- Timothy EndicottTimothy EndicottFellow in Law, Balliol College, Professor of Legal Philosophy, University of Oxford
At common law, the judges will hold administrative conduct to be unlawful on any of three grounds: error of law (and certain sorts of error of fact), lack of due process, and the improper exercise of discretionary power. This chapter discusses how (and to what extent) the three grounds of judicial review are supported by constitutional principle. Each ground must be controlled by the principle of comity. The principle of comity requires judges to defer to administrative authorities on some issues, to some extent; the chapter explains the limits of deference and the difference—and the connections—between the rule of law and the rule of judges.
At common law, the judges will hold administrative conduct to be unlawful on any of the following grounds: error of law and certain restricted forms of error of fact (Chapter 9), lack of due process (Chapters 4–6), and improper exercise of discretionary power (Chapters 7–8). This chapter raises the basic question of how those grounds of judicial review are supported by constitutional principle. Each ground must be moderated by the principle of comity.
Look for • • •
The question that the public authority must answer in order to do its job.
The court’s approach to that question: will the court impose its own answer, or review the public authority’s decision on some more restrained ground?
Deference by judges to administrative authorities, and the limits of deference.
The difference—and the connections—between the rule of law and the rule of judges.
The core rationale for judicial review.
‘It may be misfortune for the applicant that the court … cannot begin to evaluate the comparative worth of research in clinical dentistry; but it is a fact of life.’R v Higher Education Funding Council, ex p Institute of Dental Surgery  1 WLR 242, 262 (Sedley J)
In May 1995, a Serbian tank fired a shell at the accommodation block in Bosnia where British members of the United Nations Peacekeeping Force were staying. Trevor Walker was one of the peacekeepers sleeping in the compound. The shell injured his right leg so badly that it had to be amputated above the knee.
Walker applied for compensation under a criminal injuries compensation scheme that the Ministry of Defence had introduced in 1979. The scheme was not enacted in a statute and did not create any legal right to compensation; the Ministry merely decided, and announced, that the Army Board would make payments to members of the Armed Forces injured abroad in crimes of violence. The Ministry published special guidelines for peacekeepers in Bosnia, which stated that compensation was not available if a peacekeeper was injured as a result of ‘military activity by warring factions’.
The Ministry rejected Sergeant Walker’s application for compensation, on the ground that the attack on the accommodation block was military activity by a warring faction. Sergeant Walker applied for judicial review of the decision, and his application was rejected by a High Court judge. The Court of Appeal and (in a split decision) the House of Lords upheld the decision to dismiss his challenge (R v Ministry of Defence, ex p Walker  1 WLR 806).
Think of the question of whether Sergeant Walker should receive compensation. If the Law Lords rejected his appeal only because they agreed with the Ministry on that question, then the court in judicial review would simply be replacing the administrative decision with the decision that the judges would have made. Sergeant Walker would be able go to court to get the judges’ decision, instead of the Army Board’s decision. The Law Lords did not claim to be replacing the Army’s decision with their own decision in Walker. Instead, they claimed to apply the traditional grounds for judicial review. Understanding those grounds means understanding the difference between the rule of law and the rule of judges.
Remember the restriction on compensation: there was to be no compensation for injuries caused by ‘military activity of warring factions’. Walker claimed:
that the Ministry misinterpreted the restriction, so that the decision was based on an error of law;
that it had been ‘irrational’ for the Ministry to introduce the restriction at all; and
that the restriction had been introduced by an unfair process.
Note on styles of cause
Why wasn’t Walker’s case called Walker v Ministry of Defence? The answer is that Sergeant Walker had no cause of action—that is, he could assert no legal right to a remedy in an ordinary claim. But although he could not bring an ordinary claim, he was given permission (see 10.3.2, p 399) to bring an application for judicial review (since the Civil Procedure Rules 1998, it has been called a ‘claim for judicial review’) of the lawfulness of the decisions of the Ministry of Defence. The idea behind the name of the case—R v Ministry of Defence, ex p Walker—is that the Queen (‘R’ is short for ‘Regina’, the Latin for ‘The Queen’) initiates the proceeding on the application of Walker to ask her judges to review decisions of the Ministry of Defence to decide whether they were lawful. The ‘ex p’ (short for ‘ex parte’) meant that the proceeding was initiated without the Ministry of Defence being a party (the Ministry became a defendant when the court gave permission for the application to proceed).
The court was not asked whether Sergeant Walker had a right to compensation, but whether the Ministry misinterpreted its own rules, or used its discretion unlawfully in making its rules, or made the rules by a process that was unfair to Sergeant Walker. On the differences between a claim for judicial review and an ordinary claim, see 10.3, p 396.
Since the Civil Procedure Rules 1998 came into force, the courts have stopped using ‘ex parte’, and started using ‘on the application of’: for example, R (on the application of Begum) v Denbigh High School  EWCA Civ 199.1 The reason is that the initial application for permission to seek judicial review is no longer heard ex parte (that is, without the defendant being involved in the process). But the new style of cause maintains the medieval heritage: the Queen, in an exercise of her ancient prerogative to administer justice, brings the claim for judicial review on behalf of the claimant, in order to ask her judges to determine the official conduct that the claimant challenges was lawful.
These are not just technicalities; they reflect a feature of judicial review that has great practical importance: a claimant in judicial review does not need to assert any right to a remedy.
2.2 Error of law
Walker’s lawyers said that the attack that injured him was a crime, and that it could not count as ‘military activity’ if it was criminal. Lord Slynn said that the issue was ‘whether the Ministry of Defence has correctly interpreted the scheme … or whether p. 46↵its decision involves an error of law’ (Walker, 810). Along with the majority, he decided against Walker on this point because ‘in my opinion the exclusion from compensation as a matter of interpretation covers the injury to Sergeant Walker’ (812).
Lord Hobhouse dissented because he interpreted the scheme differently (at 819): ‘My Lords, Sergeant Walker is right to say that applying the Government’s own criteria his case falls on the right side of the line and he should be compensated … The attack was a criminal act not an act of war.’
When a public authority applies a scheme of standards to a case in order to decide whether to confer benefits or to impose burdens on a person, judicial review is available to correct an error in the agency’s understanding of those standards. So the judges in Walker asked: how are the guidelines for the scheme to be interpreted? The majority agreed with the Ministry’s interpretation of the guidelines, and Lord Hobhouse dissented because he agreed with Walker’s interpretation.
This ground of judicial review is remarkable. It is unlike the approach that judges take to interpretations of the law by public authorities in other common law countries, such as Canada and the United States (see 9.1.8, p 349). In England, the judges will not merely ask if other authorities’ interpretations of the law are unreasonable or arbitrary; a decision will be quashed if it runs contrary to the judges’ own interpretation of the standards the agency is applying. In English law, judges are to show no deference to other public authorities on questions of law.
Judges and academics use the word ‘deference’ in various ways. They sometimes use it to mean a spineless attitude. Everyone agrees that judges shouldn’t be spineless. Here, ‘deference’ is used instead for the attitude of a judge who treats a decision with respect because of the fact that it was made by another public authority. Some judges have been reluctant to use the term ‘deference’ in this way (see 7.2.4, p 258).
The fact that another public authority made the decision can require the judges to keep their hands off the decision altogether (as, for instance, when the Queen appoints a cabinet minister on the advice of the Prime Minister), or to scrutinize the decision and the reasons for it very closely (as they do when an administrative decision is made to detain a person—see 7.1.2, p 247). So in accord with the principle of relativity (see 1.3, p 13), judges can, and do, defer more or less to other public authorities, depending on the type of decision. When there is reason to defer, it is the constitutional principle of comity (see 1.5.4, p 22) that requires judges to do so.
Chapter 9 explains the doctrine of review for error of law, and why administrative authorities still have some leeway in applying the rules that they are responsible for applying. On a question of how to interpret the standards that other public authorities apply, the judges will impose their own view. As long as those standards are correctly p. 47↵interpreted, though, the judges may defer to other public authorities on particular questions of how to apply them to the facts of a case (the difference between interpreting and applying standards is explained in 9.3, p 366). Was Walker’s injury caused by military activity of warring factions? That was a question for the Ministry and not for the court. But if the Ministry decided it on the basis of a misinterpretation of the rule, then the court would strike down the decision.
Yet judges do sometimes say that they will quash a decision if the law was incorrectly applied in the case, and not only on the ground that it has been misinterpreted. So Lord Hobhouse said in Walker: ‘If the ministry fails correctly to interpret and apply the terms of the scheme, the decisions it takes are open to judicial review’ (817). Taken literally, those statements would make the administrative decision in a case like Walker merely provisional—to be replaced by the decision of a judge.
On the other hand, judges sometimes suggest that they cannot hold that a public authority has erred in law, unless there was something very wrong with their interpretation of the law. As Lord Slynn put it (813):
‘If I had come to the view that this phrase was imprecise enough for several meanings to be adopted, then I would not accept that the minister’s interpretation of it was such as to be “so aberrant that it cannot be classed as rational”.’2
The best way to make sense of the various judicial accounts of the judges’ role is this: English law requires them to apply a standard of correctness on questions of how to interpret the law (that is, on questions of law), but only a standard of reasonableness on the question of how to apply the law to the case. The judges often make it clear that they really won’t impose their own view on the question of how to apply the law. The public authority’s decision can go either way without being quashed on judicial review, as long as it is reasonable. Consider Lord Hoffmann’s reason for rejecting Walker’s argument on error of law (Walker, 815):
‘He was fired upon by a Serbian tank. I do not see how it can be said that the ministry could not reasonably take the view that this was military activity by a warring faction.’
This statement suggests that, in order to succeed, Walker would have had to persuade the judges that ‘the ministry could not reasonably take the view’ that it took—either because the Ministry misinterpreted the rules, or because the Ministry could not reasonably apply the rules (correctly interpreted) to his case in the way that it did. Essentially, in judicial review the judges require correct interpretation and reasonable application of the law. Chapter 9 explains the ways in which the judges give effect to those requirements.
Isn’t it odd that there were no legal rights at stake in Walker?
Lord Slynn asked ‘whether the Ministry of Defence has correctly interpreted the scheme … or whether its decision involves an error of law’. But the Ministry was not interpreting the law; it was only interpreting a scheme of payments made ex gratia (that is, given as a favour, not as a right). The judges did not even comment on this remarkable fact about Walker, which shows two respects in which the judges’ approach to judicial review is very broad: they will review a decision that does not affect the legal position of the claimant in any way, and they will quash a decision for ‘error of law’ if it is based on a misinterpretation of rules that create no legal rights and, in fact, have no legal status at all—apart from the legal status that they gain from the judges’ willingness to control their application.
Sergeant Walker also argued that if the restriction did apply, the Ministry never should have adopted it in the first place. But this argument was different: he could not simply claim that the Ministry had made an error in adopting the restriction, because error in the use of a discretionary power is not a ground of judicial review. The Ministry had no legal duty to create a particular sort of compensation scheme, or to create a scheme at all. So when the Ministry decided to offer compensation, the terms on which it would do so were up to the Ministry. It had discretion (see 7.2, p 253). The courts had no legal power to choose the terms on which compensation ought to be awarded. But they did have legal power to control the choice that the Ministry made if there was a shortcoming in the Ministry’s decision that made it unlawful. The courts are meant to decide whether the exercise of a discretionary power was lawful, without taking away the discretion.
Lord Slynn explained this ground of review in Walker by saying that: ‘It is not for the courts to consider whether the scheme with its exclusion is a good scheme or a bad scheme, unless it can be said that the exclusion is irrational or so unreasonable that no reasonable minister could have adopted it’ (812). That is the traditional approach to supervising the exercise of discretionary powers, and it gave Sergeant Walker a difficult test to meet. He argued that the restriction was ‘irrational’ because troops in Northern Ireland were entitled to compensation if they were injured by a terrorist attack, and, he argued, there was no rational basis for distinguishing between their situation and his. But Lord Slynn said, ‘the line may be fine, but to adopt it as a general rule cannot be said to have no rational base despite what seemed to me to be common features between the two situations’ (812).
So the judges’ approach to discretion, to some extent, is to keep their hands off it. They won’t interfere merely because they would have made a different decision, and they will not generally interfere on the ground that the public authority made the wrong decision. So Lord Hoffmann suggested that the Ministry of Defence used its p. 49↵discretion in the wrong way—and there still was no ground for interfering with its decision: ‘Speaking entirely for myself, I find the distinction a fine one … in neither Northern Ireland nor Bosnia were the British soldiers engaged in warfare … But I cannot say that the distinction drawn by the ministry is irrational. That is too high a hurdle to surmount’ (816).
What does it take to surmount that high hurdle? It takes a special flaw in the exercise of discretion—a flaw that allows judges to say that the decision was improper even though the public authority had a discretion. Sometimes this notion of a specially flawed exercise of discretion is summed up as an ‘unreasonable’ exercise of discretion. But as Lord Greene pointed out in the famous case of Associated Provincial Picture Houses Ltd v Wednesbury Corporation  1 KB 223 (CA), this word ‘unreasonable’ has to be treated carefully, in a way that is consistent with the lawful discretionary power of the public authority. To explain the special defects in a discretionary administrative decision that may justify judicial interference, Lord Greene said that an administrative decision maker must ‘direct himself properly in law’ (that is, must not base a decision on an error of law), and must:
‘call his own attention to the matters which he is bound to consider’; and
‘exclude from his consideration matters which are irrelevant to what he has to consider’ (229).
Doing otherwise is acting ‘unreasonably’ in the sense that justifies judicial interference with a decision. Then Lord Greene added (229):
‘Similarly, there may be something so absurd that no sensible person could ever dream that it lay within the powers of the authority. Warrington LJ in Short v Poole Corporation  Ch 66 gave the example of the red-haired teacher, dismissed because she had red hair. That is unreasonable in one sense. In another sense it is taking into consideration extraneous matters. It is so unreasonable that it might almost be described as being done in bad faith; and, in fact, all these things run into one another.’
The ‘something so absurd’ ground of review is often called ‘Wednesbury unreasonableness’. Remember that the principle that the court may overturn a decision if it is Wednesbury unreasonable is just one of the various grounds of review that Lord Greene mentioned, which are often collectively called ‘the Wednesbury principles’. He explained Wednesbury unreasonableness in three different ways. To be unreasonable in this sense, the exercise of a discretionary power must be so unreasonable that:
‘no sensible person could ever dream that it lay within the powers of the authority’ (Wednesbury, 229);
‘it might almost be described as being done in bad faith’ (229); or
‘no reasonable authority could ever have come to it’ (230).
p. 50↵He added that ‘to prove a case of that kind would require something overwhelming’ (230). Lord Greene was responding to the argument that unreasonable exercises of discretion are unlawful—and he did not simply accept that argument. He held that certain extraordinarily unreasonable decisions are to be struck down in judicial review.
This means that there is no general rule that unreasonable actions of public authorities are unlawful.
What is reasonableness?
To act reasonably is to be guided by reasons that ought to guide your action (see 7.1.3, p 250). Some of the reasons on which public decisions ought to be based can best be identified and assessed by the public authorities to whom a decision-making responsibility was assigned, rather than by the judges (who are responsible for the rule of law, and not for good government in general). An action is unreasonable in the special, restricted sense that provides a ground of judicial review if it is not guided by reasons that the law authorizes judges to insist on.
Only certain sorts of unreasonableness are grounds for judicial review. The reason is comity (see 1.5.4, p 22). No one should ever make an unreasonable decision, but a rule that judges should quash any unreasonable decision would give the judges a role in the work of other public authorities that would not actually improve administration; it would only shift administrative decision making into the courts. It is not generally the judges’ job to decide which administrative decisions would be reasonable. If there is no other specific ground for judicial interference, an action has to be unreasonable in a way that enables the judges, from their detached perspective on the court, to see that no person in the position of the public authority in question can seriously defend the action as the authority’s legitimate use of a lawful power.
Lord Greene’s speech is the most influential statement of the judicial control of discretionary power in English legal history, and the Wednesbury case has been cited thousands of times. But although Lord Greene said that he did not need to refer to authority ‘when once the simple and well known principles are understood’ (231), the decision can only be understood in the context of earlier cases on the control of administrative discretion. In Chapters 7 and 8, we will see the ways in which the law has developed before and since Lord Greene’s decision, and why (in spite of many premature announcements of the death of Wednesbury) his statement is still an integral part of the law today. The crucial points to keep in mind are (1) that the case involved a challenge to a local authority licensing decision and does not give a general code for the control of all discretionary powers of public authorities; and (2) that ‘Wednesbury unreasonableness’ was only one of the grounds of judicial review listed in Wednesbury, which also include the doctrine of relevance and the rule against bad faith.
Twenty years after the Wednesbury decision, in Padfield v Minister of Agriculture  AC 997, Lord Reid added another item to the list of special flaws that call for p. 51↵a discretionary decision to be quashed by the judges. In Padfield, the House of Lords quashed a decision that a politician had made for political purposes (see 8.2, p 294). The addition to the grounds of review stated in Wednesbury was that a statutory power can only be used ‘to promote the policy and objects of the Act’—and that determining those purposes ‘is always a matter of law for the court’ (Lord Reid, 1030).
Padfield was a striking and creative decision (and controversial: Lord Morris dissented vehemently). And yet it relied on two very old principles that, presumably, Lord Greene was taking for granted in his speech in Wednesbury—that a person who has a discretion must genuinely consider exercising it (see 8.1.2, p 289), and that it must be used for the ‘objects’ of the statute and not for the public authority’s extraneous purposes:
‘The question whether a judge, or a public officer, to whom a power is given … is bound to use it upon any particular occasion, or in any particular manner, must be solved … from the context, from the particular provisions, or from the general scope and objects of the enactment conferring the power’.3
Special flaws in a discretionary decision may justify judicial interference
Genuine exercise: a public authority must not refuse to consider exercising a power, or merely pretend to exercise it.
Relevance: matters that are irrelevant must not form the basis of the decision, and matters that the authority is bound to consider must be considered.
Proper purposes: the authority must not use the power for purposes that are incompatible with the reasons for which it was given the power.
The judges will quash a decision if ‘no reasonable authority could ever have come to it’ (Wednesbury, 230).
As Lord Greene pointed out, these grounds of review tend to run into one another. More precisely, acting for improper purposes, Wednesbury unreasonableness, and bad faith are all instances of acting on irrelevant considerations. If a public authority uses a power for some object that the legislation excludes, it is acting on irrelevant considerations. And without acting on some irrelevant consideration, a public authority cannot come to a decision that ‘no reasonable authority could ever have come to’. And acting in bad faith is simply an extreme instance of acting on irrelevant p. 52↵considerations: the fact that an action may hurt your enemies may seem appealing to you, but it is irrelevant to your exercise of a public power.
So the common threads in all of these grounds of judicial review of discretionary action are (1) the rule that a public authority must act on relevant and not on irrelevant considerations; and (2) the judicial determination to identify a hands-off way of controlling the considerations on which public authorities must act.
Now return to Walker, and consider whether he had a good argument that the Ministry of Defence had used its discretion unlawfully. The Ministry had discretion as to whether to set up a scheme, and if so, as to what restrictions to impose. So the court could only quash the decision if the Wednesbury principles authorized the court to do so. What were the considerations on the basis of which the Ministry introduced the restriction? The Ministry had decided that the original restriction (ruling out compensation for injuries from acts of violence ‘committed by an enemy’ in war) was too generous, since the peacekeeping role in Bosnia exposed the troops not to enemies of the UK, but to two hostile factions. The consideration that the scheme should not be too generous was a relevant consideration, and because of the Ministry’s discretion on the question of how generous it should be, the court deferred to the Ministry.
Statute and prerogative
Does it matter that the powers exercised in Padfield and Wednesbury were statutory powers, rather than prerogative powers? It seemed to make a difference when those cases were decided, because it was not until Council of Civil Service Unions v Minister for the Civil Service  AC 374 (see 7.1.3, p 250) that the House of Lords unequivocally held that the courts could review the exercise of the prerogative on the same broad grounds as the exercise of statutory powers. Today, the grounds of judicial review are essentially the same for the exercise of prerogative power as for the exercise of statutory powers (for one exception, see the discussion of R (Sandiford) v Secretary of State for Foreign and Commonwealth Affairs  UKSC 44 in 8.1.2, p 289).
• Pop quiz •
What was the source of the power that the Ministry of Defence exercised in creating the criminal injuries compensation scheme?
2.4 Due process
Sergeant Walker did not persuade the House of Lords that the Ministry had misinterpreted the restriction, or that it was an abuse of power for the Ministry to adopt the restriction. But he also challenged the process by which the Ministry had adopted the restriction.
The criminal injuries compensation scheme had been introduced with fanfare. The Ministry announced it to service commanders, to let Armed Forces personnel know p. 53↵that a generous new benefit was being created for them. When the Ministry decided to restrict the scheme in 1994 by excluding compensation for injuries caused by military activity of warring factions, the Minister announced the change in the House of Commons with no fanfare. It was not announced directly to soldiers or to their commanders. Sergeant Walker argued that this process was unfair. He claimed that he had a legitimate expectation (see 8.4, p 312) that he would be compensated under the scheme as it had originally been announced, and that it was unfair to disappoint that expectation without having first communicated the new restriction to the services as openly as the original scheme was communicated.
Walker was arguing that the Ministry’s conduct was unlawful because it was procedurally unfair. A decision is unfair if it wrongly neglects the interests of someone affected by the decision. The substance of the decision here was the new restriction. The procedure that Walker complained of was adopting the restriction without communicating with services personnel. The decision was substantively unfair if it was unfair for the Minister to restrict the scheme; it was procedurally unfair if it was unfair for the Minister to do so without announcing it to services personnel.
Decisions cannot be quashed for substantive unfairness in general. Unfairness is a form of unreasonableness, and we have seen that unreasonableness is not a general ground of judicial review. To ask a court to quash a substantively unfair decision, it is necessary to use the standards for the control of discretion (the special, restricted forms of reasonableness) that Lord Greene set out in Wednesbury, which are addressed in Chapter 8. Public authorities should not make decisions that wrongly neglect the interest of an affected person, but that does not mean that judges should replace the administration’s assessment of people’s interests with their own assessment. That is the basic reason for the restrained standards of control of discretionary power outlined in 2.3, p 48. Unfairness in the substance of an administrative decision is not in itself a ground of judicial review (see 8.6, p 325).
Procedural unfairness, by contrast, is a ground of judicial review, and a relatively clear and uncontroversial ground—although in Chapter 4 we will see fierce controversies about just what counts as unfair. Procedures are steps that a public authority takes in the course of making and announcing a decision, to communicate with people affected, and to hear from people who might have information to contribute or opinions that ought to be taken into account. Fair procedures give the persons affected an appropriate role in the process. Think of the decision to convict an accused person of murder. To treat the accused fairly, a criminal court needs to make sure that the accused is told what allegations he or she is facing and with what offence he or she is charged. The accused needs to be present when the prosecution makes its case, and needs to have an opportunity to respond to the evidence by challenging it and by introducing contrary evidence of his or her own. And if the accused is not able to do that, he or she needs help from someone who is competent to do it. In English criminal law, those requirements take the remarkable form of a jury trial with elaborate rules of indictment, disclosure before trial, a right to lead evidence and to choose whether to testify, a right to cross-examine prosecution witnesses, and a right to be represented by counsel and to have legal aid.
p. 54↵Associated with these demands of fairness is the requirement that the decision maker not be biased. In an English murder trial, that requirement is met by rules preventing a biased juror from participating in the decision and preventing a biased judge from presiding over the trial.
Fairness to a person affected by a decision is not the only reason for the law to impose procedural requirements. Good procedure in a murder trial is important to the public, too. It promotes good decisions, and makes the process open to the public, in a way that helps to hold the criminal justice system accountable.
Administrative decisions are all more or less radically different from the decision to convict a defendant of murder. And the procedures that are necessary for good decision making differ correspondingly. They depend on the nature of the decision, and on the relevant considerations (see 4.4, p 137). Administrative law only requires due process. It does not require the equivalent of a murder trial to test every action a public authority takes. The interests in participation, in openness, and in fairness call for different procedures in different decision-making regimes.
So we can return to Walker with an understanding of the special role for judicial review on grounds of due process: a court that is concerned not to usurp the Ministry’s discretion to decide the terms of the compensation scheme can still require the Ministry to use fair procedures in making its decisions. And Lord Slynn said that ‘it would have been better if the ministry had given a degree of publicity of the change similar to that given to the original proposal’ (814). If judges can impose procedural requirements without usurping the public authority’s role, and if they thought it would have been better for the Ministry to announce the change to services personnel, why didn’t Walker succeed on this ground of judicial review? Why didn’t the House of Lords hold that the Ministry’s procedure was unlawful?
The answer is that the requirement of due process does not mean that a court will interfere simply because the process could have been better. Lord Hoffmann said: ‘I do not think that your Lordships are concerned to decide in general terms whether it would have been better administration to make the announcement in a different way’ (815). The Law Lords’ conclusion was that the process was not unfair to Sergeant Walker. He did not know the 1980 policy, and he had not been told that his case would be treated under it. There is no doubt that the Ministry would have been acting in a more open and accountable fashion if it had publicized the new restriction as widely as it had publicized the new scheme. But judicial review is not a general technique for improving administrative process.
2.5 Constitutional principles and judicial review
The system principles of the constitution (see 1.5) ought to govern judicial review—in particular, parliamentary sovereignty, the separation of powers, comity among public authorities, and the rule of law. What is the link between grounds of judicial review and constitutional principles? Parliamentary sovereignty requires that judges pass p. 55↵judgment on the lawfulness of administrative decisions if Parliament directs them to do so (and requires them not to do so if Parliament has directed them not to do so). The separation of powers requires judges not to take over the powers of the administration, on the one hand, and on the other hand it requires judges to prevent administrative authorities from taking over powers that are not lawfully theirs. Comity requires judges to respect and to support the ability of administrative authorities to do their jobs.
The most important principle behind judicial review is the rule of law, and its demands are related to the demands of all of the other system principles. The rule of law requires that public authorities adhere to consistent, open, prospective, non-arbitrary standards, and it also requires decision-making processes that distinguish government action from the mere arbitrary whim of the people in power. But given that public authorities ought to act in those ways, why should courts interfere? Why not leave administrative agencies to construct and follow their own consistent, open, prospective, non-arbitrary standards, and to devise their own processes?
Here is an easy answer, which may seem plausible: law is for the judges, you may think, so that any question of the standards that administrative agencies ought to follow is a question for judges. The main point of this chapter is that it would be a mistake to think this. Judicial review can promote the rule of law. But we are not necessarily closer to the rule of law just because a question—even a question of law—is decided by a judge, rather than another public official.
2.5.1 How judicial review can promote the rule of law: the imprisonment of people convicted of murder
Replacing the decision of one official with the decision of another does not necessarily promote the rule of law. But it can do so. We can find the best examples in the courts’ claims to have imposed the rule of law on imprisonment after a conviction for murder. Parliament gave the Home Secretary power to decide how long to imprison underage murderers (detained ‘at Her Majesty’s pleasure’: Children and Young Persons Act 1933 s 53(1)) and, in the case of adults, the power to decide the ‘tariff’, or period to be served for the purposes of retribution and deterrence before consideration for parole (Crime (Sentences) Act 1997 s 29). In the 1990s, the judges, who openly opposed what amounted to a sentencing decision by a politician, imposed a variety of restrictions on its exercise. Each restriction promoted the rule of law. They held that:
the Home Secretary could not impose a secret tariff (R v Home Secretary, ex p Doody  AC 531);
the prisoner had a right to make representations (Doody);
the prisoner had a right to be given reasons for the decision (Doody);
the tariff could not ordinarily be increased retrospectively (R v Home Secretary, ex p Pierson  AC 539); and
p. 56 it was unlawful for the Home Secretary to base his decision on public clamour in favour of a harsh sentence for a particular offender (R v Home Secretary, ex p Venables and Thompson  AC 407).4
The European Convention on Human Rights went further than the common law could go in imposing the rule of law on the Home Secretary. First, the European Court of Human Rights declared that detention of child murderers at Her Majesty’s pleasure is contrary to the guarantee of an independent tribunal in the European Convention on Human Rights Art 6 (V v United Kingdom (2000) 30 EHRR 121).5 Then, after the Human Rights Act 1998 came into effect, the House of Lords held that it is incompatible with Art 6 for the Home Secretary to decide the tariff for an adult murderer (R v Home Secretary, ex p Anderson  3 WLR 1800). The government responded to Anderson by asking Parliament to amend the legislation, and the power to set the tariff was given to the judges.
Why say that these developments in English law promote the rule of law, rather than merely the rule of judges? Because each of them, in varying ways, stands in the way of arbitrary (see 1.1.3, p 10) decisions about imprisonment. The judges did what they could to insulate the Home Secretary’s decision from the pressures of the media that are liable to motivate a Cabinet minister. And taking the power away from the Home Secretary under the Human Rights Act detached it from those pressures. The result is that what happens to a person convicted of murder is ruled by a process of decision making that can be distinguished from a whim of the Home Secretary.
The rule against secret tariffs and the prohibition on retrospective increases promote the rule-of-law values of openness and prospectivity; the right to make representations promotes the rule of law by requiring the decision maker to face up to the factual claims and the arguments of the offender. The right to reasons supports that function of representations. A requirement of reasons increases openness, and also opposes arbitrary decisions, through the pressure it puts on the decision maker who has no rationale to offer. And the control on the considerations on which the Home Secretary may act promotes the rule of law by insulating the decision, to some extent, from political influence.
Note that there are more straightforward ways in which sentencing could be ruled by law: a statutory tariff, like the tariff of twenty-five years before parole in Canada for first-degree murder (Criminal Code s 745), would achieve legal control of the sentencing decision much more simply and effectively than all of the subtle judicial developments in the English cases.
The idea that the English judicial and legislative developments and the Canadian legislative tariffs promote the rule of law depends on an assumption. The assumption is that p. 57↵the period an offender spends in prison is the sort of decision that ought to be controlled by law. There are many sorts of decision that the judges will not review on the grounds they used in the life-sentence cases and to which Art 6 of the European Convention has no application: decisions about government expenditure, the appointment of ministers, and the disposition of the Armed Forces are the obvious examples. Why should the law control the time that life prisoners spend in prison in a way in which it does not control decisions about government expenditure, or the appointment of a minister, or the disposition of the Armed Forces? A good understanding of the rule of law requires an understanding of which governmental decisions need to be controlled by artificial and cumbersome legal processes in order to achieve responsible government. For the appointment of ministers, we can achieve responsible government through parliamentary controls on the Prime Minister and through general elections; judicial review of the decision would add nothing. In fact, judicial interference would detract from responsible government. To achieve responsible government in the imprisonment of murderers, on the other hand, the control that Parliament and the electorate can exert on government would not be enough. It would be arbitrary to subject a particular murderer’s life to political forces. We need the decision to be made by an independent tribunal.
Although all administrative decisions should be reasonable, a legal requirement of reasonableness would require courts to decide (to some extent) what reasons are good reasons. The courts have no general jurisdiction to do that, and it would be a breach of comity for them to invent one for themselves. Yet it is not a breach of comity for them to invent for themselves powers to control administration, where (1) they are capable of controlling a decision without damaging another public authority’s capacity to carry out its tasks, and (2) the alternative is to leave a public authority free to engage in arbitrary government. The invention of habeas corpus (see 1.1.1, p 5) is perhaps the oldest example of the use of the judges’ creative role, and in the twenty-first century the courts retain a dynamic jurisdiction to decide what new forms of judicial control are required for responsible government. Like any governmental power, the courts’ jurisdiction will itself be a tool for arbitrary government, if the judges do not use it responsibly.
2.5.2 Judicial deference to administrative authorities
In Walker, why were the judges reviewing the compensation decision at all, when the rules of the scheme had always said that ‘whether or not to make such a payment, and if so, the amount, shall be wholly within the discretion of the Army Board’? On the one hand, it seems that the courts should respect the fact that a decision has been committed to another public authority (whether the power is conferred by a statute or, as in p. 58↵Walker, by a prerogative act by the government). On the other hand, a very wide grant of power should not be turned into a tool for abuse. These two principles are in tension, but they do not contradict each other. In Walker, Lord Hoffmann pointed out that the scheme gave discretion to the Army Board, and simply added: ‘But the discretion may not be exercised arbitrarily’ (815). The central challenge for judges in judicial review is to prevent other public authorities from using their powers arbitrarily—which the judges need to do in order to impose the rule of law—without imposing their own judgment as to how other public authorities should use their discretionary powers. This means that judges must very often defer to other public authorities—to some extent—on the question of how to use a power. To defer to (e.g.) the Ministry on (e.g.) the question of whether injuries caused by military activity of warring factions should be excluded from the compensation scheme is to treat the fact that the Ministry decided that those injuries are to be excluded as a reason why they should be excluded.
Courts cannot do everything. They have a supervisory jurisdiction over public bodies such as the Ministry of Defence, but they are not a defence agency. They have a supervisory jurisdiction over the Higher Education Funding Council, but they are not a higher education funding agency. So, in R v Higher Education Funding Council, ex p Institute of Dental Surgery  1 WLR 242, the Court of Appeal refused to evaluate the worth of research in dentistry; it could not decide whether the Council’s funding decision was based on a sound evaluation, so it would not interfere with it. The Court deferred to the Funding Council on the question of how to evaluate dentistry research. This may seem tragic, because if the Funding Council misjudged the worth of the claimant’s research, the Court’s deference left the claimant with no remedy for an injustice. But it is not tragic: it just reflects the fact that courts cannot be expected to achieve perfect justice. The funding of higher education would not be better governed if the judges tried to do so.
How far should the courts defer to an administrative decision? Sometimes very radically, sometimes not at all, and more or less depending on the type of decision, the body that is making it, and the context in which it is made. This feature of judicial review explains why courts will quash administrative decisions for certain forms of unreasonableness, but not for unreasonableness in general. Much of Chapters 7 and 8 are concerned with the considerations requiring greater or lesser deference by courts to the judgment of public authorities whose decisions they are reviewing.
2.5.3 Isn’t it Parliament’s job to control the executive?
There is a myth in modern British constitutional lore: that the courts have had to control the government because Parliament has been failing to do so. According to this myth, the electoral system tends to create strong majorities, and the party whips control the backbenchers through threats and promises, with the result that the House of Commons is under the control of the executive, instead of the executive being under the control of Parliament. Judicial review did not develop as a substitute for proper parliamentary control, and it can never act as a substitute because:
judicial review is ancient, much older than the tradition of parliamentary supervision of administration; and
The roots of the courts’ authority to control administration are medieval. And by the seventeenth century, the courts were asserting that authority in a very creative way. It was only after the seventeenth century that Parliament was able to exercise any reliable and creative control over the government. It is certainly more than 300 years since English judges doubted their power to supervise the legality of executive action (although they have dramatically revised their views as to what makes executive action unlawful).
From the Mists of Time
Nearly 400 years ago, the Chief Justice, Sir Edward Coke, could claim that ‘to this Court of King’s Bench belongs authority, not only to correct errors in judicial proceedings, but other errors and misdemeanors extra-judicial, tending to the breach of peace, or oppression of subjects, or to the raising of faction, controversy, debate or any manner of misgovernment’ (Bagg’s Case (1615) 11 Co Rep 94 (Coke CJ)). That was putting it too highly, because it has never been the judges’ job to correct ‘any manner of misgovernment’; to do that, they would have to decide (e.g.) the value of research on dentistry for themselves, instead of leaving that to the Higher Education Funding Council. Coke’s statement has to be read as subject to the requirements of comity.
Lord Chief Justice Holt put it less extravagantly in Groenvelt v Burwell (1700) 1 Ld Raym 454, holding that ‘by the common law’, ‘this court will examine if other courts [including statutory authorities such as, in the case, the censors of the College of Physicians of London] exceed their jurisdiction’.
The judges who developed judicial review were not making up for Parliament’s failings. Judges and MPs have different capacities and different opportunities to supervise the executive. Judges have special competence for controlling certain forms of misgovernment (unlawful uses of power), and Parliament has special competence for controlling other forms of misgovernment (the pursuit of policies that are contrary to the national interest).
Parliament has control over administration, first, because the Prime Minister is the leader of the party that has the confidence of the House of Commons, and he or she needs to conduct the administration in a way that helps to sustain that confidence, and to sustain political support for his or her party. Debates in the House of Commons enable the opposition parties to fight the next election from day to day throughout the lifetime of a Parliament through constant criticism of the government’s administrative decisions.
More specifically, too, the practice of Parliament allows MPs to ask questions of ministers concerning administration in departments and executive agencies for which they are directly or indirectly responsible. The possibility of asking embarrassing questions in the House of Commons supports a practice of written questions from p. 60↵MPs to departments, which helps the MP to get a response to particular complaints on behalf of constituents. And since the 1960s, MPs have been able to refer complaints to the Parliamentary Ombudsman (see Chapter 13). And the work of the Parliamentary Ombudsman is supported by the Public Administration Select Committee of the House of Commons, which receives reports from the Ombudsman and makes recommendations to the House.
Apart from the role of individual MPs in responding to complaints, since the late 1970s select committees of MPs have had the task of overseeing the administration of departments. There are now eighteen departmental select committees.6 They are made up of backbench MPs, and they help Parliament to control the executive by examining policy, expenditure, and administration in each of the main government departments and their associated agencies. Even though each committee has a government majority, the committees provide a technique for general scrutiny of administration that is more effective than debates in the House of Commons itself. Select committees do for the House of Commons what it cannot do in a debate among all 650 MPs, by taking written and oral evidence. The committees can summon witnesses to give evidence or to produce documents. A select committee report can generate political pressure in the House of Commons, and the onus of responding to a report can itself change government policy.
Judicial review of the lawfulness of executive action is deeply different from those parliamentary techniques of oversight. If it is done right, it is consistent with parliamentary control of the government. The work of the select committees and debates in the House of Commons impose an important control on administration in the public interest (or at least they can do so, depending on political pressures), but they do not provide effective redress for the complaints of particular persons that an administrative decision was unlawful.
While courts can only interfere with the government on legal grounds, Parliament can hold the government to account on any ground. Courts only promote good government indirectly; Parliament can do so directly. Parliament can also insist on the rule of law. A complaint of unlawful administrative action could (like any complaint of bad government) be good material for questions in the House of Commons, or even grounds for a vote of no confidence. Or it could play a role in a parliamentary election campaign. But elections and debates in the House of Commons are not remotely as well designed as a claim in the High Court for responding to a particular person’s legal complaint.
Consider R v Bolton (1841) 1 QB 66 (see 1.5.6, p 25), in which Mr Bolton’s council threw him out of his house. He could have complained to his MP, who could have raised the matter in the House of Commons; the government could have taken executive action to help Mr Bolton in response to clamour in the House, or Parliament could have passed a statute retrospectively benefiting him and imposing new duties on the parish council (or new duties on the courts to control the parish council). All p. 61↵that is fanciful, though, as a way of dealing with a particular person’s complaint—especially if the person is, like Mr Bolton, a pauper who has been thrown out of a council house. When a particular personal problem ends up being discussed in Parliament, it is because of some particular media frenzy that may have nothing to do with the justice of the case. Parliament is not an effective forum for securing justice according to law for particular individuals. The courts, too, have access problems (the expense of legal services is the most significant). But once Mr Bolton is in court, he has an opportunity to make an argument to a person independent of government, and the public authority is put in the predicament of having to meet his argument with a lawful justification of its action. That arrangement is a necessary technique for imposing the rule of law on the executive, even if Parliament is doing its job well. Administrative law does not make up for any modern failure by Parliament to carry out its responsibility; Parliament never did the things that we need the courts to do.
And on the other hand, judicial review can never be a remedy for inadequate parliamentary control. Parliament controls the executive in the national interest, as a representative assembly. Legal institutions impose the rule of law on the executive, and hear particular cases. So if Parliament is not doing its job well, the legal institutions cannot solve that problem. Judges do not have any techniques to fill a constitutional vacuum left by spineless backbenchers, excessive party discipline, or a weak opposition.
2.6 The twentieth-century judicial adventure
The story so far is that the judges are prepared to pass judgment on other public authorities’ interpretation of the law, to impose due process on them, and to quash a decision that they can identify as an abuse of power. This dynamic pattern of judicial control of other agencies of government is compatible with the role of Parliament in the constitution, and it can potentially be justified by the constitutional principle of the rule of law. Whether it is actually justified always depends on whether the judges exercise their power with comity toward other public authorities, which requires the judges to defer to the initial decision maker, sometimes radically, on many questions that the constitution and the law commit to the initial decision maker. But if the judges find a lack of due process, or if they can identify a decision, from their rather isolated point of view in the court, as an abuse of power, they do not need to defer to another public authority’s opinion that it was the right thing to do.
In the development of their role, the judges repeatedly need to assess what deference is due to other public authorities on particular questions; they make a mistake if they think either that they must generally defer to administrative decision makers, or that they can generally decide for themselves the grounds on which administrative decision makers should act. It depends on the nature of the decision and the role and the capacities of the initial decision maker.
In the first half of the twentieth century, the courts developed a tendency towards general deference to administrative authorities. After the Second World War, the p. 62↵courts went through a very important period of expansion and new articulation of the grounds of judicial review that was so far-reaching that it has seemed to many lawyers as if administrative law was invented in the second half of the twentieth century. It should already be apparent that many of the most dramatic judicial inventions are centuries old. But the developments in the decades after the Second World War were remarkable, and even a partial list will give you a picture of the capacity of the courts to develop new rules of administrative law.
Timeline of the twentieth-century judicial adventure
Conway v Rimmer  AC 910: instead of having a ‘Crown privilege’ against disclosing documents in litigation, the government must ask a court to decide whether public documents should be withheld from a party in litigation for reasons of the public interest (‘public interest immunity’—see 1.6.1, p 32).
Council of Civil Service Unions v Minister for the Civil Service  AC 374: the three general grounds of judicial review apply to an exercise of prerogative power.
M v Home Office  3 WLR 433: the courts can award injunctions against ministers of the Crown, and declare them to have acted in contempt of court.
R v Home Secretary, ex p Doody  1 AC 531: cases on disclosure and duties to give reasons have developed a general trend towards openness in the making of administrative decisions.
In the control of discretionary powers, there was a marked move away from general deference to administrative authorities towards deferring only for specific reasons. The Wednesbury principles can be traced back for centuries before the 1940s (see 7.1, p 240). But after the 1940s, the courts gave them further articulation, and in the p. 63↵course of doing so they extended them. That movement includes Padfield v Minister of Agriculture  AC 997 (see 8.2, p 294), and the prisoner cases discussed in 2.5.1, p 55. It also includes the following:
Inland Revenue Commissioners v Rossminster  AC 952: the House of Lords held that Liversidge v Anderson  AC 206 was wrongly decided (see Lord Diplock, Rossminster, at 1011). In Liversidge, a majority of the House of Lords had decided that the courts had to defer radically to the Home Secretary’s judgment as to whether there is good reason for a detention under a statutory power;
R v Home Secretary, ex p Simms  2 AC 115: the courts have articulated the principle of legality, which is a rule that powers conferred in general terms by a statute do not empower public authorities to act contrary to certain legally protected values.
The courts’ control of discretionary powers has also been extended since the 1940s through tort liabilities of public authorities (see Chapter 14). This development is complex and does not amount to a general doctrine that judges will award compensation whenever a claimant suffers loss because a public authority used its power unlawfully. The judges have (with notable exceptions) been very cautious not to turn causes of action in tort into tools by which claimants can get courts to decide what administrative decisions ought to have been taken in general (see 14.2, p 552).
From the Mists of Time
The judges sometimes suggest that judicial review was invented in the twentieth-century adventure. In the Page case, Lord Browne-Wilkinson referred to ‘the great development that has recently taken place in the law of judicial review whereby the courts have asserted a general jurisdiction to review the decisions of tribunals and inferior courts’ (R v Lord Chancellor, ex p Page  AC 682, 700). But that development was centuries old. The tradition of general judicial supervision of administrative agencies predated Keighley’s Case (1609) 10 Co Rep 139, which imposed judicial control over the Commissioners of Sewers, who had been given legal power by a statute that said nothing about judicial control. The Keighley decision implicitly asserted a general jurisdiction of the Court of King’s Bench (today, the High Court) to review the decisions of tribunals and inferior courts. Blackstone’s Commentaries treated that jurisdiction as a fundamental of the constitution in the 1760s (see 1.1.3, p 10 and 2.8.1, p 71).
In fact, English judicial history is unified across the centuries by excessive judicial claims of responsibility for controlling administrative decision making (see 7.1, and on why the claim in the Page case was excessive, see 9.1.6, p 341).
p. 64↵In addition to their role in developing the grounds of judicial review in the twentieth century, the judges led reforms of the process for seeking judicial review (see Chapter 10), and they developed a very liberal approach to standing (see Chapter 11). These judicial reforms have facilitated ‘public interest litigation’, turning judicial review into an instrument for political campaign groups to ask the courts to make decisions that they cannot persuade the government to make (see 11.2, p 438).
Much of this book is concerned with the achievements and the drawbacks of the judicial adventure of the twentieth century, and with the ways in which the judges are building on it in the twenty-first century.
2.7 Twenty-first-century frontiers
In this century, the creative role of the judges in administrative law has increased. When the Human Rights Act 1998 came into force in October 2000, it established an entirely new ground of judicial review of administrative action: ‘It is unlawful for a public authority to act in a way which is incompatible with a Convention right’ (s 6). The judges accepted Parliament’s invitation to remodel administrative law. Their new role under the Human Rights Act is explained in Chapter 3, and illustrated throughout the book.
The Human Rights Act is not the only way in which Parliament has augmented the role of the judges: the Freedom of Information Act 2000 generated a new administrative duty of openness; the judges’ creativity has enhanced that duty, and they have also added new common law duties of disclosure of information that were not mandated by the Act (see 1.6.1, p 32). The Equality Act 2010 consolidated the various pieces of anti-discrimination legislation since the 1970s, and has changed the practice of public administration (and generated much litigation) by imposing duties on policy makers to give special consideration to the position of groups that have been discriminated against (see 8.3.2, p 307).
In addition to those legislative upgrades in the role of judges in governing the UK, the Supreme Court itself is prepared to assume new powers when the justices conclude that they should give new legal effect to constitutional principles. In order to understand the attitudes of the senior judges towards the power of the executive branch of government and their attitudes to their own power, every student of administrative law needs to understand the Supreme Court’s decisions in Gina Miller’s two remarkable administrative law cases: R (Miller) v Secretary of State for Exiting the European Union  UKSC 5 (‘Miller No 1’) and R (Miller) v Prime Minister  UKSC 41 (‘Miller No 2’).
These two landmark decisions are so far the only cases to have been decided by all the judges of the Supreme Court sitting together. In Miller No 1, by an 8–3 majority, the Court accepted Ms Miller’s argument that, unless Parliament authorized her to do so by a statute, Prime Minister Theresa May could not lawfully begin the Brexit process by triggering Article 50 (the provision in the Treaty on European Union that provides p. 65↵for withdrawal of a member state). It was a revolutionary decision because the powers to make and to terminate treaties and to engage in processes established by treaties are core examples of prerogative power (that is, the Crown’s constitutional power to make decisions on behalf of the UK on the advice of ministers who are accountable to the House of Commons). The judges had never before interfered with the Prime Minister’s decisions on treaties.
In Miller No 2, all eleven Supreme Court Justices accepted Ms Miller’s argument that it was unlawful for Prime Minister Boris Johnson to advise the Queen to prorogue Parliament (that is, to end one session of Parliament and announce the date when a new session would commence) for five weeks during the two months before a deadline for Brexit. The Court held that the prorogation had, therefore, never happened, and Parliament resumed sitting the day after the Court’s decision. It was a revolutionary decision because the courts had never before played any role at all in regulating the operation of Parliament.
These two decisions arose in unprecedented conditions generated by the divisive politics of Brexit, after a slim majority of UK voters voted to leave the European Union in the 2016 referendum. No court had ever interfered with the Prime Minister’s decisions on treaties before Miller No 1, but then, no Prime Minister had ever been planning to take a treaty decision that would change the governmental structure of the UK through withdrawal from the EU. No court had ever interfered with a prorogation of Parliament before Miller No 2, but then, there had never been any similar situation in the House of Commons. The prorogation in September 2019 was an unprecedented gesture of scorn towards the House of Commons by a Prime Minister who could neither get approval of his approach to the negotiation of Brexit (he insisted that the UK should be prepared to leave without having to make a withdrawal agreement with the EU), nor agreement to his proposal to hold an early general election to resolve the issue.
One reason for the decision in Miller No 1 was that the majority interpreted the European Communities Act 1972 (which gave effect in UK law to EU law as established under the treaty) as taking away the power of the Prime Minister to terminate the treaty. Nothing in the 1972 Act supported that interpretation; the judges evidently reached it because they felt that the matter was too important for the executive branch of government:
‘We cannot accept that a major change to UK constitutional arrangements can be achieved by a ministers alone [sic]; it must be effected in the only way that the UK constitution recognises, namely by Parliamentary legislation. This conclusion appears to us to follow from the ordinary application of basic concepts of constitutional law to the present issue’ .
The majority did not identify those basic concepts. Where a significant change results from the cancellation of a treaty, there was no authority in the law and no warrant in constitutional principle for the Court’s conclusion that only parliamentary legislation can lawfully authorize it. Powerful dissenting reasons by Lord Reed, Lord Carnwath, p. 66↵and Lord Hughes demonstrated that there was no basis for the legal proposition that, if a treaty makes a major change to UK constitutional arrangements, the treaty cannot be cancelled without an act of Parliament. And this is where the case offers us an important insight into the judges’ attitude: eight out of the eleven justices shared a feeling that executive power is subordinate to Parliament not only in the sense that Parliament can change it and control it, but in the sense that it is up to the judges to decide when a decision is too important to be made without legislation.
Miller No 1 had no practical effect except to delay Theresa May’s plans for four weeks while Parliament passed the legislation that the Court required.7 Miller No 2 is a different matter. In Miller No 2, the unanimous Supreme Court accepted Ms Miller’s argument that by quashing the prorogation they would not be seizing a new power for themselves, but only protecting the centuries-old role of Parliament. The Court held that the prorogation was unlawful because it had ‘the effect of frustrating or preventing, without reasonable justification, the ability of Parliament to carry out its constitutional functions as a legislature and as the body responsible for the supervision of the executive’ (; see also ). No court had ever before treated the timetable of Parliament as a matter of law. The reasons of Lord Reed and Lady Hale for the unanimous Court presented this extraordinary innovation as something that was required by ancient principles of the constitution. And it has indeed been a principle of our constitution, since the Glorious Revolution of 1689, that Parliament must meet in such a way as to carry out its constitutional role. But never before September 2019 had there been any principle that the judges should decide when Parliament needs to meet, imposing their answer to that question not only on the Prime Minister, but also on Parliament itself (which could have reversed the prorogation, and did not do so8).
Miller No 2 raises a new question as to which matters of constitutional principle the judges will turn into matters of administrative law. The boundaries to their new role are unpredictable and will depend on the creativity of claimants like Ms Miller and their barristers, but also on the judges’ sense of how far it is legitimate for them to make themselves into referees of the political practice of the constitution. The courts will undoubtedly be circumspect at some points; their caution is demonstrated in their refusal of permission for an innovative claim for judicial review of Theresa May’s deal with the Democratic Unionist Party after the 2017 general election (McClean v First Secretary of State  EWHC 3174; see the discussion of McClean and other failed attempts to use judicial review to regulate politics in the ‘Menagerie of hopeless claims’ in 10.3.2, pp 404–5). The crucial implication of Miller No 2 is that the courts will decide for themselves which constitutional principles to turn into laws. Miller No 2, in particular, was a judicial initiative to convert the vague and flexible p. 67↵constraints imposed on the executive branch of government by constitutional convention and by constitutional politics into legal constraints (that is, into vague and flexible rules that the judges are empowered to identify and to enforce). That turns part of our constitutional politics into a sub-branch of administrative law. It is a very significant change in English administrative law, and its implications are waiting to be worked out through the judges’ response to claimants who will undoubtedly argue in future that the constitutional propriety of other aspects of the Prime Minister’s conduct is a matter for judges to decide, and should not be left to the decisions of the MPs in the House of Commons and the electors in general elections.
But we should not exaggerate the role of judges. You will learn from the rest of this book that the judges do not accept every argument that creative barristers offer in favour of increasing judicial control over government. In recent years, they have resisted strongly argued invitations from claimants to make each of the following into a general ground of judicial review of administrative decisions:
In the twentieth and twenty-first centuries, judges have undoubtedly assumed powers that they had never exercised before to interfere with other public authorities. But they do not govern the country directly. The first reason for this is that they only shape the government of the UK by passing judgment on the lawfulness of the conduct of other governmental authorities, and they do not do so on their own initiative, but by deciding cases brought to them by claimants. In the two Miller cases, the judges did not set out to change the constitution, or to subject the constitutional practice of the government and Parliament to their own control. They were persuaded to do those things by the arguments of Gina Miller’s lawyers, after hearing argument to the contrary from the government’s lawyers.
The second reason that we should be careful not to exaggerate the role of judges is that courts are not the primary forum for complaints against administrative decisions. We can only understand the place of judicial review in administrative law if we understand its limits.
Other techniques for controlling administrative decisions are much more important than judicial review to millions of people. Most important of all are discussion and negotiation with public authorities themselves over their decisions and policies and plans. And, during the same period in which the judges have reinvigorated judicial review, Parliament has enhanced old techniques and instituted new techniques for resolution of disputes and for supervision of administration, which have restructured the law and practice of administration:
the system of supervision of administration by parliamentary select committees in the House of Commons has been enhanced; and
audits scrutinizing the economy, efficiency, and effectiveness of central government spending are carried out by the National Audit Office, and the Local Audit and Accountability Act 2014 imposed a privatized scheme of audits for local authorities (see 15.4.6, pp 623–4).
Ombudsmen and auditors have capacity for forms of investigation that courts and tribunals cannot carry out. Ombudsmen, unlike courts or tribunals, can uphold a complaint even where there has been no unlawful decision. Their reports can get results for a complainant that are more flexible than judicial remedies, and they are better equipped than courts or tribunals to recommend general improvements in administrative practice. Auditors have a systematic role that judges could never fulfil, in holding government departments and other public bodies to account for their finances. Parliamentary committees have a role in representative politics that enables them to hold public administration to account for matters of policy on which judges would need to defer. Judges are restricted to identifying unlawful procedures. Ombudsmen, parliamentary committees, and auditors can recommend better procedures.
As for adjudication of legal claims, far more cases are resolved in tribunals than in courts (see Figure 12.1, p 469). And if a claim reaches a court, it is far more likely to do so in a statutory appeal from a tribunal or a statutory appeal from a government decision9 than in a claim for judicial review. Tribunals can operate more informally than courts, and their judges may have specialized expertise that High Court judges do not have.
p. 69↵For these reasons, and others to be explained in Chapters 12, 13, 14, and 15, other forms of recourse are much more important, for people who have a complaint about government, than judicial review.
The alternatives to judicial review are not merely used more often than judicial review; their role is essential to the principles of the subject. You cannot understand judicial review itself if you think it is the primary form of legal control of administrative action. It is not even the secondary technique. It is a last resort. As Lord Scarman put it: ‘[A] remedy by way of judicial review is not to be made available where an alternative remedy exists. This is a proposition of great importance. Judicial review is a collateral challenge: it is not an appeal’ (R v Inland Revenue Commissioners, ex p Preston  AC 835, 852). In the Preston case, the House of Lords held that judicial review should not generally be used when a statutory appeal process is available. Today, that principle applies generally to the new tribunal system (see Chapter 12), so that permission to bring a claim for judicial review will ordinarily be refused when the claimant could appeal to a First-Tier Tribunal (R (Boulting) v HMRC  EWHC 2207), or where the defendant’s decision could be referred to the Upper Tribunal (R (Willford) v Financial Services Authority  EWCA Civ 677). The same principle applies if the public authority has an internal complaints process (R (S) v Hampshire County Council  EWHC 2537).
The courts can refuse to give judicial review even if the alternative way of pursuing a complaint does not do everything that judicial review could do. In St George’s v Rafique-Aldawery  EWCA Civ 2520, the Court of Appeal held that because judicial review is a last resort, students whose university programmes were terminated could not seek judicial review of the terminations until after pursuing complaints through the Office of the Independent Adjudicator (a higher education ombudsman service –see 13.9.2, p 530). The Court held that alternative complaint processes have to be pursued instead of judicial review even if the grounds of decision and the remedies available are not the same as in judicial review .
This approach to judicial review has survived the judicial adventure of the twentieth century, and in fact it has been taken further, since the Civil Procedure Rules 1998 gave the courts more flexible authority and encouragement to control and to reduce litigation. In R (Cowl) v Plymouth City Council  EWCA Civ 1935  and , Lord Woolf, whose report on civil procedure led to the 1998 reforms, emphasized the ‘paramount importance’ of avoiding litigation through alternative dispute resolution, and warned against ‘over-judicialising’ complaints against public authorities. The Court of Appeal refused permission for judicial review, told the parties to negotiate about the future instead of arguing about the past, and included the resulting agreement between the parties in its final decision. That important decision simply rejects the idea that all complaints of unlawful conduct are to be heard by a court. For all the importance of judicial review, it remains a last resort.
It may seem tempting to think that, because the rule of law is a fundamental constitutional principle, the judges must be prepared to pass judgment on any allegation of official unlawfulness, so that the government will be held to the law. As surprising as it may seem, that would be a serious mistake. Whether judicial review should be allowed p. 70↵in any case is itself a question of due process. If there is another way in which to resolve a dispute with the government, due process does not require judicial review.
From the Mists of Time
In R v Barker (1762) 3 Burr 1265, 1267 (see 11.1.1, p 430), Lord Mansfield held that: ‘A mandamus … ought to be used upon all occasions where the law has established no specific remedy, and where in justice and good government there ought to be one.’ Note the implication: a mandamus (today, a mandatory order requiring a public authority to exercise a legal power) will not be issued unless justice and good government require the order to be made.
Chapter 10 addresses the judicial review process. For now, it is worth outlining some of the restrictions on judicial review, to put it in its place as an extraordinary process.
Restrictions on judicial review
No one has a right to seek judicial review; a claimant must first ask the court for permission.
An unlawful process is not in itself a reason for quashing a decision. So, for example, if a hearing does not meet lawful requirements, the decision will not be quashed if the irregularity was not substantial enough to make it unfair (NJ v Essex County Council  EWCA Civ 545). And under the Criminal Justice and Courts Act 2015 s 84, unless there is a reason of exceptional public interest, courts must refuse relief in a claim for judicial review ‘if it appears to the court to be highly likely that the outcome for the applicant would not have been substantially different if the conduct complained of had not occurred’ (see 10.4.7, p 419).
Power to remit: if the judges find that a decision was unlawful, they will not necessarily order the outcome the claimant seeks. The courts have very wide power under the Civil Procedure Rules to send the matter back to the initial decision maker for it to make a lawful decision.
p. 71↵In spite of these basic restrictions on judicial review, the government has sometimes wanted to set up bodies whose decisions cannot be challenged in the courts. But when Parliament has enacted provisions designed for that purpose (called ‘privative clauses’ or ‘ouster clauses’), the courts have offered no cooperation at all.
2.8.1 How to interpret an ouster clause
In 2016, Privacy International, an advocacy group that opposes government surveillance, complained to the Investigatory Powers Tribunal that one of the UK’s intelligence services (Government Communications Headquarters or ‘GCHQ’) had engaged in unlawful computer hacking. The hacking was authorized by the Home Secretary, but Privacy International argued that the Home Secretary’s statutory power under the Regulation of Investigatory Powers Act 2000 was limited to authorizing GCHQ to hack one particular property (such as a particular mobile phone) at a time, and that she had been unlawfully issuing ‘thematic’ warrants allowing GCHQ to hack whole classes of devices (such as all the mobile phones in London). The Tribunal rejected the claimant’s interpretation of the legislation. Privacy International brought a claim for judicial review in the High Court to ask the judges to overrule the Tribunal’s interpretation of the legislation.
But the Regulation of Investigatory Powers Act provided that ‘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’ (s 67(8)). Privacy International argued that this ‘ouster clause’ should not be interpreted to oust the High Court’s jurisdiction to hear the claim, and that if the clause had to be interpreted to oust the Court’s jurisdiction, then the clause was invalid because Parliament had no power to prevent the Court from hearing the claim. In a deeply divided Supreme Court decision, four of seven justices held that the ouster clause should not be interpreted to oust the High Court’s jurisdiction; three of those four (Lord Carnwath, Lady Hale, and Lord Kerr) held that Parliament had, in any case, no power to oust the High Court’s jurisdiction: R (Privacy International) v Investigatory Powers Tribunal  UKSC 22 (on the question of Parliament’s power, see 1.5.1, p 16).
How could the Supreme Court interpret a statute as allowing the Tribunal’s decision to be questioned in a court, if the statute said that decisions of the Tribunal shall not be questioned in a court? To understand this, you need to be familiar with the famous case of Anisminic v Foreign Compensation Commission  2 AC 147 (HL). In Anisminic, Parliament had set up a commission to decide on applications for awards from a fund for compensation of British companies that had lost property in Egypt as a result of the Suez Crisis. The statute provided that:
‘The determination by the Commission of any application made to them under this Act shall not be called in question in any court of law.’10
p. 72↵Yet the House of Lords, in another deeply divided decision, quashed the Commission’s determination of an application made to them under that Act. The majority’s ground was that the Commission had misinterpreted the legislation in such a way that its decision did not count as a genuine determination for the purpose of the statute. The Court was not allowed to call a determination into question, so the majority in the House of Lords decided that it wasn’t a determination!
The implications of Anisminic will need close examination in Chapter 9, which explains the role that the judges play when a claimant argues that an administrative agency misinterpreted the law (see 9.1, p 333). We will see that the law on this matter morphed after the Anisminic decision into a general doctrine that judges will not defer to administrative authorities in the interpretation of the legislation that they administer. Anisminic did not establish that doctrine, but you may say that it paved the way: it was a paradigm example of a tendency among some judges to exaggerate the importance of judicial review. If the House of Lords had given proper effect to the ouster clause in Anisminic, the result in that case would have been that foreign compensation awards would have been made in accordance with the Commission’s interpretation of the legislation (an interpretation with which one of the Law Lords agreed), rather than in accordance with the interpretation favoured by the majority of the judges. It would still have been possible for courts to interfere—in spite of the ouster clause—with an abuse of power by the Commission (see 9.1.4, p 338). Likewise in Privacy International, if the decision had gone the other way, the interpretation of the Regulation of Investigatory Powers Act would have been decided by the independent judges of the Investigatory Powers Tribunal, rather than by the independent judges of the High Court, the Court of Appeal, and the Supreme Court. We do not actually have any general ground to think that judicial review for errors in interpreting the Act will improve decision making in complaints against government computer hacking.
The radically sceptical approach that judges take to ouster clauses is incompatible with parliamentary sovereignty (see 1.5.1, p 16). But it has a long tradition. William Blackstone wrote in the 1760s that because administrative tribunals derogate ‘from the general jurisdiction of the courts of common law’, their powers ‘cannot be extended farther than the express letter of their privileges will most explicitly warrant’.11 One way in which things have changed since the 1760s is that it would now be a mistake to think of tribunals as derogating from the common law courts’ jurisdiction. When a new administrative scheme is created (e.g. to award compensation from a foreign compensation fund, or to authorize computer hacking to fight terrorism, or to award criminal injuries compensation to soldiers), Parliament is taking nothing away from the common law courts if it creates a new tribunal to operate the scheme, or to resolve disputes over its operation. We will see in Chapter 12 that one central feature of the tribunals system is that finally, after centuries, English law has escaped from the presumption that the ordinary courts are the default forum to determine all legal disputes. Administrative tribunals do not necessarily detract from the role of the p. 73↵High Court, and the role that the High Court can play in controlling administrative tribunals should be worked out by asking what, if anything, the addition of another judicial process can do to improve decision making. It should not be worked out by presupposing that the rule of law requires the judges of the High Court to impose their own interpretation of the Regulation of Investigatory Powers Act on the Investigatory Powers Tribunal, when Parliament has provided that the Tribunal’s decisions should not be questioned in a court.12
The High Court does have a general responsibility for restraining abuse of power, to the extent that it can lawfully do so. The courts should do whatever they can lawfully do to prevent other public authorities from turning an ouster clause into a cloak for abuse of power. If Parliament has enacted that decisions of a tribunal are not to be questioned in court, the courts certainly should be vigilant to establish that the ouster clause does not protect abuses of power or mere sham decisions against judicial review. But there is no justification for judges to quash a decision of the tribunal merely on the ground that the judges of another tribunal (the High Court) take a different interpretation of the rules that it must apply. Yet we will see in Chapter 9 that the law of English judicial review requires judges to do just that. In developing that rule, the English judges have gone beyond their responsibility to impose the rule of law and have imposed rule by the judges of the High Court.
And there is one rule-of-law consideration that always weighs against judicial review: the interest in the finality of proceedings (see 7.1.1, p 244). The role of courts is, on the one hand, absolutely essential for the rule of law, since we need an independent institution that will impose the law. And yet, there is a tension in the ideal of the rule of law, because litigation suspends the legal effect of decisions and reduces legal certainty (see 1.5.5, p 23). The availability of litigation over the lawfulness of an executive decision always leaves matters up in the air and makes it impossible, for a time, to take an official decision at face value. And litigation often creates incidental risks that the relevant information will no longer be available, or will be misunderstood, in the judicial hearing. As the European Court of Human Rights held in Pullar v United Kingdom (1996) 22 EHRR 391 , it is ‘an important element of the rule of law’ that ‘the verdicts of a tribunal should be final and binding unless set aside by a superior court on the basis of irregularity or unfairness’.13 In cases of ‘irregularity or unfairness’, as the Strasbourg Court called it, judicial review is essential to the rule of law, in spite of the drawbacks of litigation. But to justify judicial review, we cannot simply say that the rule of law must subject everything to the rule of courts. We need to find special reasons of constitutional principle that make the litigation a proportionate response to the claim.
The core rationale for judicial review may seem to be that if an administrative action is unlawful, courts must step in to impose the law. That is an easy mistake to make, because judges are absolutely essential to justice and the rule of law—not just judges, but independent, creative judges. The judges’ capacity is limited, however, and so their constitutional role is limited too. They ‘cannot begin to evaluate the comparative worth of research in clinical dentistry’ (R v Higher Education Funding Council, ex p Institute of Dental Surgery  1 WLR 242), for example. Where fidelity to the law requires an administrative authority to evaluate research in clinical dentistry, the judges are not well equipped to improve the authority’s fidelity to law. The core justification for judicial review of the conduct of another public authority should be constrained by the principle of comity.
The core rationale for judicial review
Where there is no other process for imposing the rule of law, independent judges should support the rule of law by imposing due process on administrative authorities, and by opposing arbitrary decision making. But this role is limited by the principle of comity, which requires judges not to interfere with another public authority unless they can do so effectively and without damaging that other public authority’s capacity to do its own job.
Control of administrative action by law goes well beyond the core rationale for judicial review. For example, the law sets up ombudsmen to improve administration directly, and it establishes tribunals with expert membership to resolve disputes with public authorities, and auditors to improve economic efficiency in administration. But the core rationale for judicial review identifies a central set of tasks for judges—tasks that they are justified in taking on themselves (as they took upon themselves the habeas corpus jurisdiction in the Middle Ages; see 1.1.1, p 5). Within the core rationale, it is right for the courts themselves to have assumed and developed their own role in holding other public authorities to account. As Lord Mansfield said in R v Barker, ‘The interposing [of judicial review] where there is no other specific remedy, is greatly for the benefit of the subject and the advancement of justice’ ((1762) 3 Burr 1265, 1266; see 11.1.1, p 430). If the courts leave behind the limited role that is supported by that rationale, they may or may not make better decisions than the administrators, but they will not be imposing the rule of law.
Every aspect of administrative law involves processes by which one public authority exercises some form of legal control over the conduct or decisions of another. The central problems for administrative law are as follows.
What processes and remedies ought to be available for the control of one public authority’s decision by another public authority?
How can that interference be reconciled with the comity that one public authority owes towards another?
All of these problems are addressed (articulately, with remarkable results) in the law of judicial review, and that is why Chapters 4–9 unpack the grounds of judicial review outlined in this chapter, and Chapters 10 and 11 explain judicial review processes and remedies.
Take-Home Message • • •
Respect for parliamentary sovereignty requires judges to interfere with other public authorities on any grounds on which Parliament enacts that they should do so.
Respect for the rule of law justifies judges in imposing due process on other public authorities in their decision making (Chapter 4). Because the rule of law is opposed to the arbitrary use of power, the rule of law also justifies judges in developing a doctrine of review for abuse of power (Chapter 7).
The rule of law also requires judges to control the way in which administrative authorities interpret the law. But the judges’ duty to impose the rule of law on the administration does not justify the doctrine of review for error of law (see Chapter 9).
Where judges are able to identify a decision as unreasonable while acting with respect for the decision-making advantages of the initial decision maker (that is, with comity), there is ground for judicial review. But the judges have no general jurisdiction to decide what would be reasonable.
Comity requires judges to supervise the conduct of other public authorities in a way that shows respect for those authorities’ role, and that reflects the reasons why the power to make those decisions was given to a body other than the court.
Critical Questions • • •
What is the purpose of judicial review: is it to police the lawfulness of administrative action? To right injustices to claimants?
Can you explain the difference between review for error of law and review for the improper use of discretionary power?
Can you find examples of judges explicitly basing their decisions in judicial review on the rule of law?
Do judges play a political role in judicial review? Should they do so?
Can you think of any grounds on which a court ought to be prepared to review a decision by the Prime Minister to recommend the appointment of a particular person as a minister of the Crown?
Would judicial review become unnecessary if Parliament were better at scrutinizing the executive?
Associated Provincial Picture Houses Ltd v Wednesbury Corporation  1 KB 223 (CA).
R v Ministry of Defence, ex p Walker  1 WLR 806 (HL)
On Miller 2:
Note that Professor Paul Craig anticipated the Supreme Court’s holding that judicial supervision of a prorogation is required to protect the legal principle of parliamentary sovereignty: Paul Craig, ‘Prorogation: Constitutional Principle and Law, Fact and Causation’ UK Constitutional Law Blog (2 September 2019), https://ukconstitutionallaw.org/2019/09/02/paul-craig-prorogation-constitutional-principle-and-law-fact-and-causation/.
1 The citation of cases is abbreviated in this book by omitting ‘on the application of’.
3 Julius v Bishop of Oxford (1880) LR 5 AC 214, 235 (Lord Selborne).
5 To carry out the UK’s treaty obligation to comply with declarations of the European Court of Human Rights, the Home Secretary immediately announced that he would exercise his statutory power in Venables’ and Thompson’s cases and in future cases on the basis of the recommendation of the Lord Chief Justice.
7 The European Union (Notification of Withdrawal) Act 2017, one of the shortest acts on the statute books, provided that ‘[t]he Prime Minister may notify, under Article 50(2) of the Treaty on European Union, the United Kingdom’s intention to withdraw from the EU’ (s 1(1)).
8 See Timothy Endicott, ‘Making Constitutional Principles into Laws’ (2020) 136 LQR 175.
9 To give one important example, the Housing Act 1996 s 204 gives a right of appeal to a county court on a point of law from an internal review of decisions as to what housing is suitable for homeless people. See Runa Begum v Tower Hamlets LBC  UKHL 5.
10 Foreign Compensation Act 1950 s 4(4).
11 Commentaries on the Laws of England, 1769 (18th edn, S Sweet 1829) Book iii, ch 6.
12 Note the strategy in the tribunals system, which is—wisely—not to try to insulate the tribunals system from the ordinary courts by means of an ouster clause, but to allow statutory appeals to the Court of Appeal (see 12.4.9, p 485).
13 Cited in Montgomery v HM Advocate  1 AC 641 (PC) 669. The comment concerned a judicial tribunal, but the same reasoning applies to administrative decision makers.