I. Introduction
Surely nothing can be more frustrating to a new student of administrative law than judicial review for errors of law. The student will be taught that, whenever an administrative decision-maker misinterprets or misapplies a statue or some other legal provision, the question of whether that misinterpretation or misapplication causes a decision to be invalid depends on whether the alleged error of law that the decision-maker has committed is considered “jurisdictional”. The distinction between jurisdictional and non-jurisdictional errors of law, the student is told, is the principal gateway to judicial review. Having learnt this, the student might, quite naturally, seek some knowledge as to which errors fall into which category. But he or she will be told not to worry about all this: that from the decision of the House of Lords in Anisminic in the late 1960s,Footnote 1 or at the very latest from the decision in Page in the early 1990s,Footnote 2 practically all errors of law have been considered jurisdictional, save for a few minor exceptions that no one really needs to worry about.Footnote 3 In reality, all errors of law now go to jurisdiction, and thus all errors of law can lead to a decision being quashed. Nonetheless, it will soon become apparent to the student that this apparently simple “orthodoxy” continues to be hotly contested. The Supreme Court recently had occasion to revisit it in R. (Cart) v Upper Tribunal,Footnote 4 and administrative law scholars as prominent as David FeldmanFootnote 5 and Christopher ForsythFootnote 6 continue to express doubts over its continued application. Indeed, ever since Anisminic and Page, scholars have felt that English administrative law's approach to error of law review does not allow sufficient room for administrators’ autonomy: if any question of law is ultimately one for the courts to resolve, what room is there, it is asked, for administrative decision-makers independently to administer the statutory schemes that have been entrusted to them by Parliament? Many scholars have called for a more nuanced approach to error of law review, where it is recognised that some errors of law might not lead to a decision's complete and utter invalidity.Footnote 7 When the student innocently asks why the pre-Anisminic approach to error of law review might not be returned to, however, the administrative law community guffaws. “The old law used to split hairs in its application of this elusive, esoteric distinction between jurisdictional and non-jurisdictional errors”, the student will be told. “Yes, of course there are some concerns about how the law applies today, but nothing can justify our returning to the wilderness of single instances that Anisminic led us out of.” And so the student is left to wrestle with the current state of the law that seems far from satisfactory, on the assumption that no other approach can be taken: legal history, says the collective narrative of administrative law scholarship, shows as much.
The problem with this collective narrative is that it is not really based on any legal history at all. Administrative law scholars and practitioners continue to labour under the belief that, before Anisminic, the distinction between jurisdictional and non-jurisdictional errors of law was applied in so haphazard and unreasoned a way that it was, in truth, meaningless: the administrative law equivalent of the emperor's new clothes – the legal doctrine that, in practice, did not exist. Yet, when it comes to a detailed analysis of this history, no such thing can be found. Analyses of administrative law in the days before Anisminic have, broadly speaking, been forgotten. Judicial review, it is commonly thought, is a modern phenomenon – one the development of which started slowly at the beginning of the twentieth century, reaching its stride in the later part of the 1960s with cases like Anisminic, Padfield,Footnote 8 and Ridge v Baldwin,Footnote 9 and attaining its real rationalisation in the GCHQ case,Footnote 10 increasing almost exponentially ever since. This view of administrative law is testament to an ahistorical Zeitgeist that continues to dominate English public law.Footnote 11 This ahistoricism is unfortunate, to say the least: without a proper understanding of administrative law's history, we run the risk of not fully understanding how and why administrative law operates today, and how it should be made to operate in the future. We run the risk of not being able to give today's administrative law student a satisfactory explanation of why this law is – indeed, for many, why it has to be – this way.
The aim of this article is to address this deficiency, by plumbing the depths of administrative law's history to allow for a fuller understanding of the modern approach to error of law review. English administrative law has a history going back to the thirteenth century,Footnote 12 and it is unrealistic to explore the history of English administrative law in its fullness. Instead, we will concentrate on what is probably the most important error of law case in the modern history of judicial review: the 1841 decision of the Court of Queen's Bench in R. v Bolton. Footnote 13 The decision was, on the face of it, rather mundane: a simple dispute between parish officials and an alleged pauper over the latter's occupation of a parish house. But, as we will see below, from such a prosaic dispute emerged one of the most important developments in English administrative law – one that radically altered the court's approach to error of law review and which coloured the following 127 years to the decision in Anisminic and beyond. The Queen's Bench in Bolton set out a very limited understanding of jurisdiction, limiting it to a “threshold” concept that was unrelated to the substance of an administrative inquiry. “Jurisdictional” errors of law were those that conditioned a decision-maker's ability to begin an administrative inquiry, but were not errors committed during the course, or at the end, of that inquiry. As such, administrators were given a high degree of autonomy: seldom could their decisions be challenged for jurisdictional error of law.
In this article, Bolton will be used as an anchor to study the history of error of law review in England. In doing so, we will be able to get a fuller idea of the approach to error of law review that the courts adopted before Bolton, as well as a sense of how the Bolton principle developed in subsequent years. This will give us a fuller understanding of the legal context from which Anisminic or Page represent a departure, and will show what the potential contours of administrative law might be if we were to revisit the development of error of law review that took place in the latter part of the twentieth century. It is hoped that this in-depth study will cause us to question the idea, first put forward by D.M. Gordon, that pre-Anisminic administrative law was a “wilderness of single instances”.Footnote 14 And, by questioning this view, it is hoped that some of the current problems surrounding error of law review might be seen in a new light. R. v Bolton is one of the most important cases in English administrative law, and has so far escaped close scrutiny.Footnote 15 It is hoped this article will help correct this deficiency.
II. Judicial Review Remedies
Before we consider the decision in R. v Bolton in any depth, we need some idea of the workings of judicial review at the time the case was decided, and the remedies through which judicial review was employed. Broadly speaking, judicial review of administrative decisions by the superior courts of law could be effected in three ways. The first was by means of the prerogative writs of certiorari, mandamus, prohibition, and habeas corpus. These were special remedies granted mainly by the Court of King's Bench to review the decisions of lower courts and other administrative decision-makers. Second, administrative decisions made or confirmed by inferior courts, mainly justices of the peace sitting at quarter sessions, could be challenged by way of a statutory appeal to the King's Bench. Third, administrative law proceedings could be challenged collaterally in tort proceedings. More will be said about statutory appeals and collateral challenges in tort later in this article, but for now we will concentrate on the prerogative writs. If we look back at the nineteenth-century history of administrative law, it will be seen that the main prerogative writ by which judicial review was effected was certiorari,Footnote 16 and so certiorari will be our main focus.
Until the middle of the nineteenth century, certiorari issued mainly against justices of the peace, who occupied a combined role of summary judge and county administrator with extensive statutory powers. For most of its history, certiorari was mainly sought on the ground of some error of law being disclosed on the face of the record of the decision being challenged. The record would be sent by the justices of the peace to the Court of King's Bench, sitting in Westminster Hall, so that it might be reviewed and, if any error of law was disclosed on the record's face, it would be quashed. No distinction was drawn between those errors of law that were jurisdictional and those that were not: in principle, any error could lead to a decision being quashed. In effect, the writ functioned more like an appeal than judicial review: certiorari proceedings were concerned with any error committed by a public decision-maker, whether that error related to the merits of the decision (usually the province of an appeal, which was only available when expressly allowed by statute) or the legal powers of the decision-maker to make the decision (the province of judicial review).
Although this sort of review for “patent” errors of law served as a useful way of keeping justices of the peace and other administrative decision-makers in check, often decision-makers would make decisions in areas over which they had no power to act. Perhaps equally often, especially in the context of justices’ administrative orders (as opposed to criminal convictions), in which context records were much less detailed,Footnote 17 the record of their decision would hide this problem: often it could not be discovered, from the face of the record alone, that the decision-maker had no jurisdiction over the case at hand. The constitutional principle of the rule of law – in its conservative sense of government according to the law – required decisions made without jurisdiction to be quashed, however, and so, from the middle of the eighteenth century, the King's Bench began to accept affidavit evidence, external to the record, to prove that the decision was one over which no statutory powers had been given to the decision-maker.Footnote 18 This is why a distinction came to be drawn between jurisdictional and non-jurisdictional errors of law: absent any error of law disclosed on the face of the record, a decision could only be quashed if affidavit evidence could be introduced, and this could only be done if the error was alleged to be jurisdictional.Footnote 19
Over time, the distinction came to be crucial in terms of delimiting the scope of judicial review. As will be examined below, as the nineteenth century progressed, the formal records of justices of the peace came to be less and less detailed, and thus judicial review for patent errors of law diminished in its importance. If an error of law were to be reviewed, such review had to be on the basis of affidavit evidence, and so the error of law had to be jurisdictional. The distinction between jurisdictional and non-jurisdictional errors of law thus became the gatekeeper of judicial review: only if a court could be persuaded that an error committed by a decision-maker related to his or her jurisdiction could review take place. Applicants for judicial review naturally came to exploit the distinction between jurisdictional and non-jurisdictional errors to procure the writ in particular cases. The more malleable the distinction, the easier this would be. It is this distinction between jurisdictional and non-jurisdictional errors of law, created in a particular historical context for what might be seen as particular historical reasons, which, even after the decision of the House of Lords in Anisminic, continues to have a profound influence on the law.Footnote 20
Before moving on to this later history of certiorari review, however, we need to look more closely at R. v Bolton. We will begin by considering the facts of the case.
III. The Facts of R. v Bolton
Sometime around 1820, James Bolton, a gardener,Footnote 21 took possession of a cottage in a small hamlet called Hampton Wick, located just outside Teddington in Middlesex. For some time, the cottage had been put to use by the parish to house paupers entitled to poor-law relief.Footnote 22 It seems, though, that Mr. Bolton was an exception to this: he paid rent to the parish and therefore lived in the house as a tenant at law. During his tenancy, however, things seem to have gone awry. According to one version of events, Mr. Bolton was imprisoned for smuggling and, for two years, his family, with little means of support, became chargeable to the parish under the poor law. As part of their relief, however, they were allowed to stay in the house without paying rent. Mr. Bolton, on his release, also became chargeable to the parish, and rejoined his family in their rent-free home.Footnote 23
Mr. Bolton and his family were all allowed to stay in the house until 19 February 1838, when the churchwardens and overseers of the parish served notice seeking dispossession of the cottage.Footnote 24 When Mr. Bolton and his family failed to give up possession, on 22 March 1838 a formal information was submitted to one of the justices of the peace for the county, Sir Andrew Halliday.Footnote 25 A hearing was held before justices of the peace at the Red Lion Inn in Hampton eight days later, with Mr. Bolton in attendance. The justices found that Mr. Bolton was occupying the house as a pauper, and had failed to leave the premises within one month of notice being served. The magistrates therefore issued an order requiring the chief constable of the hundred of Spelthorne and the petty constable of Hampton Wick to take possession of the cottage,Footnote 26 purportedly acting under s. 24 of the Poor Relief Act 1819.Footnote 27 For unknown reasons, the order was not enforced immediately, and Mr. Bolton and his family went on living in the cottage for nearly another year.Footnote 28 On 2 February 1839, however, the overseers and churchwardens sent a notice to Mr. Bolton seeking possession of the cottage once againFootnote 29 and, on 22 March 1839, an information against Mr. Bolton was submitted to James Morgan Strachan, another justice of the peace, who summoned Mr. Bolton to appear before him on 30 March 1839 at the Red Lion.Footnote 30 Mr. Bolton appeared before the justices on that date. After considering the issues fully, James Morgan Strachan and another magistrate, John Kirkland, made a second order seeking possession of the cottage.Footnote 31
Mr. Bolton refused to leave things there. Instead, he brought an application before the Court of Queen's Bench, seeking a writ of certiorari.Footnote 32 The writ, if granted, would have required the magistrates to send the record of their possession order up to the Queen's Bench to be certified. If an error was disclosed in the possession order, then the court could have quashed the decision on that basis. This would have meant that Mr. Bolton and his family would have been able to return to their home, at least until a valid possession order was served on them.
That such a run-of-the-mill case as that of Mr. Bolton's made it to the Queen's Bench, and that Mr. Bolton was willing, and indeed able, to finance the proceedings, is quite remarkable.Footnote 33 More remarkable, however, is the effect this case had on the historical development of administrative law. We will consider the proceedings before the Court of Queen's Bench in depth, before considering the significance of the case to the history of administrative law.
IV. The Judgment in R. v Bolton
A. Proceedings in the Bail Court
Mr. Bolton applied for a writ of certiorari in the Easter term of 1839. His application was supported by affidavit evidence showing that he had paid poor rates, highways rates, church rates, and other taxes to the parish throughout his possession of the cottage.Footnote 34 He also showed in his affidavit that from 1836–37 he had served as headborough for Hampton Wick – something which might have been unusual for an alleged destitute pauper to have done.Footnote 35 Evidence introduced at the hearing before the magistrates that the cottage had previously belonged to the parish was also questioned by Mr. Bolton on the basis that the witness who testified to this fact “was a man of advanced age and feeble mind” who answered the questions in a contradictory and careless manner.Footnote 36 Mr. Bolton's case was at least persuasive enough for a rule nisi for certiorari to be granted by the Queen's Bench on 8 May 1839.Footnote 37 The rule called upon the justices of the peace to appear before the court in the next term and argue why a writ of certiorari should not be issued. A hearing was held in the Bail Court of the Queen's Bench in the following Trinity term.Footnote 38
The hearing in the Bail Court ought to have provided us with our first opportunity to assess the legal grounds on which Mr. Bolton sought the writ of certiorari. Unfortunately, however, counsel for Mr. Bolton were prevented from making argument by Coleridge J., the judge in the Bail Court proceedings.Footnote 39 Coleridge J. said simply that certiorari should be granted, thus ordering the justices of the peace to send their possession order up to the Queen's Bench. Coleridge J.'s ground for doing so was simple: the justices had acted without jurisdiction.Footnote 40
Coleridge J. based his decision on the wording of s. 24 of the Poor Relief Act 1819. Broadly speaking, this section provided that possession orders could be made by justices of the peace in circumstances where a person was residing in premises provided by the parish for the habitation of the poor, or otherwise where a person had unlawfully intruded upon any premises held by the parish, and where notice to quit had been served on such a person in the proper manner, as specified by s. 24. Mr. Bolton submitted an affidavit stating that he was not occupying the property as a pauper, and so the house in question could not be considered as being a house provided by the parish for the habitation of the poor as required by s. 24: “rent was paid for it in the same manner as for any other property of the parish.”Footnote 41 Secondly, Mr. Bolton had not unlawfully intruded upon the premises, but had been permitted by the parish to reside there. The possession order was therefore held to be the product of a jurisdictional error of law: “[i]t must be shewn that the party here is to be considered as belonging to one of the classes of persons mentioned in the section, in order to make the jurisdiction of the magistrates attach.”Footnote 42
It is difficult to tell whether Coleridge J. thought, in substance, that the justices’ order had been made without jurisdiction, or whether he was just saying that Mr. Bolton's submission of his affidavit alleging a jurisdictional error of law was enough to justify continuing the legal proceedings. Coleridge J.'s judgment in the Bail Court did not have the effect of quashing the possession order. Instead, the rule nisi for the writ of certiorari that had initially been granted to Mr. Bolton was made absolute. This meant that the writ issued to the justices of the peace, requiring them to send up the possession order to the Queen's Bench so that it could be more fully considered in proceedings to determine whether the order should be quashed. The possession order was sent up to the court and filed, along with the writ and affidavit evidence, in November 1839.Footnote 43 In January 1840, Mr. Bolton secured a subsequent rule nisi, calling upon the justices to come before the court and show cause why the order ought not be quashed.Footnote 44 Argument as to the quashing of the order was then heard before the Queen's Bench en banc (Lord Denman C.J. and Williams and Coleridge JJ.) on 11 November 1840.Footnote 45
B. Proceedings before the Queen's Bench
1. Counsel's argument
Before the Queen's Bench, the quashing of the possession order was opposed by the Attorney General, Sir John Campbell, and William Wightman. Mr. Bolton again argued that the justices of the peace had been wrong to grant the possession order because he had not been occupying the premises as a pauper. This argument was supported by the introduction of affidavit evidence demonstrating that Mr. Bolton had previously paid rent to the parish officials. The Attorney General and Mr. Wightman rejected this argument. It was said that the Queen's Bench was incapable, in cases like the present, of considering evidence relevant to the case – neither evidence that had been submitted to the justices of the peace at the original hearing, nor additional evidence that had not been submitted. On the face of the record, there seemed to be jurisdiction: the case was of the general kind over which the magistrates had power to hold a hearing and reach a decision. The magistrates, furthermore, were to be considered sole arbiters of the evidence in the case: their evaluation of the evidence had to be respected and could not be revisited by the Queen's Bench in certiorari proceedings. The decision had to be treated as final and, as such, there was said to be no ground for quashing the decision.Footnote 46
Contrariwise, William Erle and Charles Petersdorff, representing Mr. Bolton, criticised the Attorney General and Mr. Wightman for “laying down too wide a proposition”,Footnote 47 especially in cases, like the present, in which there was no statutory right of appeal against the impugned decision. They rejected the idea that simply the appearance of jurisdiction was enough: what mattered was whether in substance the justices had jurisdiction. From the report of the argument, however, it is unclear exactly how Erle and Petersdorff envisaged the ambit of review in these circumstances. They began by noting that the whole point of certiorari proceedings, and affidavit evidence submitted in support thereof, was to determine the jurisdiction of the decision-maker whose decision was challenged. Citing the turn-of-the-century case of R. v Inhabitants of Great Marlow,Footnote 48 it was then argued that where, in certiorari proceedings, the jurisdiction of a decision-maker to make a particular decision was impugned, the court could “look into the facts, and receive affidavits”.Footnote 49 Turning to the facts of this case, it was said that there was no evidence that could support the order.Footnote 50 As reported, these arguments do not seem, on their face, to be in disagreement with those advanced by the Attorney General and Mr. Wightman, who had accepted the possibility of challenging the order where it was made in the absence of jurisdiction. The rejoinder of Erle and Petersdorff, that affidavits were receivable where the challenge was based on the assertion that the justices had acted without jurisdiction, was no real rejoinder at all.
We need to cut through the obscurity of the reported argument to discern what the real issue was. At the heart of Bolton, it seems, was an argument about how the jurisdiction of the justices of the peace was to be conceptualised. We can infer from their reported arguments that the Attorney General and Mr. Wightman thought that the justices’ determination of whether Mr. Bolton had been occupying the premises as a pauper was in no way determinative of their jurisdiction to make the possession order. Because of this, the order was to be treated as conclusive and inviolable in certiorari proceedings: the error alleged on the part of the justices, as to Mr. Bolton's status as a pauper, did not relate to the jurisdiction of the magistrates, and thus could not be questioned through the invocation of affidavit evidence. Though more ambiguous, we might read Erle and Petersdorff as arguing the contrary proposition: that the justices only had jurisdiction if they had correctly determined Mr. Bolton to be a pauper, and affidavit evidence could be admitted to show that their determination was incorrect. Crucial to the resolution of R. v Bolton, then, was the way in which the jurisdiction of the justices of the peace was conceptualised: whether it depended on a correct determination of Mr. Bolton's status as a pauper or not. It was to this question that the justices of the Queen's Bench turned in resolving the case.
2. The judgment of the Queen's Bench
Lord Denman C.J. delivered the judgment of the court on 12 January 1841.Footnote 51 He began by noting that the case at hand had, at its heart, a principle as central to administrative law then as it is now: Parliament had entrusted a decision-maker with an original jurisdiction over the merits of a decision; there was no statutory right of appeal against that decision to the Queen's Bench or another superior court; the Queen's Bench had no concurrent original jurisdiction over the merits; and the court was therefore only able to review the decision where the decision-maker lacked jurisdiction to make the decision (evidenced by affidavits or otherwise) or where there was some error (jurisdictional or not) disclosed by the record.Footnote 52 The difficulty faced by the court in applying this principle, however, was the age-old problem of how exactly the jurisdiction of the decision-maker should be conceptualised. As the Lord Chief Justice said, “the principle upon which [review] turns is very simple: the difficulty is always found in applying it”.Footnote 53
Lord Denman went on to set out the court's understanding of how the principle should apply. He said:
[W]here a charge has been well laid before a magistrate, on its face bringing itself within his jurisdiction, he is bound to commence the inquiry: in so doing he undoubtedly acts within his jurisdiction: but in the course of the enquiry, evidence being offered for and against the charge, the proper, or it may be the irresistible, conclusion to be drawn may be that the offence has not been committed, and so that the case in one sense was not within the jurisdiction. Now to receive affidavits for the purpose of shewing this is clearly in effect to shew that the magistrate's decision was wrong if he affirms the charge, and not to shew that he acted without jurisdiction: for they would admit that, in every stage of the inquiry up to the conclusion, he could not but have proceeded, and that if he had come to a different conclusion his judgment of acquittal would have been a binding judgment.Footnote 54
In other words, when addressing the question of whether a decision-maker has jurisdiction, the focus for the reviewing court should be on events occurring at the beginning of the decision-making process. The court should consider the complaint or charge that was submitted to the decision-maker, and ask whether it raised a problem that the decision-maker had a legal power to resolve: that is, whether the subject matter of the proposed inquiry fell within the scope of the decision-maker's power. Jurisdiction was thus presented by the Queen's Bench as a single threshold to be crossed by a decision-maker at the start of the decision-making process. Once it was shown that the decision-maker had been presented with a charge that raised an issue that he was, in principle, jurisdictionally competent to determine, he would be free to make any decision in respect of the charge which he wished to make: nothing the decision-maker did after crossing the initial jurisdictional threshold could lead to a conclusion that his decision lacked a jurisdictional basis. All of this was neatly summed up by Lord Denman when he said that “jurisdiction does not depend on the truth or falsehood of the charge, but upon its nature: it is determinable on the commencement, not at the conclusion, of the inquiry”.Footnote 55
Lord Denman went on to apply this principle to the facts of the case, limiting his assessment of the validity of the justices’ order to the wording of the information that had been submitted at the start of the proceedings. The information alleged that Mr. Bolton was occupying premises belonging to the parish, and had failed to leave the premises in the appropriate time after being requested to do so by parish officials. The information, then, was limited to “circumstances required by the statute to found the jurisdiction” of the justicesFootnote 56: there was nothing in the information to suggest that the justices had been asked to make an order in circumstances over which they did not have jurisdiction. On the limited conception of jurisdiction put forward by the Queen's Bench, therefore, the justices of the peace had made a valid determination even if, on the facts, that determination might be considered a bad one. Because of this, and the fact that there was no error on the face of the possession order, there were no grounds on which the possession order could be quashed: no jurisdictional error of law had been committed.
3. Assessing the Bolton approach
The importance of R. v Bolton lies in the conception of jurisdiction put forward by the court. Jurisdiction was presented by Lord Denman as a concept associated with the beginning of an administrative inquiry: something which could not be lost once the inquiry had properly commenced. Implicit in this idea was a limiting of the concept of jurisdiction to questions concerned with the scope of a decision-maker's power. A reviewing court had to ask what subject matter the decision-maker had power to consider, and whether, at the commencement of an inquiry, the inquiry could be said to concern that subject matter.
It can be noted in passing that this conception is very different from the approach currently adopted by English administrative law. The modern approach, which might be traced to the decision of the House of Lords in Anisminic,Footnote 57 but which was only clearly accepted much later by the House of Lords in R. v Lord President of the Privy Council, ex parte Page,Footnote 58 can be said to be this: that any misinterpretation or misapplication of a statutory term by a decision-maker, at any stage of the administrative inquiry, will deprive the decision that is eventually reached of its jurisdictional basis.Footnote 59 In effect, the modern conception, following Page, is that practically all errors of law are to be considered jurisdictional.Footnote 60
But this comparison runs the risk of being trite unless we can show that Bolton had a wider significance for nineteenth- and twentieth-century administrative law. To assess this fully, we need to establish three things: first, the extent to which R. v Bolton differed in its conception of jurisdiction from cases that preceded it; second, why the change in Bolton came about; and third, the extent to which R. v Bolton influenced the development of the law that came after it. It is with these three questions that the remainder of this article will be concerned.
V. To What Extent did Bolton Change the Law?
A. Jurisdiction in Certiorari Proceedings before Bolton
When it comes to considering the way in which the superior courts, when reviewing the legality of administrative decisions, conceptualised the jurisdiction of decision-makers in cases predating Bolton, we need to distinguish between different areas of the case law. The area that is most directly relevant to Bolton is the case law determined under the prerogative writ of certiorari, as this was the remedial context in which Bolton itself was decided. We will therefore focus on that area of the case law.
R. v Bolton established an approach to error of law review that was substantively limited in its nature. It is clear, furthermore, from the pre-Bolton case law (at least that of the nineteenth century), that, where certiorari was sought on the ground of jurisdictional error of law, the King's Bench was very wary of allowing such applications to succeed, and thus adopted a very limited conception of the sort of error of law that would be considered jurisdictional. We might say, therefore, that Bolton represents continuity rather than change. However, when we probe deeper into the case law, it becomes clear that certiorari review before Bolton was different in one significant, and ultimately crucial, respect. While we might detect a limited approach to the review of jurisdictional errors of law in the pre-Bolton case law, the precise limits of that approach are deeply ambiguous, and it is unclear from the case law what the exact borderline was between jurisdictional and non-jurisdictional errors of law. R. v Bolton can therefore be seen as an important case because it marked out for the first time in express terms the conceptual limits of jurisdiction in certiorari proceedings for error of law, thus making it clear how a jurisdictional error of law might be distinguished from a non-jurisdictional error of law.
The approach taken in certiorari proceedings before R. v Bolton can be seen in the case of R. v Inhabitants of Great Marlow.Footnote 61 Justices of the peace had appointed four overseers for the parish of Great Marlow in Buckinghamshire. Two weeks after their appointment, after one of the new overseers had claimed that he was exempt from service because he was a yeoman in ordinary of the king, the justices purported to appoint a fifth overseer to replace the yeoman. This appointment was challenged in certiorari proceedings on the ground that the justices did not have jurisdiction to make it. It was said that, once the justices had made four valid appointments of parish overseers, their jurisdiction to make any more was at an end. The only circumstances in which the justices could appoint a fifth overseer were set out by s. 3 of the Poor Relief Act 1743Footnote 62: the justices only had power to appoint a replacement parish overseer on the death, removal, or insolvency of a previously appointed overseer. As none of these events had happened in the present case, there was no jurisdiction. Any further challenge to a valid appointment had to be made by way of appeal to the quarter sessions.
The King's Bench decided to quash the fifth appointment on the basis that it had been made without jurisdiction. The court's approach to the concept of jurisdiction can be seen most clearly in the judgment of Lawrence J. He said that “if there were a proper number of overseers legally appointed before, according to the provisions of the statute, a subsequent appointment of another overseer is merely void; the magistrates had no jurisdiction to make it”.Footnote 63 As Lawrence J. later went on to explain, “after the former appointment it was not competent to the other magistrates to receive [the yeoman's] excuse; but [the yeoman] should have appealed to the sessions, who might have allowed his excuse”.Footnote 64 Similarly, Le Blanc J. said that “the first appointment being good, all was at an end, and the other magistrates had no jurisdiction to make another appointment”.Footnote 65
The conception of jurisdiction adopted by the King's Bench could be seen as very similar to that adopted in Bolton. The focus for the court was on the beginning of the magistrates’ decision-making process. Jurisdiction seems to have been assessed without any consideration of the merits of the order of appointment. The role of the King's Bench was simply to ask whether one of two circumstances in which an overseer could be appointed had been satisfied: first, if there had not yet been valid appointments of four overseers; and second, if there had, if one of the criteria identified by s. 3 of the Poor Relief Act 1743 as being necessary for a valid reappointment – if an overseer had died, left the parish, or become insolvent – was satisfied. If these circumstances could not be made out, then the justices of the peace ought never to have considered appointing the fifth overseer; they had no power to commence their administrative inquiry, and thus their resulting appointment was invalid. But there is a danger here that we are reading too much into the case, and retrospectively refashioning the limited dicta of the justices of the King's Bench in light of R. v Bolton. It is certainly true to say that the approach in Great Marlow suggests a limited approach to certiorari review: the justices of the King's Bench said, in effect, that the substance of the magistrates’ appointment – whether it was right to appoint a particular person as an overseer and so on – could never be challenged in certiorari proceedings. But it is difficult to say any more than that. What if the justices of the peace had considered, for example, whether an overseer had become insolvent, and concluded, perhaps against the evidence, that he had? Would the King's Bench have been able to quash any subsequent appointment order on this basis, or was this question one to be determined conclusively by the justices? There is little in Great Marlow to shed light on this point. While the judgment demonstrates a judicial tendency towards limited certiorari review, the conceptual framework by which this is to be secured is not clearly spelled out: we are not told how to distinguish jurisdictional questions of law from non-jurisdictional questions of law; we are not, as in Bolton, given a framework that says that jurisdictional questions are those concerned with the commencement of an administrative inquiry.
This phenomenon – the combination of a tendency towards limited review and the lack of a descriptive methodology for securing it – can be seen running through the certiorari case law predating R. v Bolton. Thus, where justices of the peace had made an order upholding the annual accounts of a highways surveyor without those accounts first being submitted to a justice of the peace sitting at a special hearing for such purposes, as required by the Highways Act 1773,Footnote 66 their order was quashed in certiorari proceedings.Footnote 67 No explanation was given, however, of why the order was considered non-jurisdictional. All that was said was that “sufficient was not done to satisfy the words of the statute, and … the allowance [of the accounts] by the petty sessions was therefore invalid”.Footnote 68 We can see this also in a case where certiorari was sought to quash a decision of the quarter sessions, on appeal, to overturn a magistrate's order removing a pauper from one parish to another.Footnote 69 The quashing of the quarter sessions’ appellate decision was sought on the ground that a procedurally proper notice of appeal had not been given before the sessions commenced their hearing: a ground, we might think, that came under even the limited conception of jurisdiction put forward in Bolton, given that the quarter sessions had to satisfy themselves at the start of the hearing that proper notice had been given before going on to consider the substantive issues raised in the appeal. Nonetheless, certiorari was held not to be available in the case. In a confused judgment, Lord Denman C.J. said that the quarter sessions had jurisdiction to act because the appeal was “duly lodged”,Footnote 70 but failed to explain why questions as to the procedural correctness of the notice of appeal were questions of merits for the quarter sessions to determine and not factors that had a bearing on whether an appeal was duly lodged, and thus the quarter sessions’ jurisdiction.
These and many other certiorari casesFootnote 71 all demonstrate reluctance on the part of the King's Bench to interfere with the merits of an original decision: a constancy in favouring limited review. What these cases omit, however, is an explicit articulation of the conceptual tool through which limited review could be secured. That tool was a conception of jurisdiction which expressly confined questions of jurisdiction to those concerned with the scope of an administrative inquiry, determinable at the inquiry's commencement. It was not until the decision in R. v Bolton that such an approach was consciously articulated and adopted in certiorari proceedings.
B. Jurisdiction and Proceedings in Trespass
The foregoing has demonstrated that the Court of Queen's Bench in R. v Bolton put forward a novel approach to certiorari review for jurisdictional error of law. Given the marked lack of support for this approach in earlier certiorari cases, we might wonder what the source of inspiration was for the justices of the Queen's Bench in Bolton when they came to formulate their judgment. An answer can be gleaned by looking at the non-certiorari cases that were cited by Lord Denman C.J. in his judgment.
Two such cases were relied on in Bolton for the decision of the Queen's Bench, both decisions of the Common Pleas in tort proceedings. The relevance of tort should be unsurprising. Statutory powers could allow for quite invasive conduct on the part of administrators and their officers, and exercises of such powers could constitute, prima facie, trespass to goods or to the person. While decision-makers could justify their conduct on the basis that they had a statutory power to act in a way that would otherwise be unlawful, liability would arise in tort if it could be shown that their decision had been made without jurisdiction: namely that the statute did not apply to their act, rendering it unlawful.Footnote 72 Trespass actions were thus a convenient way of collaterally challenging the legality of administrative decisions.
In the first trespass case cited in Bolton, Brittain v Kinnaird,Footnote 73 an action had been brought against magistrates who, purportedly exercising powers conferred by the Thefts Upon the Thames Act 1762,Footnote 74 had made an order for the possession of a vessel and its 500 pounds of gunpowder. The plaintiff brought the action on the basis that the justices’ order had been made without jurisdiction: the Act only applied to “boats”, and the plaintiff's larger vessel clearly did not satisfy this description. Counsel for the justices objected, arguing that the defendants’ order had to be treated as conclusive of the facts for the purposes of an action in trespass: “… as long as a [decision] remains unquashed, it is conclusive of the facts stated in it.”Footnote 75 While it was admitted that the order would not be conclusive if it had been made without jurisdiction, a limited conception of jurisdiction was put forward to prevent such an argument. “Whether the subject matter of this conviction were a boat or not,” it was said, “was the very question to be decided before the magistrates, and upon which his decision was final.”Footnote 76 This defence found favour with the justices of the Common Pleas, who saw this argument as sitting consistently with a long line of authority in similar trespass cases. The court's conception of jurisdiction was made clear by the judgment of Dallas C.J.: “The magistrate, it is urged, could not give himself jurisdiction, by finding that to be a fact, which did not exist. But he is bound to enquire as to the fact, and, when he was enquired, his conviction is conclusive of it.”Footnote 77
A similar conception of jurisdiction was put forward in Cave v Mountain.Footnote 78 The plaintiff had been taken into custody following an allegation that he had cut down trees belonging to a local cleric. In the Common Pleas, the plaintiff sought damages from the justice of the peace who had committed him to gaol on the basis that his committal had been made without jurisdiction: the statutory provision in question,Footnote 79 it was argued, gave the defendant jurisdiction to commit the plaintiff only where the alleged conduct caused more than £1 in damage. Here, it was said, no evidence had been laid before the defendant that the loss caused by cutting down the trees exceeded £1.Footnote 80 The Common Pleas refused to hold the defendant liable, however, again invoking a limited conception of jurisdiction:
[T]here can be no doubt but that if a magistrate commit a party charged before him, in a case where he has no jurisdiction, he is liable to an action of trespass. But if the charge be an offence over which, if the offence charged be true in fact, the magistrate has jurisdiction, the magistrate's jurisdiction cannot be made to depend upon the truth or falsehood of the facts.Footnote 81
In both Brittain v Kinnaird and Cave v Mountain, we thus see an explicitly limited conception of jurisdiction being used by the superior courts of common law in trespass proceedings. This is the same limited approach adopted later in R. v Bolton.
We might wonder, though, why until R. v Bolton different conceptions of jurisdiction apparently existed in trespass and certiorari proceedings. Indeed, in many cases, certiorari and trespass actions often worked in tandem, certiorari being sought to quash a decision before proceedings in tort were commenced: it was common for plaintiffs first to secure the quashing of an administrative order that might otherwise be used as a defence in an action in trespass.Footnote 82 We might therefore expect it to be obvious that the conception adopted in the latter would be mirrored in the former.
If we approach jurisdiction as something which ought only to have a single, universal conceptualisation, then the shape of the pre-Bolton case law might be questioned. However, it is clear that, as a matter of judicial policy, there were strong reasons why the courts might have adopted a more clearly limited conception of jurisdiction in trespass proceedings while at the same time maintaining a looser conception of jurisdiction for the purposes of certiorari. Two in particular can be identified: preserving administrative efficiency, and minimising the serious personal consequences of liability in tort.
In terms of administrative efficiency, if justices’ orders were not treated as relatively inviolable in trespass proceedings, at least until they were quashed in certiorari proceedings, then this could have resulted in serious administrative difficulties. To take the context of the poor law, seen in Bolton, as an example, every time an order removing a pauper from one parish to another was made, the pauper could conceivably have brought an action for trespass to the person. Only a very limited conception of jurisdiction would prevent decision-makers’ orders in such circumstances being collaterally reviewable in trespass proceedings. Of course, issues of administrative efficiency were equally relevant in the context of certiorari proceedings: an overly generous availability of certiorari would also have been damaging to the efficiency of public administration. But certiorari was subject to many procedural constraints, such as time limits and the requirement of recognisances,Footnote 83 to which a plaintiff in trespass proceedings would not have been subject. The adoption of a high threshold for liability in tort, through a tighter conception of jurisdiction, would be a sensible course for the courts to take.
Minimising the serious personal consequences of liability in tort was another reason for adopting a more limited conception of jurisdiction in trespass proceedings. Liability in trespass attached to the defendant in his personal capacity, marking him out as someone who had committed a wrongful act. In comparison, certiorari focussed on the office occupied by the defendant. Although morally repugnant acts could of course be challenged in certiorari proceedings, a finding that the defendant, as magistrate or other office-holder, had acted without jurisdiction did not necessarily import a sense of moral wrongfulness: a justice of the peace could easily make mistakes. Connected to this, and perhaps more importantly, liability for trespass rendered the defendant personally liable to pay damages to the plaintiff, whereas a finding that a decision-maker lacked jurisdiction for the purposes of certiorari meant nothing more than the decision could be quashed. In other words, the consequences of liability in tort were, from the perspective of individual decision-makers, much more serious. Of course, justices would find certiorari proceedings discomforting, too.Footnote 84 Nonetheless, the greater reputational and financial ramifications of an adverse verdict in an action in trespass would perhaps have made certiorari somewhat less undesirable.
There were, then, good reasons for adopting different conceptions of jurisdiction in trespass proceedings and certiorari proceedings respectively. Indeed, some attention was paid to this point in Bolton itself. In argument, Erle and Petersdorff submitted that the whole purpose of certiorari was to substitute the reviewing court's evaluation of the evidence for the original evaluation by the magistrates, and hence a conception of jurisdiction wider than that taken in the trespass case law should be taken in certiorari proceedings.Footnote 85 This argument was not, however, addressed by the Queen's Bench. After citing Brittain v Kinnaird and Cave v Mountain, Lord Denman simply commented that “[t]hese cases were both of them actions of trespass against the magistrate convicting; but they are authorities not on that account the less in point on the present occasion”.Footnote 86
VI. Explaining the Change: Bolton and its Wider Context
We have seen above exactly why R. v Bolton was historically significant to the development of English administrative law. The case was important in that it represented the conscious transplantation into certiorari proceedings of a conception of jurisdiction traditionally used to determine liability for trespass. This was despite the fact that the reasons for the limited conception adopted in trespass proceedings were not especially relevant in the context of certiorari. Noting Bolton's significance, however, does not explain why this development occurred. An answer to that question, it is suggested, might be found not in the case law itself, but in the context of the broader reforms to the legal system that occurred in the nineteenth century. More specifically, it is suggested that the adoption of a limited form of review in R. v Bolton, and its continuation from the 1840s onwards, represents a concern on the part of the Queen's Bench to reflect a new scheme of review and appeal provided by Parliament in various statutory reforms concerned with the jurisdiction of magistrates to convict for summary criminal offences.
To understand these developments fully, we need to begin by considering the various processes that were available at the start of the nineteenth century for defendants who had been convicted in summary proceedings to challenge their conviction. In total, a defendant had four options open to him or her. The first two options were in substance appeals rather than review, being concerned with the merits of a conviction rather than with the jurisdiction of the magistrates to make it. A defendant could either seek a writ of certiorari to challenge his or her conviction on the basis of some error of law – jurisdictional or not – disclosed on the face of the record. Alternatively, he or she could appeal to the quarter sessions and then petition the sessions to state a case for the opinion of the King's Bench at Westminster.Footnote 87 The former option would be easy to procure and was very wide-ranging: at the beginning of the nineteenth century, the records of magistrates’ summary proceedings – known as “speaking records” – contained a wealth of information, meaning that many errors of law were disclosed on their face (in contrast to the records of justices’ administrative orders, which were much less detailed and thus demonstrated fewer errors on their face)Footnote 88; and the King's Bench had a reputation for being meticulous in its scrutiny of such records when they were brought up for review.Footnote 89 The latter option would be equally generous, but much harder to procure: it would depend on there being an express right of appeal to the quarter sessions in the statute creating the summary offence, the absence of a no-certiorari clause in the statute (as the conviction and stated case were technically brought up to the King's Bench by a writ of certiorari), and, importantly, the voluntary assent of the quarter sessions to the process.Footnote 90
As well as these options, a defendant had two other ways of challenging his or her conviction. Where the defendant had been imprisoned as a result of a magistrate's conviction, or where his goods had been seized, he could bring an action in trespass against the magistrate.Footnote 91 However, if the wrongdoing alleged by the defendant lay within the jurisdiction of the justices, he would need to have the conviction quashed in certiorari before commencing his actionFootnote 92 and, following the enactment of the Justices Protection Act 1803,Footnote 93 he would only be able to recover a maximum of twopence damages for any claim for intra-jurisdictional wrongdoing on the basis of a quashed conviction, unless he could show malice on the part of the convicting justices and lack of probable cause.Footnote 94 A fourth and final option for a defendant would be to seek certiorari on jurisdictional grounds, introducing affidavit evidence. Given the quality of “speaking records”, however, this would be an unusual course of action.
For a defendant at the start of the nineteenth century, then, the best way of challenging a summary conviction would be appeal-like certiorari for error of law disclosed on the face of the record. The types of challenges that could be brought were much more wide-ranging than those available in the case of jurisdiction-based certiorari or trespass actions, and certiorari for patent errors of law was much easier to secure than appeal to the quarter sessions and review by consensual case stated. From the very start of the nineteenth century, however, and perhaps earlier,Footnote 95 the availability of this form of challenge was gradually restricted by Parliament. Statutes such as the Summary Proceedings Act of 1822Footnote 96 and various statutes passed under Peel's administrationFootnote 97 restricted the availability of “speaking records” by instituting short forms of conviction that disclosed very little detail, barring certiorari, and substituting appeal to the quarter sessions.Footnote 98 By 1839, Stone, in his treatise on the petty sessions, did not consider it necessary to discuss certiorari in detail “on account of the unfrequency with which summary convictions are now brought under the notice of the Judges”.Footnote 99
As opportunities for certiorari review for errors on the face of a conviction came to be restricted, defendants came to be pushed more towards other ways of challenging their conviction. Because of the difficulties in appealing to the quarter sessions and encouraging them to state a case for the King's Bench, more pressure came to be put on affidavit-based certiorari for jurisdictional error of law and actions for trespass. The problem for defendants was that the conception of jurisdiction adopted in these circumstances was limited, and thus the potential grounds for challenging convictions were far fewer. In the context of trespass actions, this phenomenon was particularly pronounced, given the explicitly limited approach to jurisdiction adopted in those proceedings. In pre-Bolton certiorari, however, the conception of jurisdiction was not so precisely defined. Defendants seeking certiorari could in theory, therefore, have exploited the doctrinal ambiguity in the case law, securing a greater chance of challenging their convictions.
The Queen's Bench was thus faced with two options. In the spirit of certiorari's sixteenth-century history, whereby the writ was developed as a strong-headed common law response to the rise of the summary powers of justices of the peace,Footnote 100 the court could have exploited the ambiguity in their conceptualisation of jurisdiction, widening the conception so as to allow for the breadth of review that had been available on the face of the record before the introduction of statutory short-form convictions. To do so, however, would have been constitutionally difficult: Parliament had clearly intended a restriction of the inherent power of the Queen's Bench to review convictions, and an arrogation of reviewing powers in the face of this legislative intention would have had the effect of undermining this. Instead, the Queen's Bench chose to cement its conception of jurisdiction in certiorari proceedings, adopting the explicitly limited approach to jurisdiction that was already adopted in trespass cases. This adoption of a more explicitly limited conception of jurisdiction in certiorari proceedings came, as we know, in R. v Bolton. Given that it was in certiorari proceedings concerned with justices’ administrative orders that the conception of jurisdiction was most commonly addressed (not all summary offences had, even at this stage, prescribed short-form convictions, and thus patent review for jurisdictional and non-jurisdictional errors of law was still common in that context), it should not be very surprising that this important decision for the review of summary convictions came about in that somewhat different context.
VII. Bolton and the Development of Judicial Review
We have considered above the change that R. v Bolton brought about in administrative law, and suggested some historical reasons as to why this change might have occurred. We cannot fully understand the significance of Bolton, however, without examining its influence on the later development of the law. It is to this question that this section will now turn.
The general restriction of review of summary convictions in certiorari, discussed above, came to be greatly accelerated under the Acts of Sir John Jervis in 1848.Footnote 101 These reforms, in particular the Summary Jurisdiction Act 1848Footnote 102 which introduced a uniform and sparsely detailed form of conviction, rendered speaking records forever non-existent.Footnote 103 Especially important for our purposes, the Act applied also to many of the administrative orders of justices.Footnote 104 As well as the Summary Jurisdiction Act, the Justices Protection Act 1848Footnote 105 emphasised Parliament's desire to restrict proceedings in trespass as well as certiorari, removing “most of the opportunities to use tort actions against enforcers [of invalid decisions] to reopen the validity of [those] upstream decisions”.Footnote 106 These restrictions on the availability of certiorari were, furthermore, only part of Parliament's plan to reform magistrates’ statutory powers and the ways in which exercises of them could be challenged. Parliament also enacted various statutes which made appeal to the quarter sessions, coupled with consensual case stated to the Queen's Bench, the standard way of challenging such decisions. Contemporaneously with the Jervis Acts, the Poor Law Procedure Act 1848Footnote 107 vested the quarter sessions with the sole ability to judge the sufficiency of the paperwork that had brought appeals on poor-law removal orders to it.Footnote 108 The Quarter Sessions Act 1849Footnote 109 extended this to nearly every other administrative context where the original Act gave an appeal. In 1857, Parliament passed another Summary Jurisdiction Act,Footnote 110 which marked out the consensual case stated process as the proper way of addressing errors committed by justices of the peace.Footnote 111 And, in 1879, the Summary Jurisdiction ActFootnote 112 extended the facility to appeal by case stated to all convictions, orders, determinations, and other processes of any court of summary jurisdiction, removing the requirement of certiorari to bring up the stated case to the Queen's Bench.Footnote 113
In light of these developments, it should be no surprise that reasons for the concept of jurisdiction in R. v Bolton, as well as the consequences of the decision itself, continued to have a real effect on the development of certiorari review. When we look at cases decided after 1841, the vast majority of them show the concept of jurisdiction being articulated in error of law cases in strictly limited terms. Jurisdiction came to be seen exclusively as a concept concerned with the scope of an administrative inquiry, and as something which could be determined conclusively at the time that inquiry was commenced. In R. v Justices of Buckinghamshire,Footnote 114 for example, in rejoinder to an argument that the factual findings of justices of the peace in poor-law proceedings could be re-examined by the Queen's Bench under the writ of certiorari, the Queen's Bench invoked an expressly limited conception of jurisdiction:
Here, if there had been no complaint, the magistrates would have had nothing before them on which they could make an order: but, there being a complaint, they were bound to inquire into all the facts on which an order must be grounded: chargeability, residence, and settlement. It appears that they have done so: and, that being the case, the late cases shew that this Court will not have the examinations brought before it for the purpose of instituting a nice inquiry whether all the necessary facts were fully stated. That may be a matter of appeal, but is not a subject to be considered here.Footnote 115
Similarly, in R. v Buchanan,Footnote 116 in which the validity of a poor-law removal order came into question, the Queen's Bench felt able, less than a decade after Bolton, to state the Bolton approach to jurisdiction as simple orthodoxy:
It is impossible for us to lay down the rule that we will inquire in every case whether the magistrates have arrived at the right conclusion, and if we think they have not, say they had not jurisdiction. Such a rule would impose on us the duty of reviewing all proceedings before magistrates. The course to be taken is that prescribed in the admirable judgment in R. v Bolton; we must see that the charge gives jurisdiction, and, if it does, that the magistrates proceed to determine the charge.Footnote 117
These cases are simple examples of a much larger phenomenon. They have been selected from a large group of casesFootnote 118 as typical examples of the court's consistent approach throughout the nineteenth century. After R. v Bolton, the court began to conceptualise jurisdiction in more limited terms, securing a very limited form of review. Error of law review in certiorari, at least in those cases not concerned with errors of law on the face of the record (practically defunct after the legislative changes just mentioned), was restricted to those errors that related to the scope of the inquiry, assessed from the theoretical perspective of the inquiry's commencement and thus independent of the correctness of determinations reached during the course of the inquiry.
Moving into the twentieth century, the Bolton approach can be seen running through the judicial review case law throughout most of the century. In R. (Limerick Corporation) v Local Government Board, Footnote 119 a decision of the Local Government Board to award superannuation allowances to certain office-holders upon resignation could not be challenged on the basis that the beneficiaries of the decision had not, in fact, been office-holders at the time they had resigned: whether or not they were office-holders at the time was a question of merits, to be determined conclusively by the compensation authority, and a valid decision on this question could be reached even if marred by an error of law. Similarly, a decision of magistrates in Weston-super-Mare that a catering company had failed properly to police food rationing regulations in its canteens could not be quashed on the ground that the company was not a “person having control or management of an establishment”, as required by art. 13 of the Rationing (General Provisions) Order 1942Footnote 120: “… the question whether the [company] were in control or management was not collateral, but was part of the very issue which the justices had to inquire”, and so certiorari could not issue if that question was answered erroneously.Footnote 121 The dominant approach to error of law review was made clear in the third edition of Halsbury's Laws of England: “When the inferior tribunal has jurisdiction to decide a matter, it cannot (merely because it incidentally misconstrues a statute, or admits illegal evidence, or rejects legal evidence, or misdirects itself as to the weight of evidence, or convicts without evidence) be deemed to exceed or abuse its jurisdiction.”Footnote 122
The conception of jurisdiction put forward in Bolton, and adopted by the later case law, though limited, still allowed for a meaningful form of judicial review. Thus, in R. v Bradford, in which justices of the peace had a statutory power to authorise highways surveyors to take materials from certain enclosed areas of land, provided that such an area was not a park, their order could be challenged on the basis that they had wrongly decided that a given area of land was not a park: “… the question whether a place is a park or not is … not for the justices finally to determine”Footnote 123; “… if the place is a park in fact, they cannot give themselves jurisdiction by finding it is not a park.”Footnote 124 Likewise, where a judge of the county court allowed for a number of university students to be taken off the city's register of electors without giving them written notice of the judicial hearing that led to his decision, his decision was quashed: failing to abide by the procedural requirements for the proper commencement of a decision-making process led to a decision being made without jurisdiction.Footnote 125
To us, reading about these cases at a high level of abstraction, it is of course difficult to assess why it was, for example, that, in Limerick, the Local Government's Board decision as whether the applicants to the superannuation scheme were office-holders at the time they resigned was a question of merits, while, in Bradford, the question of whether the land was a park was a jurisdictional question to be answered before proceeding to the merits. But it is wrong to see this as a deficiency in the R. v Bolton distinction per se. Like any legal concept, the Bolton distinction requires substantiation by the context of the particular case in which it is applied.Footnote 126 In that sense, it is no different from the resolution of a contractual dispute through the application of the broad principles of contract law to the details of a sophisticated commercial transaction, or the broad principles of the tort of negligence to the events surrounding a personal injury or the commission of property damage. Or, to take modern administrative law as an example, it is no different from taking the general principle that now (almost) all errors of law lead to a decision being quashed and asking, on the facts and statutory norms of a particular case, whether an error can be said to have been committed. All abstract legal principles direct the court's interpretation and resolution of the details of a particular case. In this sense, the Bolton approach is no exception. The important point is that the courts took as a general framework the idea that jurisdictional questions were those that had to be ascertained at the commencement of an administrative inquiry, and then sought to apply that to the unique statutory framework of the case before them by seeking to determine which questions Parliament really intended the decision-maker to answer conclusively. That we cannot predict the outcome without as detailed an analysis of the governing statute does not mean that the Bolton approach leads to a wilderness of single instances. The variety of outcomes is nothing more than a product of the variety of statutory administrative schemes that the courts came into contact with, and the courts’ genuine attempt to give effect to how and, crucially, by whom Parliament intended that scheme to be administered.
Throughout the twentieth century, the Bolton approach continued to set the dominant tone for judicial review for jurisdictional error of law, at least when it came to certiorari proceedings. Lord Sumner, in the Privy Council case of R. v Nat Bell Liquors Ltd.,Footnote 127 said that “the law laid down in R. v Bolton has never since been seriously disputed in England”.Footnote 128 Later in his opinion, he described Bolton as “undoubtedly … a landmark in the history of certiorari”.Footnote 129 Similar views had been expressed in an earlier decision of the High Court of Ireland, where Lord O'Brien C.J. had commented that the Bolton approach had constituted “an unbroken line of authority from 1841 to the present day”, before going on to apply that principle to the case before him.Footnote 130 The Bolton approach also came to shape legal scholarship as well as the case law. It was the Bolton approach that came to be rigorously analysed and defended by Gordon in a series of articles on the concept of jurisdictionFootnote 131; it was an approach that came to be endorsed, albeit with some caution, by de Smith as the “pure theory” of jurisdiction.Footnote 132 Some commentators, however, were more sceptical. For example, Rubinstein, in his monograph on the concept of jurisdiction published in 1965, while accepting that the theoretical approach put forward by Gordon had “some logic to it”, nonetheless thought that “if carried to its logical extreme … [would] prove self-defeating”.Footnote 133
To an extent, Rubinstein was right. As the twentieth century progressed, the very limited form of review allowed by the Bolton approach came to be seen as less desirable. No doubt this is due in part to the increased responsibility borne by the post-war state, and the concomitant increase in statutory powers that this necessitated.Footnote 134 As public decision-makers were given more opportunities to influence the lives of ordinary citizens, the light-touch approach to judicial review the courts had hitherto adopted came to look out of place. In a century influenced heavily by Dicey's characterisation of broad discretionary powers as something antithetical to the rule of law,Footnote 135 the common law came to be seen as being ill equipped for its constitutional role of ensuring government according to law.
Any problem of conceptualising jurisdiction was thus, as noted by de Smith, “one of public policy rather than one of logic”.Footnote 136 As Wade said, the Bolton approach could be abandoned for “good and self-evident reasons”,Footnote 137 given that “it opens the door to arbitrary power”. Accordingly, towards the middle of the twentieth century, the courts appeared to pull away from Bolton in certain administrative contexts, expanding the concept of jurisdiction and allowing for more error of law review under the writ of certiorari. This was done by a subtle manipulation in the Bolton approach, massaging the boundary between jurisdictional and non-jurisdictional errors of law. Though it continued to be accepted that, theoretically, jurisdiction was a concept concerned with the commencement of an administrative inquiry, and could not subsequently be lost through a wrong decision on the merits, judges began to characterise as preliminary or collateral questions what might most logically be seen as questions of merits: the merits of an inquiry came quite illogically to condition when an inquiry was considered to have been properly commenced. This was not a phenomenon seen across the breadth of administrative law, but it was noticeable in certain important contexts, in particular in the context of the judicial review of rent tribunals, exercising powers conferred by Acts of Parliament instituting a system of rent controls.Footnote 138
Nonetheless, throughout most of the twentieth century, the conception of jurisdiction adopted in R. v Bolton continued to dominate judicial review, especially in certiorari cases. We can see this right up to the decision in Anisminic v Foreign Compensation Commission, with Bolton still acting as a starting point for analysis of the concept of jurisdiction in the Divisional CourtFootnote 139 and Court of Appeal.Footnote 140 In the judgment of the latter, for example, Diplock L.J. insisted on the importance of drawing a distinction between “the description in the statute of the kind of case into which an inferior tribunal has jurisdiction to inquire, and the description … of the kind of situation the existence or non-existence of which that tribunal has jurisdiction to determine”.Footnote 141 Sellers L.J. put it more squarely, referring directly to R. v Bolton and affirming that “the question of jurisdiction is determinable at the commencement, not at the conclusion, of the inquiry”.Footnote 142 It seems that it was only with the decision of the House of LordsFootnote 143 that the Bolton approach to jurisdictional errors of law, adopted as orthodoxy for nearly 130 years, came seriously to be challenged.Footnote 144
VIII. Assessing the Significance of R. v Bolton
We have seen that the decision in R. v Bolton was crucially important to the development of administrative law. The case put forward, in the context of certiorari proceedings for jurisdictional errors of law, a conception of jurisdiction that was explicitly limited. The Bolton approach, which limited the concept of jurisdiction to questions concerned with the scope of an administrative inquiry – the subject matter over which a decision-maker was given power to determine – and which provided that the existence of jurisdiction was something that could be determined conclusively at an inquiry's commencement, had already been used for some time in tort proceedings. Despite there being a number of policy reasons for not adopting a similar approach in the context of certiorari, the Queen's Bench decided to do so in 1841 because of new pressures put on certiorari for jurisdictional error of law, brought about by statutory changes to the way in which the summary convictions of magistrates could be challenged. Once adopted, the approach dominated judicial review for jurisdictional error of law for the next 130 years, especially in the context of certiorari. It can be seen to be having an impact on the functioning of English administrative law right up to the decision of the House of Lords in Anisminic and, to some extent, beyond.
This historical picture ought to be interesting in its own right, but it should also provide food for thought for scholars and practitioners of modern administrative law. There are a number of lessons we can learn by looking at the history of public law. From the foregoing, it is hoped that the following conclusions might have some bearing on the current workings and future development of the law.
The first point to make is that history shows there to be no such thing as a doctrinally correct or logically necessary conception of jurisdiction. There is no pure conception of jurisdiction, existing in the abstract, whose discovery only requires a more concerted straining of our legal and logical reasoning skills. We have seen that, throughout the modern history of administrative law, different conceptions of jurisdiction – all equally workable – have been applied in different remedial and administrative contexts. It was nothing more than variations of legal and administrative context that initially meant that different conceptions of jurisdiction were adopted for the purposes of certiorari review and trespass actions respectively. It was nothing more than the demands of context that eventually led to the uniform adoption of a limited conception of jurisdiction in Bolton. We should not be tricked into thinking that the apparent uniformity in our current law's conception of jurisdiction, stemming from Anisminic and Page, is any less context-specific: by the time Anisminic came to be decided, the limited approach to review enshrined by Bolton was being put under strain because of the expansion of the administrative state during the early decades of the twentieth century, and it is no doubt the result of a perceived need for increased judicial review, from Anisminic to today, that explains the current conception of jurisdiction adopted by the courts. As has been argued elsewhere, the adoption of the Anisminic approach to jurisdiction can relatedly be rooted in the rise of the declaration as an administrative law remedy.Footnote 145
Nonetheless, there are serious problems with the post-Anisminic conception of jurisdiction, solutions to which might be revealed after deeper reflection on the law's history. Currently, administrative law doctrine treats practically all questions of law as relating to the jurisdiction of a decision-maker, and therefore reviewable on that basis. This deprives decision-makers of the autonomy to determine legal questions for themselves. Of course, in certain contexts – perhaps in the majority of contexts – such an approach may be appropriate: the desire for administrative autonomy ought not to trump the importance of upholding the rule of law, through judicial review of an administrator's legal determinations, in every case. But, in recent years, this approach has come under increased strain, especially in certain contexts where administrative autonomy definitely is seen as more desirable.Footnote 146
If we recognise the importance of context to these different conceptions of jurisdiction in administrative law, however, we can be freed from some of the fetters which are supposedly put on modern law by the so-called orthodoxy of Anisminic. We will instead be able to shape the conception of jurisdiction adopted in judicial review proceedings to the context of particular cases. In one context, an Anisminic-style conception of jurisdiction might be appropriate; in another context, for different reasons, a more limited conception might be better. Provided legal doctrine is developed and applied in an open and clearly articulated manner, this must be preferable to the heavily pragmatic resolution of these issues that is starting to be favoured by English courts – a resolution through which the supposed logical demands of doctrine are recognised yet in the same breath pushed aside where various extra-doctrinal factors pertain. To take the litigation in R. (Cart) v Upper Tribunal as an example,Footnote 147 the Divisional Court and the Court of Appeal took a doctrine-centred approach, adopting an alternative conception of jurisdiction to that ordinarily adopted in judicial review proceedings to limit the reviewability of the Upper Tribunal. Within the context of the statutory scheme, which created an independent and expert judicial body which was intended by Parliament to possess a high degree of autonomy, the Anisminic-style of error of law review was considered unsuitable. Something more like the Bolton-style limited review was thought better. The Supreme Court, however, favoured a different approach. While it was accepted that review of Upper Tribunal decisions should be restricted, the use of the doctrinal distinction between jurisdictional and non-jurisdictional errors to effect this restriction was rejected in favour of more pragmatic considerations – whether the issues of a particular case raised an important point of principle or practice, or whether there were other compelling reasons for review.Footnote 148 The result was conceptually muddled, removing questions as to the scope of review from its conceptual rooting in questions of vires and jurisdiction.Footnote 149 A similarly regrettable approach has also been seen in the context of the distinction between law and fact.Footnote 150
As well as misunderstanding the role of doctrine, the modern approach to jurisdictional error of law is also problematic because it is wilfully ignorant of what Parliament might have intended as to the balance between agency autonomy and protection of the rule of law. Is it not reasonable for us assume that, in certain contexts, Parliament does intend, when it vests administrative powers in a decision-maker, that certain legal questions are to be answered autonomously by that decision-maker? If parliamentary sovereignty is to be treated as a meaningful principle in our constitution – a principle that has a real as opposed to merely nominal impact on the way in which cases are decided – then this is a question that needs to be addressed every time the courts come to judicially review an alleged error of law, and thus come to determine what exactly a decision-maker has jurisdiction to decide. While the modern approach to error of law review allows little opportunity to do this, it is hoped that from reflecting on the history of R. v Bolton, and seeing how ultimately it was respect for parliamentary intention that influenced the development of the law from 1841, a renewed respect might be found for the role of the legislature in public law adjudication.
In some cases, perhaps the historical approach to limited error of law review, enshrined in Bolton, should not be completely ruled out. While the Divisional Court and Court of Appeal in Cart favoured such an approach, seeing it as more accurately satisfying the twin demands of the rule of law and the autonomy of the Upper Tribunal in the context of that case, Lady Hale and the other justices of the Supreme Court feared that such an approach could return the law from the wilderness of single instances that supposedly existed before Anisminic.Footnote 151 Once it is appreciated, however, that no such wilderness of single instances existed, taking such a limited approach to review is again made possible, at least in those cases like Cart in which it seems appropriate.Footnote 152 We can escape from the idea of the wilderness of single instances once and for all, by confronting the history of judicial review and showing the wilderness to be nothing but an illusion.
