Previous chapters contain numerous instances of courts passing judgment on the legality of decisions, regulations, and actions made in pursuance of statutory authority, which together constitute the division of law in the twentieth century called judicial review or, in one of its senses, administrative law.¹ The twentieth-century texts by S. A. de Smith or H. W. R. Wade have no nineteenth-century counterparts, however, because to have such a general concept of review depends on a perception of state or official power which itself transcends the particular context of its exercise.² That was twentieth-century thinking, rooted, in England, in the politics that followed the third Reform Act, 1884. In the nineteenth century legal writers embedded their analysis of courts’ powers to determine legality within an account of a particular context—summary convictions, railways, corporations, public health, for example.³ In addition there was a thin literature on particular legal processes. Mandamusattracted the most because, lying to compel the performance of any public duty, it spanned a broad range of contexts.⁴ Prohibition, which was used most often to prevent ecclesiastical and admiralty courts encroaching on to common law territory, had a single text.⁵

There were public lawyers before there was public law. Robert Pashley had a large presence in reported litigation in the 1840s, noticeable in marginal decisions, Robert Vaughan Richards likewise, as Nicholas Fazakerley had had (p.487) during a formative period a century earlier.⁶ Their staple was the stream of local decisions emanating from justices, whether by an order, or by a summary conviction, or through quarter sessions. That had an extensive literature, traditionally in the form of compendia that immediately break down the promise of generality into specialized subsections. These included primary jurisdictions, where justices made and enforced their own orders, but also important secondary jurisdictions to enforce orders made by local commissioners and local boards who, through into the mid-century tended not to have their own enforcement powers. Questions would then arise whether justices had a supervisory jurisdiction over the commissioners’ and boards’ orders they were asked to enforce, which might themselves be reviewed by a superior court. The law on all this became extensive and homogeneous enough to generate a text by Scholefield and Hill in 1902 on how justices’ ‘proceedings and acts may be supervised and controlled’.⁷ Within its limited context it is a recognizable forerunner of de Smith and Wade.

Justices had no political responsibility, but the local boards and councils that took over local governance from them did, and ministers of the Crown more so. It takes a degree of abstraction and some political predisposition to bracket them together as ‘persons exercising statutory powers’, hence equally subject to judicial review. Judges could never escape asking whether political accountability should make a difference or whether that was a matter for Parliament to make explicit. They would sometimes struggle, as they did in the 1830s while deciding that though mandamus would not lie against the Crown it would against even the most senior political office holder, though (retreating) only in respect of duties owed to an individual, not duties owed to the Crown.⁸ But challenges to ministerial decisions were few. Those to the Local Government Board were directly authorized by statute, the Board being the successor to the Poor Law Commissioners, whose orders were amenable to statutorycertiorari, and those at the end of our period that were enveloped in political controversy concerned appeals to central ministries from local bodies, where at least some of the difficulty arose from the ministries’ ambiguous dual role.

This chapter offers a framework for the instances of review raised in previous chapters. It begins with the world of Pashley and Richards and ends with Rice, Arlidge, and the cases tentatively asserting judicial control over discretionary (p.488) decisions.⁹ Scholefield and Hill’s book did not go into further editions—which practically any law book does that sells. Instead, when F. J. Port wrote his pioneering study in 1929 his focus was on centralized authority.¹⁰ His book was different in another respect too, in that it was a product of a university rather than the bar, which may suggest another reason why there was no general administrative law until well into the twentieth century.

1. Eighteenth-century Foundations

In addition to mandamus and prohibition there was certiorari, which brought inferior courts’ records into the King’s Bench, freezing the lower court’s process and awaiting the bench’s own exercise of authority. That might be simply a matter of judicial administration—shifting the venue of a trial or enabling enforcement through the King’s Bench of a local civil court’s order against a defendant who had skipped its jurisdiction. Importantly, it was certiorari that brought to Westminster the many special cases stated for its opinion by quarter sessions, a voluntary substitute for appeal on point of law about which more will be said below. But it might also be to scrutinize the record for unlawfulness, a power the King’s Bench held to be within its own inherent jurisdiction.

Such reviewable unlawfulness might be a deficiency of jurisdiction or it might be an error of law within jurisdiction.¹¹ Jurisdiction signified legal authority within a denoted sphere, a competence dependent on person, place, subject-matter, and, to an extent, process. Review to check that those preconditions were satisfied was uncontroversial in principle, to the point that no-certiorari clauses in statutes did not prevent it.¹² For some legal processes judges required applicants to show that the error affecting a decision was jurisdictional, for example if they were seeking damages for its consequences or wanted amandamus to compel a (re)hearing. But certiorari and prohibition were broader, subject, however, to the (p.489) need for the error of law to be demonstrable from the formal record.¹³ Indeed, for review of summary convictions on certiorari the difference between jurisdictional and non-jurisdictional error was insignificant, since very probably no error was reviewable unless it was disclosed by the record.¹⁴ Justices returning a mere order in response to a certiorari, however, were not required to produce a full record of their proceedings, and in that case the rule was different: lack of jurisdiction could be proved by affidavit, though error within jurisdiction could not.¹⁵

It was axiomatic that a decision of an inferior court or executive officer made without legal power—without jurisdiction—was a nullity. If its execution had led to imprisonment, or a fine, or to sale or impounding of goods a chain of tortious liability arose, spanning those who did the seizing back to those who made the null order. This was a central element in the eighteenth-century rule of law. It meant that this form of unlawfulness could be tested by action before a jury, which judges would sometimes think preferable to proceedings by certiorari.¹⁶ Constables and other process servers enforcing apparently valid warrants had been given statutory immunity in 1751, and the issuing and enforcing justices given procedural protections designed to allow them to make graceful amends where they had transgressed.¹⁷ The basic principle, however, remained.

It was axiomatic too that anyone exercising a judicial power, which in this context included justices in all they did, must not have an interest in the subject-matter.¹⁸ If they, or just one of them, did the decision was void.¹⁹ It was a principle easily carried over to statutory commissioners.²⁰ Equally, however, it was one that statute might adjust to allow leading local figures to hold positions on bodies whose interests the judges might think conflicted.²¹

(p.490) A second central element of the eighteenth-century rule of law, fading by the time our period begins, vigorous then only in perennial complaints about the arbitrary enforcement of the Game Laws, was that the process of summary conviction by justices needed close judicial supervision from Westminster. It had been achieved by expanding what a record of conviction required—with the concomitant understanding that justices need write it up only when summoned by certiorari.²² Over the century the King’s Bench had seized upon what they held to be critical omissions from records so as to construct an ever more detailed law of summary criminal procedure in the interests of the accused.²³ Justices must show that there had been credible evidence underlying the conviction, that it was given in the presence of the accused, that it had not been given by someone with a pecuniary interest in a conviction, that it met the elements of the offence (properly understood), that it was admissible in law…and so on to the point where late in the century they were required to detail also the evidence for the defence.²⁴ A full ‘speaking’ record came to be a formidable document, scrutiny for error of law correspondingly intense.²⁵

It seems that these two elements of liability and intensifying scrutiny interacted. In theory they ought not to have done. From the Case of the Marshalsea in 1616, at the latest, only a decision made without jurisdiction was actionable, not one that was merely erroneous.²⁶ The difficulty seems to have been with the mechanics of certiorari. If a record of conviction had been removed to Westminster and quashed, and the accused who had suffered a penalty now sued the justice for redress, there was no formal record that could be proffered in defence whether or not the error had been jurisdictional—the reason would not appear on the absence of record, as it were.²⁷ This explains the Justices Protection Act 1803, a statute which, while influential, was sometimes treated as mysterious.²⁸ It provided that if a conviction had been quashed the plaintiff in a subsequent action could recover no more than the return of any penalty plus twopence damages unless he could prove malice and lack of probable cause.²⁹ If the defending justice (p.491) proved that the plaintiff was guilty of the offence he would receive nothing at all, for all that the conviction had been quashed. Contemporary understanding, according to Burn’s Justice of the Peace, was that the first of those provisions, at least, was limited in its conception. It applied ‘to those cases only where the justice improperly convicts’, that is, only where the mistake was within jurisdiction.³⁰

That reasoning explains why the 1803 Act applied only to convictions, and not to the many sorts of order justices could also make. Quashing mere orders for error within jurisdiction was rare because for them speaking records were not required. Successive chief justices in the 1720s and 1730s had tried to persuade their puisnes that records of orders should be made to speak in similar detail to convictions, but they had failed.³¹ The policy that had required close scrutiny of summary convictions was not carried over. So although an order would be set aside for error of law within jurisdiction if the justices chose to disclose it, they could not be required to provide the wherewithal for such scrutiny.³² This important limiting feature of certiorari lasted throughout our period.

That deficit in scrutiny of orders was made good in practice by the customary jurisdiction of quarter sessions to state a case for the opinion of the King’s Bench.³³ For it to arise a magistrate’s order had first to be appealable to the sessions, which needed express statutory provision. And there needed not to be a no-certiorari clause in the statute, since that writ was necessary to remove the proceedings into the King’s Bench.³⁴ Appeal rights from justices were common, however, both in civil and criminal matters, though the practical details differed so greatly from statute to statute that John Stone, whose work on petty sessions became a standard, wondered whether Parliament’s intention was deliberately to deter their use.³⁵

Once seised the sessions could state the facts for the opinion of the King’s Bench and seek a decision on the application of the law. That was just like an appeal on a point of law save, importantly, that it was voluntary. Sessions could not be required to submit a case to the King’s Bench, no matter how clear it was (p.492) to counsel or King’s Bench that they ought to.³⁶ ‘The facility with which these cases are granted varies very much in different Courts …’, wrote one commentator.³⁷ Nonetheless use was common, especially for the Poor Law settlement cases and the various rating questions that together made up the bulk of non-criminal work at most sessions.³⁸ As a later writer put it, it was a means whereby a sessions could choose to make its order speak.³⁹ It was required, however, that the sessions did find the facts; judges would become tetchy if they thought that that buck was being passed to them.⁴⁰ Fact-finding was pre-eminently part of the ‘merits’, reviewable only by appeal, for which an express statutory jurisdiction was necessary.

2. Dismantling and Rebuilding: The Decline of Review and the Rise of Appeal

Opportunity to challenge summary convictions had eroded before the 1820s. William Paley complained in 1814 that no-certiorari clauses and statutory short forms of conviction, which relieved justices of the need to demonstrate full legality, were encouraging injustice from haste, mistake, and preconception.⁴¹ John Adolphus, leader of the Old Bailey bar, wrote caustically of the ‘“holiday and lady terms” better fitted to the pages of a romance than an act of parliament’ in the Summary Proceedings Act 1822, which aimed to neuter certiorari by instructing judges to ignore defects of form after a conviction.⁴² Peel’s statutes in the late 1820s consolidating the criminal law instituted short forms and barred certiorari, usually substituting an appeal to quarter sessions.⁴³ Protection was being frittered away by the short forms, said Bolland B.⁴⁴ In the mid-1830s Stone thought it not worthwhile explaining review on certiorari ‘on account of the unfrequency with which summary convictions are now brought under the notice of judges’.⁴⁵

(p.493) Nor would judges allow this stunting of certiorari to be sidestepped by shifting into tort. They would not let defects that would have been revealed, had a speaking record been available, be brought up in an action against the justice—for example that there was no evidence to sustain his decision on a matter properly before him. That was reasserted in Brittain v. Kinnaird in 1819, a decision ‘oftener recognized than almost any modern case’, according to Coleridge J. in 1844.⁴⁶

There are suggestions that challenge was taking a different route. Stone thought it more usual for defendants believing themselves improperly dealt with to attack instead the subsequent warrant of commitment or distress by an action against the issuing magistrate—who may or may not have made the conviction.⁴⁷ They might allege a mistake in the warrant (there was room for pickiness here) or a jurisdictional defect in the conviction. Judges could show sympathy; in 1825 the Common Pleas reiterated that justices received ample procedural protection against surprise actions, and that it was only proper that those who had made errors should make amends.⁴⁸ Stone was sceptical, preferring the sections in Peel’s Acts that protected justices in those circumstances and regretting that recent legislation such as the Poor Law Amendment Act and the General Highways Act had not followed suit.⁴⁹ Archbold thought that justices were often intimidated by open threats of actions when called upon to issue warrants.⁵⁰

The reasoning applied equally to the issue of distress warrants to collect rates and other civic liabilities.⁵¹ Parke B. explained the theory later: because justices did not adjudicate upon the validity of the prior conviction or order, its validity must be a precondition to the exercise of their enforcement authority.⁵² That was an aspect of the eighteenth-century conception of the rule of law. Since there was ample room for disagreement about the interpretation of the statutes that created the offence or imposed the liability, as also about which elements were jurisdictional, there was risk enough to enforcing justices for the King’s Bench to show caution. If in a marginal case the justices declined to issue a warrant the court (p.494) would not order them to do so by mandamus.⁵³ The opinion of even the bench in mandamus would not validate that which was void.

These elements of liability, these possibilities for review, were removed in the 1840s. Archbold drafted a section for the Mandamus Act in 1843 immunizing justices who obeyed a mandamus from subsequent liability, removing their excuse for inaction.⁵⁴ In 1848 that provision was carried forward into a new streamlined statutory alternative to mandamus—just to justices—which was part of a systematic consolidation and rationalization of summary jurisdiction.⁵⁵ That was the work of John Jervis, then attorney-general, and spanned three Acts that collectively bear his name.⁵⁶ In short, having spelled out magistrates’ manner of exercising summary jurisdiction, his Acts made appeal to quarter sessions virtually the only practicable redress, save for jurisdictional error narrowly conceived. To that end they stipulated a standard short form of conviction, albeit one that was a little more informative than the skeletal versions that had sometimes found their way into law.⁵⁷ More importantly perhaps, since particular statutory short forms had been common for some time, Jervis’s third Act, the Justices Protection Act, snuffed out most uses of actions to raise issues of validity.

That Act rewrote its precursor of 1803 in more rational terms, extended it to orders as well as convictions, provided that defects in pre-trial warrants would usually be cured by conviction and that defects in post-trial coercive orders would be immaterial if their substance could be sustained by a prior, lawful, conviction or order. Justices issuing commitment or distress warrants consequent upon an order or conviction by another justice were not to be liable for its invalidity at all. Nor any longer could liability for a Poor Law rate be tested by bringing an action against the justice who issued the distress warrant.⁵⁸ Justices remained liable for the consequences of their own acts done without jurisdiction, but with the new requirement that the offending conviction or order should first have been quashed on appeal or review. Likewise, justices remained liable for malicious acts done within jurisdiction. So, to facilitate settlement and to protect justices from (p.495) publicity, the Act also re-enacted the procedural protections justices had enjoyed since 1751.

Thus Jervis’s Acts finally stultified certiorari as a means to address errors of law within jurisdiction and removed most of the opportunities to use tort actions against enforcers to reopen the validity of upstream decisions. Contemporaneously Baines’s Act, the Poor Law Procedure Act 1848, made quarter sessions the sole judge of the sufficiency of the paperwork that had brought appeals on removal orders to it.⁵⁹ Previously, its preamble claimed, disputes over what details should have been included in what notices had produced ‘much expensive and useless litigation…so that few cases of appeals against such orders are now decided upon the merits’, with the added baleful consequence that such disputes were going a further round by case stated to the Queen’s Bench. Now all that would be stopped; it was, as his friends put it, an ‘act for the better suppression of Pashley’, leader of that part of the bar.⁶⁰ A year later the Quarter Sessions Act applied these rules to appeals to sessions in nearly all other matters too and standardized appellate procedures—where the originating Act gave an appeal.⁶¹ Together all this legislation replaced an eighteenth-century conception of legalism, in which each step in a coercive process had to be independently valid, with instead an efficient appeal on the merits (if one lay at all) which, if the sessions could see that nobody had been substantially disadvantaged, would cure many errors or oversights en route.

Where no appeal lay there was room for review on narrow jurisdictional grounds only. That, however, was much in line with judges’ own preferences, as two well-known examples show. The first was the sequel to Stockdale v. Hansard, where, in 1839, the Queen’s Bench had determined that, whatever the outrage the House of Commons displayed, a resolution of a single House of Parliament was not ‘law’ enough to justify a libel outside the House.⁶² But when the Speaker ordered the two men who held the office of sheriff of Middlesex to be detained for contempt of Parliament—not saying in the warrant that the cause was their role in enforcing the courts’ orders contrary to the wish of the Commons—the court declined to release them.⁶³ The Speaker had a lawful jurisdiction to detain; if he chose not to let his order speak the court would not look behind its face.

In the second example, also in 1839, Coleridge J. allowed a certiorari to bring up a justices’ order giving parish officials possession of a house, because the parties (p.496) appeared to agree that the man being evicted was a tenant, whereas the justices’ process lay only against pauper licensees.⁶⁴ By the time the full Queen’s Bench heard argument on the order’s validity the parties’ apparent agreement had evaporated, their affidavits were irreconcilable, and since this was an order there was no record of the justices’ findings. The court declined to re-try the facts. Invoking Brittain v. Kinnaird, it held that matters such as this man’s status were to be taken as being within the justices’ jurisdiction to decide, and since their order was regular on its face it was valid. This is R v. Bolton, which has been argued over ever since; it would have been surprising if, in 1841, the decision had been otherwise.⁶⁵ In the following year Pashley induced the Queen’s Bench to order justices to send up their examination of witnesses in a settlement dispute though no case had been stated—a transparent evasion of the principle that a speaking record could not be required for a mere order—only to be rebuked when he tried it again.⁶⁶ Such a speculative novelty must not be repeated, the court said.

The problem in R v. Bolton, as the applicant told the court, was that the statute did not provide an appeal to quarter sessions. Where one did lie the Quarter Sessions Act 1849 provided, cautiously, that if both parties to a justice’s summary decision agreed, they could lodge an appeal to sessions but then state a case to a superior court at Westminster, ultimately enrolling the result in the sessions’ records as though it were a judgment of that court.⁶⁷ It became a common way of testing legal liability to rates.

The change away from begrudging access to the Westminster courts came in the more relaxed years of the 1850s, with the Summary Jurisdiction Act 1857, a government measure passed with no recorded discussion.⁶⁸ It enabled any party to a summary adjudication to waive appeal to the sessions and unilaterally apply by case stated to a superior court. It was limited to determinations upon information or complaint, and applicants had to move fast: the time limit was three days. But this unilateral power was new in this context, and the form was modern too—the Act explicitly said that certiorari was not needed to take the special case to Westminster. Prior to this Act, said Lord Campbell a year later, the(p.497) Queen’s Bench could not get at the merits of a decision within jurisdiction, and he then proceeded to do what the eighteenth-century court would have done on certiorari with a speaking record: assess whether the facts stated amounted to the offence charged.⁶⁹

From the 1849 and 1857 Acts together there followed numerous reported decisions on the validity of rates, on liability to market tolls and on similar issues requiring interpretation of local legislation, on the construction of the Public Health legislation and the validity of byelaws and orders made under it, as well as cases arising under the master and servant legislation and the Factories Acts. In 1879 the facility to appeal by case stated was extended to all convictions, orders, determinations, or other processes of any court of summary jurisdiction—closing the gaps in the 1857 Act—and by Order made under the Judicature Act the previous tight time limit of three days was relaxed to seven.⁷⁰ This was a nineteenth-century rule of law, pressing enforcement through justices and then integrating them with the Westminster courts.

Where enforcement or adjudication was not through justices, but through a new statutory agency then explicit statutory provision needed to be made if this, or any, process for taking points of law to superior courts were to be available. On principle, perhaps, certiorari would be available; but without a speaking record the Queen’s Bench could address only jurisdictional errors. There was an eighteenth-century precedent for statutory case stated, however, from the commissioners for house and window duties, interesting for not requiring the commissioners’ consent, and though such a provision was not universal even in taxing statutes it was copied frequently enough for Chantal Stebbings to conclude in a recent study that nearly all nineteenth-century statutory adjudicative tribunals had something like it, from the tithe and copyhold commissioners onwards.⁷¹ The technique was carried forward, with variations, into legislation such as the Housing of the Working Classes Act 1890 and (as a reluctant political concession) its reinforcement, the Housing and Town Planning Act 1909, where it applied to decisions of the Local Government Board on appeal against closure and demolition orders.⁷²

Case stated was obviously apt to raise questions of statutory application, but it was used also to determine how far a justice called upon to enforce an order made by some other agency could inquire into its validity or wisdom. This was (p.498) important as the powers of local boards on matters of public health grew, where justices’ sympathies might not align with the boards’. Generally the Westminster courts reduced justices’ supervisory role. Thus in the 1860s the Queen’s Bench held that justices called upon to enforce a local board of health’s order that a privy be replaced by a water closet could not revisit the board’s opinion that the change was ‘necessary’—though Blackburn J.’s doubts show that the statute could have been read to allow such a check.⁷³ A ratepayer could not argue that the rate levied under the Public Health Act was for works insufficiently ‘permanent’ to qualify, or that it was only other people who benefited.⁷⁴ A justice or arbitrator required to settle or enforce an apportionment order of the costs incurred in works compulsorily implemented under the Public Health Act must not inquire into its reasonableness, Erle CJ stressing the trustworthiness of an elected, responsible authority publicly audited.⁷⁵ With much greater difficulty, but similar result in the end, finality was likewise given to urban building lines prescribed by boards of works and their architects—there is nothing unjust about such a skilled public officer indirectly affecting property rights, said Lord Selborne in the decisive case, ending 20 years of judicial disagreement.⁷⁶ There was no appeal route against prescription of a building line, a fact that had weighed with some of the earlier courts, so this decision was particularly important.⁷⁷

3. Accommodating the New

4. The Virtue of Fair Procedure

Judges likewise showed a persistent assumption that the common law contained principles of procedural fairness that would apply to new forms of decision making too. They had long since imposed their values on the processes of summary conviction. As Paley wrote in 1814:

…even where the statutes are silent, as to any particular mode of proceeding, the law declares, that the magistrates to whom cognizance of offences is referred, are nevertheless bound to observe the rules of natural justice, one of which is, that the accused should have an opportunity of being heard before he is condemned. This is indispensably required….¹²²

In the 1790s Lord Kenyon CJ insisted that that rule apply also to justices issuing distress warrants to recover unpaid parish rates, against what seems to have been previous practice.¹²³ Given that in the mid-century so much enforcement was routed through justices, it was only a short but very significant step to hold, as the King’s Bench did in Painter v. Liverpool Oil Gas Light Co in 1836, that a justices’ distress warrant issued under a modern statute to a statutory utility company to recover an unpaid account was lawful only if the parties had first been summoned.¹²⁴ Further, just as a justice whose disregard for procedural requirements would be liable for damages in tort, so too was the gas company for originating and executing this unjudicial process.¹²⁵ Though a trivial error might be classed as a ‘mere irregularity’, a breach of natural justice took the perpetrator outside jurisdiction.

(p.507) The common law principle requiring a hearing before deciding had been applied to a much wider range of decision makers than merely the justices; Bagg’s Case (1615), which was a disfranchising of a borough freeman, and Bentley’s Case (1723), which was a deprivation of university degrees, were cited from time to time in nineteenth-century litigation just as they are in twentieth-century textbooks.¹²⁶ Lord Lyndhurst’s Exchequer applied the principle in 1832 to the sequestration of a benefice by the Bishop of London for the incumbent’s neglect of his parish—an important incursion into Church discipline at a sensitive time, but one justified both on principle, the court said, and by implication from the statute: a penalty for cause implied a hearing to determine the true facts.¹²⁷

Alderson B. later doubted that decision, because the statute enabled the bishop to decide of his own knowledge, but his lead was not followed.¹²⁸ Instead the line was drawn at offices held merely at pleasure, a function of the judges’ general unwillingness to be drawn into questions of discretion—of which more will be said below. The upper master of Darlington School, dismissed by the trustees in 1840, the relieving officer of Godstone Union, dismissed by the Poor Law Commissioners in 1849, Dr Hayman, headmaster of Rugby School dismissed in 1873, and the district surveyor of Bethnal Green, dismissed by the London County Council in 1915 all found themselves in this category, though in all cases there was no doubt something they could have said in their defence if given the chance.¹²⁹

In the twentieth century those cases might be seen as instances of employment, with the line to be drawn between contract and office. In the nineteenth all were seen as offices; whether they were terminable at will or only for cause depended on the statute or instrument that created them. Nor do judges seem to have brought social assumptions to bear on that question of interpretation.¹³⁰ Alfred Poole was but a junior curate at the Anglo-Catholic chapel at Pimlico when he antagonized parishioners of a lower-church persuasion by asking young women indelicate questions when hearing their confessions, but, the Queen’s Bench held, the Archbishop of Canterbury was nonetheless wrong to confirm the bishop’s revocation of his curate’s licence without giving him full opportunity to (p.508)present his case.¹³¹ The statute said nothing about process, and enabled the archbishop to decide as ‘shall appear to him just and proper’, but the court cast the function as appellate, hence requiring a ‘hearing’ at which the appellant could contest the evidence and produce his own.

The principle carried over into justices’ decisions that might be labelled administrative, most notably in R v. Totnes Union in 1845, where justices had ordered the union to maintain an elderly woman at home rather than in the workhouse without having summoned the guardians or an officer.¹³² And in a similar way mid-century judges decided that because orders by commissioners of bankruptcy were not judicial, and hence could be made ex parte, they could not be final in law.¹³³ That reasoning was to recur in the building line cases mentioned above, where one reason judges gave for allowing justices to reopen the question at enforcement stage was that district surveyors did not have an obligation to hear before deciding, whereas one reason for ultimately deciding that their line was final was that surveyors (and judges) now agreed that they did.¹³⁴ At its broadest, in the loose words reported in R v. Totnes Union, the obligation attached to any order that ‘affects the interests of persons in some way’ even though not a penalty or ‘direct burden’.

The principle’s breadth and the heterogeneity of the situations to which it was being applied may explain why the decision in Cooper v. Wandsworth Board of Works in 1863, which a century later became an obligatory point of reference, caused no excitement at the time.¹³⁵ The statute, Metropolis Local Management Act 1855, gave power to demolish houses, and said nothing about according a hearing, but it nonetheless seemed orthodox to require the Board first to tell Cooper of its intention and why—he had failed to give the statutory advance notice of construction that enabled the Board to check compliance with building regulations—and give him a chance to dissuade it. None of the judges doubted the importance of building regulation, or that the power to demolish existed in the circumstances, but those were not reasons for excusing the Board from a basic obligation that would cause it no discernible harm. The decision was quickly absorbed into textbooks on subjects as varied as partnership, arbitration, and local authority by-laws, but the ‘of course’ with which the editor ofAddison (p.509) on Wrongs greeted it speaks for them all.¹³⁶ Like the gas company in Painter, the Board had to pay damages for its unjustified trespass.

At least one local authority took the point and drafted its surveyors’ emergency statutory powers explicitly to exclude need for ‘presentment, notice or other formality’, which the Common Pleas took entirely at face value just as the Queen’s Bench had taken offices terminable at will in the Darlington School case.¹³⁷ More usually, and more importantly, a middle way evolved of providing an appeal to an administrative authority. It was used in public health legislation, where, rather than presenting a need to a justice, who would then hear and decide, an official might serve a notice requiring work to be done, with a short time limit, after which the authority might do the work itself and charge for it, or perhaps make a closing order or some such.¹³⁸ There was no statutory provision for the surveyor or authority to hear the property owner, but an appeal lay to the Local Government Board.¹³⁹ Meeting a similar power for the first time in 1889 Field J. said he hardly knew how to go about classifying it.¹⁴⁰ In 1890, however, the Queen’s Bench revisited Cooper, found that it did not mandate a particular style of fair procedure, and held that the opportunity for a hearing before the appellate authority sufficed without need for the authority to ‘hear’ as well.¹⁴¹

The critical question then became how much of what the courts had imposed upon justices they would now impose on these new administrative appellant authorities. They had not previously been inflexible. The reform of municipal corporations in 1835 had entailed the removal of office-holders, some of whom were to be compensated by the new councils. Anticipating disputes, the Act gave a right of appeal to the Lords Commissioners of the Treasury. The Queen’s Bench eventually decided that that august body’s unsuitability to find facts or apply law made it unlikely Parliament meant it to adjudicate whether a town clerk was a victim of the Act (as he claimed) or had been dismissed for misconduct (as the council claimed), and that therefore the Treasury’s role was confined to valuing (p.510) the office.¹⁴² However, on the assumption that the Commissioners did have jurisdiction they could decide on the papers and not, as the applicant wanted, only after a hearing adorned by counsel and witnesses.¹⁴³ Similarly, when the Queen’s Bench had ordered the archbishop to hear the curate, Alfred Poole, it had pointedly left the manner of hearing to his discretion—it too might be done on the papers, Lord Campbell suggested.¹⁴⁴

Like Board of Education v. Rice