coercion.
Justifiability is not a separate principle but an implication of the first two. If following the law’s commands is a free choice (followability) based on knowledge and comprehension of the law (knowability), it must be possible to explain the applicability of a law to a given person in a given situation, and for the explanation to be rationally challenged. If law-compliant behaviour is not to be secured by force, it must also be the case that a rational challenge will, in some cases, succeed, and the law be amended accordingly.
Like Raz’s models, this formal model of the RoL does not encompass every standard to which the law in a just society would conform. However, it does not go as far as Raz’s argument that the RoL „has no bearing on the existence of spheres of activity free from governmental interference „, to the point that „[t]he law may […] institute slavery without violating the rule of law”. These assertions run counter to the requirement of followability: a freely followable law must guarantee some “spheres of activity free from governmental interference”, while a law to enslave rational and previously free citizens is one that could only be secured by oppressive force.
The formal model does, however, omit most of Bingham’s desiderata, including the fundamental liberal requirement of equality before the law: while it must be possible for each individual in a society under the RoL to ascertain and choose to follow the laws applicable to them, there is no requirement that the same laws apply to all. The formal RoL is thus compatible with a high degree of social stratification — and, as such, with “very great iniquity”, at least if that word is used in the archaic sense of “want or violation of equity”. It grounds only a weak universalism: discriminatory treatment of different groups is compatible with the RoL thus defined, for as long as those groups accept their treatment as just.
However, the formal RoL is arguably better suited than either the minimal or maximal models to be used as grounds for critique. Whatever its other defects, any society operating under the RoL must have laws that can be known and can be followed (and can be justified in response to challenge); defects from any of these properties are identifiable as shortfalls from the RoL.
3. Counter-Law and Counter-Terrorism: Preventive Offences
Contemporary counter‑terrorist legislation in the UK is based on the Terrorism Act 2000. While the 2000 Act’s definition of terrorism has remained largely unchanged, a series of additional Acts of Parliament have since been passed, in response to the September 11th 2001 attacks and subsequent terrorist attacks in the UK. Multiple new offences have been created, many of which can be seen to violate the requirements of knowability and followability. Examples include the groups of offences classified as inchoate (offences of attempt or encouragement), preparatory (otherwise lawful conduct in preparation for the commission of an offence) and situational (offences defined in terms of a state of affairs).
Inchoate counter‑terrorism offences can involve long and speculative causal chains: it is an offence under the Terrorism Act 2006 to state that a terrorist act is worthy of emulation, if the person making the statement intends, or is reckless as to the possibility, that hearers should be encouraged to commit or instigate terrorist acts. This offence has since been joined, under the Counter-Terrorism and Border Security Act 2019, by the offence of expressing an opinion or belief indicating support for a proscribed organisation, recklessly as to the possibility of others being encouraged to share this support. To commit either of these offences does not require that anyone is in fact encouraged in these ways, let alone that any terrorist act takes place.
The key preparatory offence in UK counter-terrorism legislation is the catch-all offence of ‚preparation of terrorist acts‘ under the Terrorism Act 2006, covering „any conduct in preparation for giving effect to [the] intention“ to commit a terrorist act. Given that the goal of counter-terrorist legislation is preventive rather than reactive, the preparation offence makes it possible to criminalise actions which were harmless and otherwise lawful, on the grounds that the offender possessed an intention which these acts would — if they had not been interrupted — have realised in the commission of acts of terrorism.
Situational counter-terrorism offences, lastly, are offences where a guilty act is inferred from a state of affairs, elevating what might more usually constitute circumstantial evidence to an offence in its own right; an example is the offence of possessing information likely to be useful to a person committing or preparing an act of terrorism (Terrorism Act 2000). This offence has been expanded by the Counter-Terrorism and Border Security Act 2019 to include having viewed any such information over the Internet.
Given the crucial importance of the offender’s—unrealised—intention in making out the elements of offences, offences like these are neither knowable nor followable. A law prohibiting the commission of any act which might subsequently be presented as having been preparatory to what might have been a future act of political violence is not a law whose scope can reliably be known. Similarly, a law prohibiting any statement which could be presented as tending to encourage support for one of a designated list of groups (even if no such encouragement had taken place) is not a law that can freely be followed: it enjoins either silence or the affirmative act of a disclaimer.
Instead of the classical requirements of ‚guilty act‘ and ‚guilty mind‘, these offences share two unusual and complementary characteristics. The acts brought to trial are not themselves harmful and could potentially be proved against a wide range of people, many or most of whom would have carried them out without criminal intent. When charges such as these are brought, the prosecution attests not only to the factual element of these offences but to the correct interpretation of the facts. These are ‚ouster‘ offences: the criminal court is ‚ousted‘ from its role of determining guilt by the public prosecutor, being presented with offences defined in such a way that only one verdict is possible.
This degree of latitude is only available to the prosecution because — secondly — these are terrorist offences, subject to their own body of legislation. What this means, however, is that not merely intent or recklessness but a specific guilty mental state has been assumed: the defendant is suspected of intending to carry out, or intending to threaten to carry out — or welcoming the possibility of others carrying out or threatening to carry out — one of the broadly defined disruptive acts listed in s2 of the Terrorism Act 2000, for the broadly-defined purposes listed in s1. The factual elements of the various offences look quite different if this assumption is not made: if the prosecutor is persuaded that the suspect is a harmless fantasist whose professions of innocence can be trusted, no terrorist intent can be inferred and no terrorist charges brought.
4. The Rule of Law: How to Use
Counter-terror legislation has created a catalogue of offences offering enormous scope for prosecutorial discretion, allowing for individuals to be convicted on the evidence of having committed innocuous acts. Moreover, in an extraordinary irony, suspicious acts and individuals qualify to be considered as ‚terrorist‘ on the basis of an act of more or less speculative inference—which is itself an exercise of prosecutorial discretion.
The Rule of Law, I have suggested, requires that the law be a reliable and non-oppressive guide to how citizens should act: as such, the laws governing every citizen must be rationally knowable and voluntarily followable (and, by extension, open to rational challenge and justification). These tendencies in counter-terrorist legislation clearly run counter to the RoL thus understood. Every move away from knowable and followable laws is a move away from the RoL — and towards a landscape in which police discretion determines not only who will be put on trial but who will subsequently be found guilty. Conversely, every move in this direction can be resisted by reference to the RoL, if this is understood as neither a technical desideratum nor a broad set of liberal ideals, but as the simple requirement that citizens should be able to know, understand and choose to follow the laws that apply to them.