An introduction to judicial review

Where it comes from, how it works, and why it’s so difficult to reform

In many branches of the civil service, new recruits are provided with a document entitled The Judge Over Your Shoulder. Now in its sixth edition, this text — whimsically and unfittingly abbreviated to ‘JOYS’ — is freely available on the gov.uk website. It bills itself as a guide on ‘what to expect when working with government lawyers, allowing for effective collaboration and lowering the risk of legal challenge’. More succinctly, it has been described as a tool for ‘judge-proofing’ state actions.

Friends and colleagues on the Right often lament judicial intervention in state policy. The height of right-wing hostility towards the courts came after the case of R (Miller) v The Prime Minister [2019] UKSC 41, better known as ‘Miller 2’, where the Supreme Court decided unanimously that Boris Johnson’s use of the royal prerogative to prorogue Parliament during the Brexit deadlock had been unlawful. Although the ultimate effect of the decision was nil after the spectacular election result the same year, the decision came at the zenith of a period of intense, feverish panic among the British political class, and a full-spectrum effort to prevent Brexit at any cost.

Yet lost among the memories of that strange year is a proper Right account of what the court was actually doing in Miller 2, and what the process of ‘judicial review’ actually is. Why is it that government policies, or even decisions of local authorities, are routinely found to be unlawful, and nullified by the courts? Where does this power come from, and what (if anything) are its limits? What, precisely, are these judges doing when they are looking ‘over your shoulder’, and why?

The purpose of this essay is to explain what this process is, why it has taken the form it has in Britain, what it demands of our governments, and why it has proved so difficult to tame. This essay does not deal with the specific problems introduced by specific legislation, such as the Human Rights Act 1998 or the Equality Act 2010. While noxious, they can get in the way of identifying exactly what this process is, and why it looms so large in our political system.

---

Judicial review is, simply put, exactly what it sounds like: it is the inquiry by a court into whether actions undertaken by public authorities were lawful. Something like it has existed for centuries. The classic case taught to young law students is Entick v Carrington, a 1765 case examining the lawfulness of searches carried out by the king’s men, but older precedents as far back as the 1610 Case of Proclamations are still routinely cited as representing an early strand of what we now call constitutional or administrative law. But what is notable about the older cases is that they did not attempt to found a distinct public or administrative branch of law; more often, they were about determining whether someone was entitled to a remedy in normal, ‘private’ law, and hinged incidentally on whether the government had some right. In Entick v Carrington, for instance, the courts ruled that Carrington could be held liable for a tort — i.e., Entick could indeed sue for the £2,000 worth of damages caused by their search — as Lord Halifax (who had ordered the search) did not have the power to do so under either statute or precedent, meaning the warrant he had issued was of no significance and Carrington was trespassing.

But first, what is ‘private’ — as distinct from ‘public’ — law? In ‘private’ law, where A sues B for something, A is generally telling the court that B has wronged him, and therefore he is entitled to get something from the court. B has done a wrong to A either by hurting him (a ‘tort’) or by breaching an obligation. As a natural consequence of that wrong, A is entitled to be compensated for that wrong. The court decides whether A or B is right, and from that decision, the outcome flows naturally. If A is right, he gets his remedy; if B is right, A does not. It is clean; it is logical; at its core, it is almost mathematical.

So much for private law. But judicial review does not work this way. Instead of a private law system of obligations and wrongs, judicial review applies what we might loosely call ‘public law’ or ‘administrative law’. The ‘central idea of administrative law’ (per the great scholar Sir William Wade) is the idea of ultra vires, i.e., whether power (vires) has been exceeded (ultra).

Let us imagine that the fictional Department for Administrative Affairs has made a certain decision: perhaps it has decided to give permission for a particular building to be constructed on government land; or perhaps it has decided to give public money to a certain project; or perhaps it has decided to intern someone as a suspected terrorist. Judicial review does not suggest that the Department has committed a wrong against someone; rather, its core inquiry is to ask whether the Department has the power to do what it has done.

All well and good. That much is straightforward. But: (a) how does the court decide whether or not the Department has the power; and (b) what happens if it does not?

Question (b) is simpler. The court has three remedies: quashing orders (formerly called certiorari), mandatory orders (formerly mandamus) and prohibitory orders (formerly prohibition). Mandatory orders force the government to do something; prohibitory orders forbid it from doing something. However, the most important is certiorari. If a government decision is quashed, the decision is termed ‘void’. What this means is not only that it ceases to have legal effect from the date of the order, but that it never had legal effect to begin with. Every subsequent decision based on that decision is unwound; every legal instrument entered into on the basis of that decision is immediately vulnerable to challenge as itself being ultra vires. A single quashing order can undo, at a stroke, years of government practice and millions of pounds of spending, requiring the Department to go back and decide each and every affected sub-decision differently in light of the court’s judgment.

From the answer to question (b) flows the answer to question (a), and for this some history is needed. In the early days of the development of modern judicial review, the courts understood there to be a difference between what they termed ‘jurisdictional’ and ‘non-jurisdictional’ errors. Jurisdictional errors were those which meant that the entire exercise of the power was invalid — because the decision-maker had no jurisdiction to make it, i.e., he should not even have been involved in that area — and consequently the whole decision was unwound from the very beginning. Non-jurisdictional errors, by contrast, were not ‘void’; rather, they were only ‘voidable’: they only ceased to have legal effect from the date of the judgment, and actions taken before that judgment on the basis of the erroneous legal decision were still lawful.

That distinction was erased by the most important case in English administrative law, Anisminic Ltd v Foreign Compensation Commission [1969] 2 AC 147. Anisminic Ltd was a mining company that operated in Egypt; when Nasser forcibly nationalised the Suez Canal in 1956, he also seized certain property belonging to Anisminic. Some years later, Egypt and the UK came to a settlement about proper compensation for those seizures, and that compensation was distributed under the Foreign Compensation Act 1950. Anisminic applied to the relevant commission for compensation under the 1950 Act, but was denied because its successor in title — i.e., the company that supposedly inherited its property — did not have British nationality, meaning it was not eligible for compensation.

Unsurprisingly, Anisminic sought review of the decision. On its final appeal to the Appellate Committee of the House of Lords (now the Supreme Court), the court first decided that the commission’s definition of ‘successor in title’ was incorrect: Anisminic was still in existence, so it had no successor in title. The concept could not even apply to Anisminic.  

But Anisminic was still faced with a problem. Section 4(4) of the 1950 Act stated simply: ‘The determination by the commission of any application made to them under this Act shall not be called into question in any court of law.’ Game, set and match to the Government, you might think: Parliament has provided that the court shall be unable to review the decisions made by the commission, and so Anisminic would have to lump it. Indeed, this clause was one of only two such clauses that Parliament decided to fully exempt from the general abrogation of ouster clauses in the Tribunals and Inquiries Act 1958.

Not so. By a 3-2 majority, the court held that the court always had power to inquire into whether there had, in fact, been a decision (‘determination’) at all, and it found that — quite contrary to the impression you might get from reading the case — the Commission had simply never made the decision it appeared to have made.

This sounds very strange: clearly, from a layman’s perspective, there had been a decision. But because the commission had made an error of law in deciding how to interpret ‘successor in title’, its decision was a nullity, and void. No decision, in law, had ever been reached: the decision was not merely wrong, it did not exist. The determination of the Commission was only a purported determination. For the majority, as Lord Morris put it in a rather sarcastic dissent, ‘what looks like a determination was in fact no determination’.

Since then, the doctrine has developed into a broad orthodoxy that holds that all errors of law are jurisdictional, and therefore completely unwind any decision which was erroneous: as the Supreme Court accepted in R (Cart) v Upper Tribunal [2011] UKSC 28, the distinction is simply between ‘lawful’ and ‘unlawful’ decisions. Only lawful decisions exist at law. Unlawful decisions are, legally speaking, ‘writ in water’ — and, naturally, the arbiter of what is lawful and what is unlawful is the court. As Lady Hale put it in her summary of the decision in Miller 2:

This Court has already concluded that the Prime Minister's advice to Her Majesty was unlawful, void and of no effect. This means that the Order in Council to which it led was also unlawful, void and of no effect and should be quashed. This means that when the Royal Commissioners walked into the House of Lords it was as if they walked in with a blank sheet of paper. The prorogation was also void and of no effect. Parliament has not been prorogued.

The political implications of this orthodoxy are considered further below, but there is a more obvious question: what is an error of law? Some are obvious: imagine an Act of Parliament authorises the Department to build a road, but it instead sends the money to a charity it considers deserving. That would be unlawful. Other types of error — what we term ‘heads of review’ — are, to put it mildly, rather less obvious.

On the uncontroversial side is the requirements of ‘natural justice’, a somewhat airy, open-sounding term for what is one might more familiarly think of as ‘due process’. A denial of natural justice occurs when someone has been subjected to state interference, or has had a certain state benefit withdrawn, without due process. This is contained in two core rules: the rule against bias, and the requirement of a fair hearing. A decision made by a biased decision-maker (such as giving a contract to your brother’s company), or a decision without a fair hearing (such as striking off a doctor without giving them a proper hearing), is legally not a decision at all, and thus can be unwound.

Another, more narrow head of review is ‘legitimate expectations’. This is a relatively new ground of review, but it likewise is largely uncontroversial in theory (although its boundaries are always up for debate). Basically, a ‘legitimate expectation’ is created when a public authority tells you in no uncertain terms it will do something, or not do something; in some circumstances, that public authority can be prevented from later going back on its word.

Neither of these, in general terms, create much of an obstacle to right-wing policy, and would be unlikely to be disrupted by any future right-wing government. It may be that a future government might want to tinker with the requirements of natural justice in some settings, or alter some of the situations in which a legitimate expectation can be countermanded. Other heads of review, however, are much more problematic.

The first is the issue of ‘considerations’, specifically ‘relevant’ or ‘irrelevant’ considerations. In theory, this is also unproblematic: the idea is that if a public authority considers something they should not have done, or fails to consider something they should have, then their decision is unlawful and null. For those considerations which are clearly set out in the statute law governing the decision, this makes sense: if a statute says the Department must consider a person’s income in giving him a certain benefit, and the Department does not do so, then that decision is unlawful. But problems emerge when the courts are not so bound by legislation.

For example, Jon Venables and James Thompson were convicted of the murder of James Bulger in 1993, carried out when they were both ten years old, and Bulger was two. They were sentenced to detention ‘at her Majesty’s pleasure’ — the youth equivalent of a life sentence — and under the laws at the time, the Home Secretary was empowered to determine the length of the ‘tariff’, the minimum period they were to serve before they could be considered for parole. The trial judge recommended eight years; the Lord Chief Justice recommended ten. The Home Secretary instead set a fifteen-year tariff, and in doing so, took into account not only the judicial recommendations, but also to a large number of letters he had received from the general public imploring him to set a high tariff — including a petition, for which The Sun collected signatures. Venables and Thompson appealed, and the matter made it up to the Lords (R v Home Secretary, ex parte Venables [1998] AC 407).

Lord Browne-Wilkinson and Lord Lloyd urged caution, suggesting that the Secretary of State should be entitled to take public concern into account and noting that Parliament had not imposed any such limits on the Home Secretary. Nonetheless, the majority of the court took the view that taking the petitions and letters into account was improper, and on that ground (among others) the decision was unlawful. In using a power that amounted to setting a sentence, the Home Secretary was required to ‘act within the same constraints as a judge will act when exercising the same function’, wrote Lord Goff. His Lordship laid special emphasis on the ‘tendency’ for the ‘perhaps natural’ desire of society for revenge against child-murderers to be ‘whipped up and exploited by the media. When this happens, it can degenerate into something less acceptable’. In this way, public clamour for harsh sentencing was explicitly identified by the highest court of the land as an improper consideration. One can only imagine how many times the Secret Barrister has read this case.

Just as dangerous, too, is where the court gradually expands the reach of the relevant considerations that must be taken into account. Nowhere has this been more challenging for the Right than on environmental policy. It appears to have reached its high water mark in the recent Friends of the Earth case ([2022] EWHC 1841 (Admin)). In that case, Friends of the Earth challenged the government’s formal Net Zero strategy, prepared to fulfil an obligation under the Climate Change Act 2008. Unfortunately for the government, while the Civil Service had produced briefings explaining the overall reduction in emissions that would result from the set of policies contained in the strategy, the briefings failed to provide predicted emissions reductions for each policy, nor identify the uncertainties. As such, the High Court concluded that this was an ‘obviously material consideration’ that had not been taken into account by the Secretary of State, and therefore the strategy was unlawful — despite the total absence of a statute that said he should do it this way. The reader is left to consider how this would work out for any environmental policy that was seriously at odds with the way we do things at the moment.

But chief among the issues is the two grounds for ‘substantive review’, i.e. not of how a decision is made, but of what the decision was. The first is ‘irrationality’ (traditionally called ‘unreasonableness’), and the second and more modern is ‘proportionality’.

The classic definition of an irrational or unreasonable decision is tautologous: it is a decision that ‘no reasonable decision-maker could have made’. Again, in theory, this could be highly subjective, but in practice it sets a very high bar, and the most common ground on which a decision is irrational is that it was based on a serious misunderstanding of the facts. While irrationality is open to abuse, in practice the worst excesses of the judiciary have tended to come from the invention of novel constitutional principles — on which more below — or the extension of more obvious grounds of review such as irrelevant considerations.

Proportionality is a different beast. Proportionality is generally understood only to apply to cases which have a European dimension, whether EU law (including retained EU law) or law relating to the European Convention on Human Rights (ECHR) and the domestic Human Rights Act 1998. However, it is all the more powerful for this apparent limitation: the basic principle of proportionality is that a right or interest recognised in these systems of law may only be interfered with if such interference is ‘necessary in a democratic society’. Lord Sumption set out the test for proportionality in detail in Bank Mellat v HM Treasury [2013] UKSC 39, holding that the court is to determine whether the public authority’s measure meets the following criteria (para. 20):

…the question depends on an exacting analysis of the factual case advanced in defence of the measure, in order to determine (i) whether its objective is sufficiently important to justify the limitation of a fundamental right; (ii) whether it is rationally connected to the objective; (iii) whether a less intrusive measure could have been used; and (iv) whether, having regard to these matters and to the severity of the consequences, a fair balance has been struck between the rights of the individual and the interests of the community.

Obviously, each of these questions is intensely political. The wicked genius of proportionality is to transform these purely political questions into a legal question. Anywhere human rights law or EU law touches becomes juridified, transmuted into a question of law, not politics — to which, implicitly, there is a right answer and a wrong answer, to be determined by the courts. Granted, the courts tend to extend a certain amount of latitude to the decisions of public authorities, but everywhere and always, the court sets the boundaries for what is acceptable and what is not. Even though the court may pay due deference to the Department on some matters, it is the court who decides how much deference is due in the first place.

The effects are obvious: a significant reduction in freedom of action of elected politicians, and a growth in the power and reach of the courts to determine what the state is to do. Above all, more and more policy becomes ultimately a question of law — hence the judge ‘looking over your shoulder’, determining whether the state’s actions are sufficiently important, too intrusive, and/or represent a fair balance. The effects of this are subtle but unmistakeable, and lead inevitably to the implicit delegitimisation of political decision-makers in favour of judicial decision-makers — away from political will and decision, and towards bureaucracy and procedure.

---

So, one might ask, given this context: why didn’t the Conservative government pass an Act of Parliament that changed all this? After all, the one thing everyone involved in this system appears to agree on — at least nominally — is that parliament is ‘sovereign’, that it has the power to make and unmake any law it chooses. That sovereignty was in abeyance during Britain’s membership of the EU, where EU law had a higher status than national law and could be used to overturn national law, but after Brexit there is no such barrier. There is, at least in theory, no reason that Parliament could not pass a Judicial Review Act to restrain or control judicial review, or at least prevent it from being applied to certain areas of government action.

Unfortunately, the problem is made knottier by the very basis of the doctrine of judicial review. The reader will recall from our earlier analysis of Anisminic that the Foreign Compensation Act 1950 provided, in very clear terms, that the decisions of the commission ‘shall not be called in question in any court of law’, but the court escaped the implications of this clear provision by holding that the decision did not exist at all. The court accepted, true enough, that it could not call a ‘determination’ of the Commission into question — but this was only a ‘purported determination’, because it did not exist. The tautology of this is obvious — how can one determine whether something is a ‘purported’ or ‘real’ decision, without calling it into question? — but that tautology has not stopped Anisminic from forming the doctrinal core of the judicial review powers of the High Court.

The depth of the taboo against excluding — or, in the legal jargon, ‘ousting’ — judicial review is perhaps hard to understand for laymen without a further example. In Page’s case [1993] AC 682, hundreds of years of legal tradition had held that the Visitor of a university — an external person or body charged with upholding university statutes, in something like a private judicial role — had an exclusive ‘eleemosynary’ jurisdiction, unappealable and unreviewable. But the House of Lords held that this was unsustainable after Anisminic: the Visitor could make errors of law too, and as such, the High Court could review them. Thus did a case about the Suez Crisis overturn three centuries of common law.

How far this doctrine could be taken was demonstrated to dramatic effect in R (Privacy International) v Investigatory Powers Tribunal [2019] UKSC 22. That case concerned the Investigatory Powers Tribunal, a special court established by the Regulation of Investigatory Powers Act 2000 (RIPA) to examine the conduct of the intelligence services and make certain decisions about their powers, such as in surveillance. RIPA contained what constitutional scholars call an ‘ouster clause’, a clause that attempts to exclude (‘oust’) the jurisdiction of the High Court to judicially review the decisions. It provided:

Except to such extent as the Secretary of State may by order otherwise provide, determinations, awards, orders and other decisions of the Tribunal (including decisions as to whether they have jurisdiction) shall not be subject to appeal or be liable to be questioned in any court.

This, the Supreme Court held (by a bare majority of 4-3), was simply not sufficiently ‘clear’ wording to oust judicial review. Indeed, Lord Carnwath suggested in his leading judgment that there was ‘a strong case for holding that, consistently with the rule of law, binding effect cannot be given to a clause which purports wholly to exclude the supervisory jurisdiction of the High Court to review a decision of an inferior court or tribunal’. In other words, the court suggested that Parliament — supposedly the sovereign in this country — may not, in fact, have the unlimited powers it is supposed to have, and simply may not have the authority to prevent the High Court from undertaking judicial review. There has been some indication in recent case law that some judges, at least, are beginning to row back on their aggressive hostility to ouster clauses, but whether this will continue into the future (or survive the Court of Appeal and Supreme Court) remains to be seen.

At the same time, however, the courts have been increasingly adventurous even where no statute is in play. A recent and striking example was the UNISON case ([2017] UKSC 51), where the Supreme Court held that an order of the Lord Chancellor that introduced fees on persons bringing claims in the Employment Tribunal was unlawful because it contravened a free-standing legal principle of ‘access to justice’. Specifically, they held that ‘even an interference with access to the courts which is not insurmountable will be unlawful unless it can be justified as reasonably necessary to meet a legitimate objective’ (para. 89, per Lord Reed). Naturally, there was no ‘Access to Justice Act’ to interpret, so who decides what is reasonably necessary, and what is a legitimate objective? Why, the courts, of course. The fact that the Lord Chancellor, the minister with constitutional responsibility for the courts, took a different view, is neither here nor there (see paras 83 to 84): on terms explicitly similar to the European ‘proportionality’ jurisdiction, the court appoints itself the final arbiter on these questions, regardless of what mere ministers have to say about it.

All of this takes place in the context of an inherent inequality of arms between the courts and Parliament. Parliament requires months, usually years, to legislate on a given issue: Green Papers, White Papers, drafts, committees, second reading, third reading, the House of Lords. By the time it is passed, it will take several more weeks for an Act to be given Royal Assent, and often several more months for it to be formally ‘commenced’, i.e., brought into force. Over none of that time is the given issue the sole concern of Parliament — it legislates on many issues all at once, and has no power to monitor or control implementation. Conversely, the courts decide on the basis of disputes brought before them, and take many weeks of detailed scrutiny of each individual case, before retiring to deliver a judgment of which every word is law. The courts can give dozens of judgments in the time it takes Parliament to draft and enact a single Act. Where Parliament wields a sledgehammer — huge and weighty, but slow — the courts wield a scalpel, nipping and tucking at all law, every day, simply because that is their function. It is therefore no surprise that even bold reforms are likely to wither on the vine: the courts simply have more opportunities to dodge the statutory sledgehammer.

What hope, then, for right-wing governments in the future? Drafting an Act to counteract all of these problems is a very tall order, and even if one were carried out, it is likely it would be interpreted into meaninglessness by the courts in precisely the areas we consider most important — especially on immigration and asylum. Indeed, in 2003, Tony Blair’s government planned to pass an ouster clause ‘the like of which had never been seen’, according to the editors of the august textbook De Smith’s Judicial Review. (They neglected to attach the gigachad.) Unfortunately, the ouster clause in question was just too based, and triggered an outraged reaction from the judiciary, with the Lord Chief Justice himself warning (or threatening) in a speech at the Cambridge Law Faculty that it ‘could be a catalyst for a campaign for a written constitution’. Blair, despite his colossal 2001 majority, backed down.

The Judicial Power Project at Policy Exchange has done important and valuable policy and academic work on what can be done to maximise chances of success. Several constructive suggestions were put forward to the Independent Review of Administrative Law under Boris Johnson’s government, suggesting a welter of corrective legislation to be fast-tracked to reverse recent, highly unpopular judgments, proposing the restoration of the concept of non-jurisdictional errors of law, and even hinting sotto voce at codifying the entire system of public law, giving Parliament back the initiative. Sadly, like anything potentially difficult undertaken by that government, it died on the vine, and we are unlikely to see anything of its like under the Rt Hon Sir Keir Starmer KCB KC MP.

Still important, however, is the art of diplomacy. Assuming Britain’s form of government remains largely the same — a bourgeois, democratic, crowned republic — a right-wing government will have to deal civilly with judges and lawyers, as an exceptionally powerful interest group with their own preoccupations and their own desires. It will be necessary for a serious government to confront this problem head on, and speak to lawyers in a language they understand: lawyers should not be treated as class enemies, but as wayward and unruly officers to be corrected and led.

Ultimately, there is no reason lawyers cannot be brought on as potential converts. Yes, their class interests present certain obstacles, but public lawyers are not universally popular in the profession, and a bold and self-confident government would engage with the profession and make the case for a reset in how judicial review functions. That would probably require an able, amiable, and quick-witted Lord Chancellor, in post for the entirety of a Parliament, or longer, who can monitor the implementation of the Government’s agenda and assuage the concerns and doubts of the profession as far as possible, while not bending on the reforms that need to be undertaken. On such a man the English constitution would turn for a century. We should find him now, while there is still time.