In Sirros v. Moore all three judges concluded that judicial immunity should not depend upon whether a court were classified as superior or inferior.74 Lord Denning and Ormrod L.J. took the position that the classification recognized at common law ought to be abandoned in modern times, and that the same protection afforded to superior court judges ought to be extended to judges of all courts.75 It was the view of Buckley L.J. that the common law had never articulated a rule which distinguished between the two types of courts, but interestingly, he indicated that the rule was perhaps applied differently to judges of superior courts, by tending to classify their errors as errors of judgment rather than jurisdiction.76 It is not clear whether Buckley L.J. agreed with this tendency which he recognized, but his actual classification of the error at issue in Sirros might be viewed as an example of it.77 Surely, if a distinction in practice is justified it would be preferable to recognize it in theory rather than to disguise it in the classification maze.
The differing treatment of superior and inferior courts in the past again turns upon the elusive concept of jurisdiction. It has been said that a “...superior court is in one sense a court which is presumed to have jurisdiction until the contrary is established by evidence; in the other sense it is a court which cannot be restrained by prohibition from exceeding its jurisdiction”.78 Although the author did not specify, it seems he intended the term jurisdiction to include preliminary jurisdiction, and probably secondary jurisdiction as well.79 The argument would then be that as the arbiter of his own jurisdiction a judge of a superior court can never exceed it, but can only commit an error within it.80 Technically this is a tenable position, but as a matter of common sense a superior court is as much bound by law as any other court, and the rejection of this argument on that basis by Lord Denning81 and Ormrod L.J.82 is far more convincing.
Once that technical distinction is put aside, there is little reason for distinguishing between the two levels of courts. Often the functions of and precise subject matter dealt with by superior and inferior courts are identical, and often judges of both types sit in the same court.83 Even where the subject matters differ, as where an inferior court has a limited sentencing power or monetary jurisdiction, the difference is more of degree than of kind. Moreover, as Sirros indicates, the classification of certain courts as superior or inferior is by no means a simple and certain process,84 and it seems absurd that liability should depend upon such an irrelevant and elusive criterion. It is submitted that the abandonment of this distinction in both theory and practice is a welcome development.
Whereas the distinction between superior and inferior courts may have determined trespass liability at common law, the distinction between courts of, and not of, record85 may have fixed liability for malicious judicial acts. One author has argued convincingly from the older authorities that judges were absolutely immune from liability for malice when acting as a court of record.86 Leaving aside for the time being the questions of whether, and when, malice ought to be an appropriate basis for an action, it is interesting to consider the author’s rationale for the distinction. He argues that the common law has never extended immunity for malicious acts to officers, other than judges, who perform quasi-judicial functions,87 and he then suggests that judges not acting as a court of record more closely resemble quasi-judicial authorities than they do judges acting as courts of record. It is probably true that a distinction based, for example, on the power to imprison and fine, poses more of a difference in kind than the distinction between superior and inferior courts, which often poses only a difference in degree of the same type of power. Therefore, if the scope of judicial immunity were to depend upon a classification of different types of courts of law, the distinction between courts of, and not of, record is preferable to the superior — inferior court distinction for both trespass and case liability. But the argument in support of that distinction depends upon whether quasi-judicial officers ought to be granted lesser immunity than judges, a matter of some debate.88 Moreover, even if quasi-judicial officers are to be treated differently, Lord Denning’s arguments for treating judges of both superior and inferior courts alike would seem also to apply to judges in courts of, and not of, record.89
The first and strongest criticism of a classification system similar to the one Buckley L.J. suggests is that it is tremendously complex and difficult to apply in practice. One wonders whether liability ought to depend upon such ambiguous distinctions; and, if the distinctions really are that difficult to draw, there is a risk that the classification of error will more likely be used to support the outcome in a case than to determine it. Buckley L.J. almost admits as much.93 Consider his determination that the judge in Sirros must have put his mind to the question of whether to detain the alien, and then simply adopted the wrong procedure. It follows then that this was an error within jurisdiction, and immune. But he might have held that the judge lacked preliminary jurisdiction because he was functus officio. Or, he might have classified this as an error as to the extent of his jurisdiction, and then possibly have held the judge disentitled to immunity. Once the error is classified as jurisdictional of the second order, the further determination of whether the error was one of fact or law arises, and if of fact, whether justifiable or careless. It appears that the judge will be strictly liable for excess of preliminary jurisdiction and for excess of secondary jurisdiction produced by an error of law, but liable only for want of reasonable and probable cause with an excess of secondary jurisdiction produced by an error of fact. Surely there must be a more persuasive justification of these complex and uncertain formulas than the reasons in the judgment of Buckley L.J.
So the question which remains is whether, notwithstanding the complexity and uncertainty of the classification of error scheme, this classification is necessary in order to rationalize the degree of judicial immunity to be granted. Starting with the clearest situation first, consider the case where the judge has, and knows he has, preliminary jurisdiction to make an order, and knows what factual and legal determinations will support the order. The question before him is whether, in his judgment, he ought to make such an order. Clearly, if immunity is ever necessary to preserve free and independent judicial decision making, it is necessary in this case.
Next, consider the same situation, except assume that the judge, exercising his judgment within jurisdiction, commits a procedural error. This is how Buckley L.J. classified the judge’s detention of the alien in Sirros, in granting the judge immunity. It is possible to regard these two examples simply as variations of judicial acts done within jurisdiction and to immunize them both on that basis. However, while it seems highly desirable to support the free and independent exercise of judicial judgment on a substantive matter, it is not so easy to argue that the same freedom and independence ought to govern a judge’s choice of procedure. Nor is it accurate to classify an error of procedure as less important than an error of, for example, jurisdiction, per se; recall that the error which Buckley L.J. regarded as procedural was nevertheless sufficient to support habeas corpus.94 It might be preferable to deal with procedural errors within jurisdiction by arguing that such errors are not the cause of the plaintiffs loss. For example, the choice of the wrong procedure in Sirros was not the cause of the alien’s detention; indeed it was the basis of his release. Had the judge in Sirros really put his mind to the issue of whether he ought to vary the magistrate’s order and detain the alien, and had he then chosen the proper procedure, the alien would have suffered a longer detention than he in fact did. A party is entitled to procedural regularity, but it does not follow that irregularity will be the cause of actionable damage.95
It is beyond controversy, however one analyses the decisions on point, that provided a case can be classified as an error of judgment or procedure within jurisdiction, there will be immunity from tort liability. The extent to which this immunity ought to be extended to errors of jurisdiction, primary or secondary, is a more difficult issue to resolve. In Sirros, all three judges attached significance to the classification of an error as jurisdictional, arriving at two different propositions. Lord Denning would not immunize a judge who knowingly exceeded his jurisdiction, and he does not appear to distinguish between preliminary and secondary jurisdiction for this purpose.96 Buckley L.J. would not immunize any act in excess of preliminary jurisdiction, regardless of how that error occurred; and he would not immunize negligent errors of fact, nor any errors of law, going to questions of secondary jurisdiction.97 What is the reason for the distinction between jurisdictional and non-jurisdictional errors?
If questions of preliminary jurisdiction were always clear and easily resolved, the distinction might make sense,98 but a jurisdictional question can be as difficult as any other. It is therefore meaningless to speak of the “correct” jurisdictional determination as an obvious absolute.99 Potential liability is more likely to affect the accuracy of jurisdictional determinations adversely than positively.100
There is, however, a conceptual distinction between jurisdictional and non-jurisdictional questions which at first glance might seem to support a different liability rule. The exercise of judgment within jurisdiction is the very essence of a judge’s function, and it is that free and independent exercise of judgment within legal bounds which society values. No public interest is served by giving a judge freedom and independence to determine his own jurisdiction.
The force of that argument is somewhat dissipated, however, when one considers that in the first instance, it falls upon the judge to exercise his judgment on factual and legal questions which determine whether or not he has jurisdiction to perform the act in question. Notwithstanding the admitted qualitative difference between jurisdictional and non-jurisdictional questions, there is the same public interest in having the judge determine both matters free and independent of potential personal liability. Moreover, although appeal and judicial review will not usually compensate the victim of the jurisdictional error, they do protect the public’s interest in jurisdictional control.101 It is, after all, this public interest which distinguishes jurisdictional and non-jurisdictional questions; the aggrieved individual is likely to be indifferent to whether the source of his damage was an error within or without jurisdiction.
The real significance of the jurisdictional error lies in its relationship with the trespass action. Any interference with the person or property of another is a prim a facie trespass for which the defendant will be held strictly liable unless he establishes a defence. Liability is strict in the sense that any error as to the circumstances which give rise to the defence, however reasonable and innocent, will destroy the defence. The strength of the argument of Buckley L.J. in favour of strict liability for jurisdictional error is its consistency with the ordinary law of trespass as applied to private citizens: any error as to the scope of judicial authority is resolved strictly against the judge, and renders him liable in trespass. This is also the very weakness of the argument, because it entirely ignores the special judicial role.
Buckley L.J. reasoned that an act done without preliminary jurisdiction is not a judicial act, and therefore attracts no immunity. That proposition is more attractive as a matter of semantics than of logic or policy. Suppose counsel bring a case before a judge, argue at length over whether he has jurisdiction, and then the judge concludes after careful consideration of all the authorities on point that he does have jurisdiction. Should a higher court’s determination that he was in error deprive this process of its judicial character, let alone expose the judge to personal liability in tort? Surely he is performing a judicial function in determining the jurisdictional point in the first instance.
Moreover, it is the very fact that the judge has a public duty to make the jurisdictional determination in any case which comes before him, as well as a further duty to interfere with the person or property of the subject, which distinguishes him from the private citizen. It is one thing to affirm the value of personal security by permitting a private citizen to interfere with another only in exceptional circumstances, and entirely at his peril. Peace in the realm, and the integrity of its subjects are best preserved by discouraging such action. But surely it is an entirely different matter to propose that the same principle govern judicial officers who are obligated by the public interest to interfere with the liberty of subjects.
Recall that in Lord Justice Buckley’s view, liability for an error of law going to secondary jurisdiction would be strict, as with an error of preliminary jurisdiction, whereas an error of fact would be judged on the negligence standard. Although the distinction between errors of law and fact has been recognized in several older cases, there is no compelling reason for the distinction: it is nebulous at best, and there is no difference in the degree of potential difficulty of the issue or the severity of the consequences of error. It is desirable to eliminate the rehearing of all factual determinations in the subsequent tort suit, but it is equally desirable to eliminate the rehearing of legal issues of secondary jurisdiction.
Buckley L.J. does not attempt to rationalize the distinction between questions of preliminary and secondary jurisdiction, but Ormrod L.J. in obiter dicta indicates that, were inferior courts to be held liable for jurisdictional errors, then he would limit that liability to errors of preliminary jurisdiction in order to preserve some meaningful scope for judicial immunity.102 This theory seems reminiscent of a view once prevalent in administrative law by which jurisdiction meant preliminary jurisdiction, and nothing more. The erroneous exercise of a power within that jurisdiction might be appealable, but would not be subject to judicial review. The theory was that preliminary jurisdiction granted the jurisdiction to exercise the given powers, rightly or wrongly. But the distinction has now lost all significance in administrative law, and virtually any error which a tribunal might make may now be regarded as jurisdictional. Indeed, the modern trend is perhaps entirely to dismiss the requirement that an error be jurisdictional as a precondition for judicial review .103 In view of these developments in administrative law, the judge’s desire to limit the definition of the term jurisdiction with regard to liability is quite understandable.
This leads one to doubt whether the distinction suggested by Ormrod L.J. would be any more likely to survive in the law of torts than in the law of judicial review. Cynically, it might be pointed out that the courts’ expansion of the jurisdiction concept in administrative law may have been motivated by their desire to increase their supervisory powers over a growing segment of governmental regulation of society. In contrast; it might be supposed that they would narrow the concept to minimize their own liability risk. Even were this an accurate prediction, it would be undesirable to define the elusive concept of jurisdiction differently in public and private law.
There is, however, a less cynical and more relevant explanation for the expansion of the concept of jurisdiction in administrative law. It is apparent that the determination of preliminary jurisdiction may be just as difficult, and the consequences of error just as severe, as the determination of errors going to secondary jurisdiction or substance. The distinction between the two types of jurisdictional question is not responsive to the policies underlying judicial immunity, nor to the private interest of the party aggrieved. The distinction in administrative law depends not so much upon the inherent nature of jurisdiction as it does upon the intent of the legislative body which created the public authority: were the courts, or the authority itself, intended to have the final word upon a particular issue?104 This is a far less relevant consideration when the courts’ supervision of one another is in issue. Within this context, the distinction between preliminary and secondary jurisdiction is neither certain, nor meaningful.
In summary, most of the distinctions suggested by Buckley L.J. are not supportable, either with reference to the potential plaintiff or to the rationales for judicial immunity. The significance of the classification of an error as jurisdictional, even assuming the term may be defined with certainty, lies in its relationship to the trespass action. But the trespass action itself is objectionable on two counts. First, it seems entirely inappropriate to impose strict liability for jurisdictional error, however bona fid e and careful, upon officers who are under a public duty to commit prima facie trespasses. Secondly, jurisdictional errors are not uncommon, so the trespass action, properly interpreted, creates a disproportionately wide ambit of actual liability.
Actual liability has been controlled in the past by distinguishing between inferior and superior courts, and either immunizing superior courts for jurisdictional error, or classifying their errors as within jurisdiction.105 But neither the classification of the court, nor the elastic use of the concept of jurisdiction, can be supported on any other ground beyond its function as a crude limiting factor. If possible, it would be desirable to develop limiting formulas with more rational bases than these.106
It is interesting to note the legislative provisions in some jurisdictions whereby a judge in quashing a decision of an inferior court may grant an order protecting the judge in error from civil liability.107 These provisions serve a function in public law because they free the reviewing judge from considering the liability consequences to the judge below, should he determine a jurisdictional error had been committed. However, the provisions would be entirely unnecessary if the trespass basis of the action were abandoned. Moreover, these provisions appear to reflect an open acknowledgment that strict liability in trespass is an inappropriate basis for judicial liability. There is very little jurisprudence indicating when the protection order should be granted, but what there is demonstrates that the courts consider the standard of care and motive of the judge in error.108 There is one advantage to approaching improper motive or breach of standard in this manner, rather than directly through a specified action on the case: that groundless suits may be precluded at the earliest possible stage.109 But ironically, the trespass suit as defined without reference to this legislation will only be groundless if the judge has not committed a jurisdictional error, and provided the jurisdictional determination is a precondition of the tort suit, the protection order is unnecessary. Although in theory the basis of the trespass action based upon jurisdictional error is far wider than the basis of liability in case, in practice the effect of these protection order provisions is to make the basis of liability identical to that for errors within jurisdiction. The potential for numerous groundless suits is far larger in the latter case, where the protection order is of no assistance. Therefore, when the argument against the protection order is substantially the same as that for liability for improper motive or breach of standard, and the essence of both is really an action upon the case, it seems foolish to have a special procedure to govern jurisdictional errors when the legislation itself appears to recognize the irrelevance of the characterization. Instead, might it not be preferable to define carefully an action based upon breach of standard or improper motive, regardless of the classification of the court or the error?
Liability for negligent judicial decision making is easily rejected as an inroad upon judicial immunity, if for no other reason than that the potential for liability would be so great as seriously to impair judicial freedom and recruitment. It would be virtually impossible to determine the issue without practically retrying the original action, and there would be little or no deterrence of the groundless suit. There would also be a fairly large basis of actual liability. Judges routinely make decisions which adversely affect one party or another, and hum an nature is such that sooner or later even the most competent judge will make an error which could be classified as negligent in the sense that term is used in other contexts. The simple breach of the reasonable man standard is not the type of conduct which provokes m oral outrage requiring an institutionalized legal outlet for appeasement, and the deterrent value of negligence law is highly questionable.112 If compensation for the victims of judicial negligence is thought to be a worthwhile goal, it ought to be effected otherwise than by negligence liability.113
In Sirros v. Moore Lord Denning suggested that an act would lose its judicial character, and therefore its judicial immunity, if a judge knowingly exceeded his jurisdiction.114 That was a trespass case, and the idea was put forward as a basis for liability in trespass, but there is no reason to restrict it to a trespass theory, nor to suppose Lord Denning would have done so had he been faced with a pure economic loss claim.
Lord Denning draws no distinction between errors of preliminary and secondary jurisdiction for this purpose, and there is no reason why he should. There may be a reason for distinguishing between jurisdictional and non-jurisdictional errors in trespass theory by considering that an act within jurisdiction, erroneous or not, is always a judicial act. But this line of reasoning does not take us very far, because one can argue — and indeed the courts might very well so interpret Lord Denning — that no judge has jurisdiction to err intentionally in law or fact, within or without his stated powers. In principle, there is no reason to distinguish among different types of intentional error, and it would be far less complex to remove judicial immunity for any knowing error, jurisdictional or not.
The difficulty with a cause of action based upon knowing error is the converse of the difficulty with one based upon negligence: in the latter case the ambit of liability could be too wide, whereas in the former, the element of knowledge might be so difficult to prove as to render the action nugatory. Therefore, it is necessary to seek some middle ground between the two standards.
This suggests the possibility of basing liability upon a standard of gross negligence, or reckless error. In the United States it has been suggested that there is a distinction between acts in excess of jurisdiction and those wholly without jurisdiction. Liability in the latter case seems to correspond to a standard of gross negligence.115 The gist of the matter seems to be that some errors are so gross that no reasonable person acting judicially could possibly have made them. There is no reason why the same approach could not be employed in an action on the case, imposing liability where a judge “has knowingly erred as to law or fact, or both, or erred with reckless disregard of making the proper decision”.116
There are two difficulties with the recklessness standard. First, the term has no precise meaning, so there is a risk that it will create, as the negligence standard, too broad a basis of actual liability.117 However, judges, who are after all most familiar with the process, ought to be able to employ the standard in such a way as to distinguish it meaningfully from simple negligence. If necessary, liability could be further controlled by employing recklessness as an indicator of knowing error rather than as a basis of liability itself. The greater difficulty is with the groundless suit which might be invited by the vagueness of the term reckless. By its very nature, an allegation of reckless error cannot easily be struck down at the pleading stage, and it may not, therefore, provide a sufficient deterrent to potential suits. One possible solution would be to discourage suits through procedural provisions, perhaps by insisting upon the plaintiffs posting security for costs,118 and perhaps by imposing cost penalties upon the unsuccessful plaintiff. The breadth of liability can also be contracted by insisting, in addition, upon proof of improper judicial motive, but that approach also has its disadvantages.
