Retreating to the History of Judicial Review?

Introduction

Judicial review of administrative action is such an established feature of the Australian legal system that its existence might be taken for granted. Indeed, the High Court assured us that we have ‘an entrenched minimum provision of judicial review’, ¹ and that was in a case which deployed traditional interpretive tools to render a privative clause virtually pointless. That was in 2003, when privative clauses needed to be pellucidly clear before the courts would take them as meaning to have any impact on the availability of judicial review. In 2017, the High Court implied a limited privative clause into an Act that had none. ² What has changed?

Legislative drafters have devised several methods to undermine the ‘entrenched minimum provision’. Clauses that oust specific grounds of review have met with mixed success. Provisions stipulating that ‘no invalidity’ results from breach of specified statutory requirements have fared better, as have provisions making it more difficult to get at the evidence needed to establish a review ground, ³ and provisions requiring litigants to pursue their claims in the High Court’s original jurisdiction. ⁴ The drafting for each of these methods needs to be very clear, however, because until recently, judicial resistance has been palpable. That may also change, although it is impossible to generalise from a single decision. In 2017, the New South Wales Court of Appeal upheld the effectiveness of provisions designed to protect the Ombudsman from judicial review from any challenge other than for acting in bad faith. ⁵ Once again, what has changed — the relevant Acts or judicial attitudes?

The history of judicial review was important for one of the judgments in the first case (Probuild) ⁶ and critical for both of the principal judgments in the second case (Kaldas). ⁷ The issue in the first case concerned judicial review for non-jurisdictional error of law apparent on the face of the record, which was never envisaged as within the protection of the ‘entrenched minimum provision’. The second case alleged jurisdictional errors, and that did require a consideration of what exactly was entrenched. In broad terms, the court found its answer in history, saying that the Constitution entrenched only orders in the nature of the prerogative writs, and only so far as they would have lain against the respondent according to the law as it was at the time of federation. This could open up a Pandora’s Box. Well-devised privative clauses could presumably eliminate the possibility of judicial review for anything that would not have been subject to the supervisory jurisdiction in 1901, at a time when the principles of the prerogative writs had not been synthesised into ‘judicial review’ and when the legal system had yet to recognise the emergence of an administrative state. If that were to occur, we would need to learn two sets of laws — the 1901 version for statutory powers protected by more cunningly devised privative clauses and the current version for the rest.

Each of the two cases under review in this article also introduced a further (and more pragmatic) factor into the mix, namely, whether the Act in question gave would-be challengers a trade-off, replacing standard judicial review with a bespoke statutory remedy. What matters for that factor is not the fact of the courts’ traditional hostility to privative clauses, but the reason for that approach, which is to protect access to the courts for those seeking protection from unlawful administrative power.

It is still early days. This article will argue that the turn to history is misconceived and could do far more mischief than good. The pragmatic factor is attractive for responding to the underlying reasons for resisting privative clauses, although its deployment in the second of the cases here under review was unconvincing.

Recognising ‘Judicial Review’

Judicial review on non-constitutional grounds is a creature of the common law. Created in a form that starts to become recognisable to the modern reader by the end of the 17^th century, the common law courts have also been the direct authors of its ebbs and flows since that time. In the early days, they might have disguised some of their more creative moves behind a veil of technicality and dubious precedent, but it has always been possible to ascribe some purpose to the developments, some rationales or justifications that go beyond the existing precedents. Precedents are important, of course, because the application of the law should be reasonably predictable. However, judicial review has never been simply about its precedents, or, indeed, itself. Framed originally as a set of remedies, judicial review always needed to ask itself what needed a remedy, and the answers have always been dynamic, whether the movements be contractionary or expansionary.

Generalised concepts of ‘judicial review’, ‘public law’ and ‘administrative law’ emerged only in the 20^th century. The antecedents of ‘judicial review’ stretch back to the Stuart era, although Jason Varuhas argues that their common threads were not drawn together until well into the 20^th century; before that, it was a collection of remedies sharing distinctive procedures. ⁸ Varuhas was writing of the English history of the common law of judicial review. In a sense, the Australian history bears him out more clearly, with the enactment of the Administrative Decisions (Judicial Review) Act 1977 (Cth) (‘ADJR’). ADJR is long on grounds of review, but has only one remedy, the ‘order of review’. ADJR reflected a general impatience with remedial technicalities and a shift of focus to the grounds of review.

Varuhas also argues that the prerogative writs were originally focused on the supervision of inferior courts (in the case of certiorari) and the restitution of private property (in the case of mandamus). Controlling ‘government’, the ‘state’ or ‘public power’ as such were not conceived as judicial review’s core concern. ⁹ If he is right, then this could become important, because there is a view that Australia’s constitutional entrenchment of judicial review is limited in two ways. First, some argue that it is limited to the named remedies — prohibition, mandamus and injunction. Secondly, some people argue that the law that is entrenched is that which existed at the time of federation.

Modern-day judicial review has two broad components. First, it consists of a number of judicial remedies for the unlawful exercise of government power, and secondly, it comprises a set of general standards or principles for the exercise of that power. In Australia, those standards or principles are treated as presumptions of statutory interpretation, to be implied into statutory grants of administrative power unless the Act gives some reasonably clear indication to the contrary. The standards are not tied to the remedies, but they are most commonly discussed and elaborated in that remedial context. As principles of statutory interpretation, they are a gloss on statutory powers that make no reference to the requirements of reasonableness, adherence to natural justice and so forth. The gloss is often justified on the ground that legislation is drafted on the understanding that the gloss will apply — some things go without saying. At other times, however, the judicial glossator is at odds with the drafters, leading to a situation recently described as a ‘power struggle’ between the courts and Parliament. ¹⁰

All sorts of high-level theories are on offer to explain, justify or rationalise both the remedies and their underlying standards. ‘Control’ of government power is often advanced, ¹¹ as is the rule of law, ¹² but these prompt questions as to the purposes of control and the content of the rule of law. The answers all contain functionalist and pragmatic elements, because the administrative state which lies at its centre is itself so variable. Further, judicial review varies not only between current regulatory contexts, but also across different periods of time.

Casting ‘administrative law’ as an accountability mechanism, Professor Mashaw said that the enduring (and ultimately irresolvable) tensions shaping his subject were ‘political accountability to elected officials; hierarchical or managerial accountability to administrative superiors; and legal accountability to individuals and firms through judicial review’. ¹³ Mashaw explained that the weighting of each opposing tension inevitably varies between contexts and political periods. America’s ‘political accountability’ to elected officials might not equate to Australia’s political accountability to the Parliament. Making the necessary adjustments, therefore, one might suppose that one of the variables in the Australian mix is the extent to which the law attempts to shield unjust decisions from tribunal scrutiny and Ministerial responsibility to the Parliament.

Over the last few years, however, a number of cases appear to have reverted to earlier times, before the development of unifying concepts such as judicial review, public law and administrative law. It is as if there are now two bodies of judicial review law and its underlying purposes — the modern law with its doctrinal developments of the 20^th century, forged in response to the emergence of the administrative state; and a pre-modern law as at 1901 that applies if Parliament strips back judicial review as far as that is constitutionally possible. The creation of a parallel universe is a development that seems to be at odds with the High Court’s decision in Re Refugee Review Tribunal; Ex parte Aala. ¹⁴

The Immigration Minister had argued in Aala that s 75(v) of the Constitution did not apply to administrative decisions made in breach of natural justice, because such breaches had not amounted to ‘jurisdictional error’ back in 1901. Gaudron and Gummow JJ (with whom Gleeson CJ agreed) said that it was indeed unclear whether natural justice violations were jurisdictional errors in 1901. Even if they were independent heads of complaint for prohibition, that had only limited relevance; it was not decisive. Their Honours said that what mattered were the ‘essential characteristics’ of the constitutional writs, and that history was relevant but not decisive in that regard. ¹⁵ They reasoned that:

The law was in a state of development. The doctrinal basis for the constitutional writs provided for in s 75(v) should be seen as accommodating that subsequent development when it is consistent with the text and structure of the Constitution as a whole. ¹⁶

Hayne J agreed, stating repeatedly that the scope and meaning of s 75(v) were not frozen at 1900. ¹⁷ Kirby J was even more dismissive of the Minister’s appeal to history, relying instead upon his view of the constitutional purpose of s 75(v) to determine the scope of the ‘essential characteristics’ of the constitutional writs. ¹⁸ Rather unfairly, his Honour castigated the ‘tedious and largely unilluminating examination of nineteenth century case books’. ¹⁹

Aala’s reading of s 75(v) of the Constitution was in an important sense ambulatory. The remedies entrenched by s 75(v) come from the common law and equity, two sources of judge-made law that necessarily evolve, and those evolutionary developments flow into s 75(v) itself.

The New South Wales Court of Appeal took the opposite approach in Kaldas, ²⁰ without mentioning Aala. I will discuss Kaldas in some detail below. It suffices for now to observe that if two bodies of judicial review law were indeed to emerge, it would be because of a very restricted approach to the question of what exactly is entrenched by s 75(v) of the Constitution — the remedies with their associated principles as developed from time to time or the law and remedies as applied at federation. Other variations are possible. Kaldas concerned the supervisory jurisdiction of the Supreme Courts of the States, which the High Court had declared in Kirk v Industrial Court (NSW) to be entrenched, not because of s 75(v) (which applied only to the High Court), but because s 73 of the Constitution preserves the Supreme Courts and such of their defining characteristics as they bore at federation. ²¹ Judicial review for jurisdictional error is one of those defining characteristics. ²² Although the High Court has not stated this explicitly, the entrenchment founded upon s 73 appears to align in practical effect with the entrenchment founded upon s 75(v), and this article proceeds upon that assumption. If I am wrong, however, then Kaldas beckons with another possibility, namely, that the s 75(v) jurisdiction is ambulatory while the s 73 jurisdiction is not. The same arguments against the emergence of two sets of supervisory jurisdictions (frozen and evolutionary) would still apply. To understand that which is frozen (whether that be federal or State), we would need to understand how unsystematised the law was back in 1901.

A Preliminary Disclaimer

Popular history has Churchill declaring that: ‘History will be kind to me, for I intend to write it.’ That appears to be apocryphal, a misquotation, ²³ but no-one really cares — it is difficult to dislodge popular history, especially when it is so straightforward.

There are, of course, different ways of writing history, many of them less self-serving, unidirectional or triumphalist than that attributed to Churchill. Histories of law and legal institutions approach their subjects from a variety of standpoints, although legal history written for lawyers tends to focus on legal institutions and legal doctrine. Histories abound, however, written from different standpoints, be they social, cultural, religious, economic, gender, racial, political, colonial or postcolonial. Different lenses add to our understanding of the complexities, foibles and struggles of our different pasts, and sometimes, these might point to different possibilities for our future.

Constitutional scholars, for example, have devoted considerable attention to the oddity of s 75(v) of the Constitution investing the High Court with original jurisdiction to hear claims against Commonwealth officers for mandamus, prohibition and injunction, but not for any of the other remedies (especially certiorari) typically associated with judicial review. Differing explanations have been offered in doctrinal terms of exactly why s 75(v) names those three remedies, and only those three, depending to a considerable extent on understandings of how those remedies operated in England and the United States in the period leading to federation. ²⁴ This article will not pursue that branch of historical scholarship, which necessarily leaves several questions unanswered. Perhaps more importantly, the scholarly constitutional histories of s 75(v) are written to help explain the provision’s very existence, and its peculiar drafting choices. Explanations were seen as needed not to advance an argument for a return to the law as it was at 1901, but because the changes in the remedial law since then are what make the provision look so puzzling.

Judicial review on non-constitutional grounds has its own long and tangled history, and I am no legal historian. The twists and turns in the history of what we now regard as ‘judicial review’ stretch back at least to the beginnings of the 17^th century. The oldest of those materials are so arcane as to have become the preserve of specialist historians, equipped with skills and learning far beyond those of most modern practitioners or text writers. For the early beginnings of the subject, we are necessarily dependent on these historians. Their source material consists of far more than the bare bones of unofficial law reports. The reports were all unofficial until the 19^th century, and the 17^th century judicial review cases were often reported in mangled Latin, French and English. Not only that, the earlier case reports sometimes consisted of only a few lines. Furthermore, the historians provide us with different interpretations and emphasise different social, political and doctrinal tensions.

Any history in what follows, therefore, is offered with no pretence on my part to have done my own original research. For anything before the 1850s, I am largely reliant on de Smith’s first edition ²⁵ and on monographs by Edith Henderson, ²⁶ Amnon Rubinstein ²⁷ and Paul Halliday. ²⁸

From Individual Remedies to Judicial Review

Until some time into the 20^th century, the history of judicial review is a collection of histories of a subject’s remedies against unlawful government action. Even the High Court’s entrenched jurisdiction in s 75(v) of the Constitution is framed in terms of three named remedies (mandamus, prohibition and injunction), rather than in terms of ‘judicial review’. Perhaps s 75(iii) (matters in which the Commonwealth is a party) could encompass judicial review more generally, although that possibility awaits further exploration.

For some considerable time, therefore, there was no over-arching concept of ‘judicial review’. In what is now called ‘tort’ (and even that is a 19^th century invention), there were claims against public entities for trespass, battery and false imprisonment, based upon allegations of excess of power by public actors. The defendants’ publicness, however, was irrelevant. In Janet McLean’s terms, it was ‘public law without the state’, ²⁹ and until statute subjected the Crown to tort liability, that was the only way of getting at justice. One can imagine some taxonomists debating whether those old tort cases belonged properly to public or private law.

And then there were claims for the prerogative writs of certiorari, prohibition, mandamus and habeas corpus, whose modern uses started to become reasonably recognisable only some way in to the 17^th century. ³⁰ Certiorari, for example, had long been used to call for records, so that they might be subjected to administrative audit. King’s Bench also used it as part of its turf warfare against competing courts, commanding the removal of cases into King’s Bench for hearing. Certiorari to quash, however, came later.

Judicial review’s history, therefore, consists in large part of a history of particular forms of action (mainly what are now seen as ‘tort’ actions) and particular prerogative writs (mainly what are now seen as claims for ‘judicial review’), rather than of generic principles applicable to them all. Until the early 20^th century, each prerogative writ, for example, had its own technicalities, including its restrictive rules about standing to sue, its limitations upon admissible evidence and cross-examination, the nature of the institutions or individuals to which it could be directed and the nature of the functions which it might supervise.

By the beginning of the 19^th century, certiorari went only against courts of limited jurisdiction. Although the meaning of ‘jurisdiction’ (and its more recent cognate, ‘jurisdictional error’) was still quite unsettled, ³¹ it clearly bore a wider meaning in claims for certiorari than in tort actions, ³² and it had some intuitive appeal to a professional audience generally. Lawyers have long spoken of a variety of limitations upon an inferior court’s ‘jurisdiction’. Such a court might be limited to claims under £50, or forbidden from determining title to land. Sometimes the jurisdictional limitations were territorial (eg, claims arising within Middlesex). By contrast, mandamus was never tethered to the supervision of inferior courts; rather, its indispensable ingredient became the enforcement of a ‘public’ duty. The language of ultra vires sat most easily with entities other than natural persons, invested with powers that bore no resemblance whatsoever to courts (eg, statutory corporations responsible for maintaining bridges or setting toll rates). The difficulties with a court/administrator divide, however, became unmanageable in the light of three sets of tectonic shifts.

First, the divide itself was weakened beyond intellectual coherence by a flood of cases extending the reach of certiorari and prohibition ever further beyond the territory of inferior ‘courts’ strictly so called, so as to include adjudicative or court-like functions. These were usually labelled ‘quasi-judicial’. Professor de Smith’s classic treatise was written for an English audience, but its early editions had some influence in Australia, at a time when our courts followed English law more closely than is now the case. Chapter two of the first edition (in 1959) was devoted entirely to the ‘Classification of Functions’, ³³ emphasising its structural importance (in de Smith’s view) to the entire subject of judicial review. Chapter two remained in place (albeit with varying page lengths) for the next two editions by de Smith, and for the fourth edition by Professor Evans, produced after de Smith’s death. ³⁴ Mason J said that for Australia, ‘the modern view, now received doctrine, is that the classification of powers is not a sound criterion for the operation of precise rules of law’. ³⁵ The new team writing the fifth edition were of the same mind, admitting that they had found it ‘tempting to omit entirely the former Chapter 2 of this work’. ³⁶ However, they thought that it retained historical interest. Making no attempt to update it, they added a couple of pages explaining why it had passed its use-by date, trimmed it and shifted it from pride of place near the front of the work, to an Appendix. ³⁷ In practical terms, they had preserved de Smith’s Classification exercise in aspic, for the history buffs, and there it remains. ³⁸

Secondly, the rules of natural justice followed suit, extending to quasi-judicial functions. The old rule was that natural justice applied only to judicial (and then quasi-judicial) functions. ³⁹ Ridge v Baldwin declared that restriction to be unnecessary, ⁴⁰ and Australia followed suit soon after. ⁴¹ By this stage, the old Classification exercise had reduced in size from an overall structuring device for the whole of judicial review, to an exercise relevant only to the scope of natural justice and certiorari.

Thirdly, the other grounds of judicial review followed the expansionary drift of the natural justice grounds. They were no longer assigned to only one category (‘adjudicative’) or the other (‘administrative’), ⁴² although, of course, the nature and function of the decision-maker still influences the extent to which it is appropriate to apply any particular generic review ground, and if so, the intensity of superior court scrutiny. This meant that narrow conceptions of ‘jurisdiction’ applicable to inferior courts lost their traction. In particular, reviewable decision-makers, whether they were adjudicative or administrative, could now commit jurisdictional errors (such as breaches of natural justice, or failing to take mandatory considerations into account) in the course of a process which they had authority to hear and decide. ⁴³

The disappearance of any substantive difference of outcome between ‘jurisdictional error’ and ultra vires roughly coincided with the development of general principles of judicial review and a substantial increase in the overall reach and depth of the grounds of ‘judicial review’. The term itself now appears to be interchangeable with ‘administrative law’ in England, where the search is turning to the development of a useful set of principles of ‘good administration’.

These shifts of analytical focus from judicial review’s remedies to its general principles should come as no surprise. They were part of a greater shift from the compendium or digest style of legal writing and analysis, to a search for overarching principles or syntheses, to be both derived from and exemplified by the decided cases. The search for general principles connecting formerly discrete forms of action is well-documented. ⁴⁴

In one sense, the development of general principles of judicial review has led to simplification. The technicalities of the older writs are disappearing from the radar screens of most practitioners and rarely get much attention from text writers. Getting away from the old technicalities was also a deliberate choice made by the framers of ADJR, ⁴⁵ which has only one remedy — the order of review.

None of this should be taken as implying, however, that substantive principles are any easier than remedial technicalities. With great prescience, for example, Professor Rubinstein wrote in 1965 that the expansion of the grounds of judicial review should not lead automatically to a corresponding expansion of government liability in tort. ⁴⁶ The principles of false imprisonment are currently in a tangle, as the courts struggle to explain their reluctance to grant not just substantial damages, but any damages at all, to those whose detentions by public authorities were ‘unlawful’ only with the benefit of considerable judicial review hindsight. ⁴⁷ That particular tangle dates back to a tort defence laid down in the early 17^th century, to the effect that judges and those acting in reliance on their orders were liable only if the proceedings were so devoid of ‘jurisdiction’ (coram non judice) as to have been a nullity. The prerogative writs were discretionary, but their usage of ‘jurisdiction’ was generally far more flexible, and produced outcomes which in a considerably later period also came to be characterised as nullities. ⁴⁸

Viewed from other perspectives, the prerogative writs were as much about a power struggle between the Royal courts at Westminster on the one hand and specialist or local courts and tribunals (sometimes administering their own forms of law) on the other. ⁴⁹ Professor Harry Arthurs saw the expanding reach of certiorari and prohibition as the contribution of the Royal courts at Westminster to a determined assault upon legal pluralism, replacing it with standardised laws and a standardised (and centralised) system of courts for the entire country. ⁵⁰

From other perspectives, the writs were the vehicles for the development, not of process standards for the exercise of public power, but of substantive rights. Professor Paul Halliday’s history of habeas corpus, for example, depicts modern conceptions of the liberty of the subject as emerging from the assertion by King’s Bench of power over its rivals and the setting of ‘incisively vague’ limitations to custodial powers deriving from status (such as husband and wife). ⁵¹

Revisiting the History and Rationales of Certiorari

These shallows of legal history are now becoming the subject of judicial interest. The current uses of certiorari are well-known. Its form is to quash the decision under attack. The only decisions that are amenable to certiorari are those which have themselves altered a person’s legal rights, obligations or status. ⁵² Certiorari cannot quash a decision whose strict legal effect is to do no more than seriously damage a person’s reputation, even though the decision is flawed by jurisdictional error, such as a failure to accord natural justice. People in that position have limited options. They might get in early, seeking prohibition (on jurisdictional grounds) or injunction (on the ground of legal error) to stop the decision being made unless or until the defect is fixed. ⁵³ If it is too late for that, then they are left with a claim for declaratory relief.

The grounds for certiorari are all types of jurisdictional error, plus non-jurisdictional errors of law if they appear on the face of the record. For convenience, I will refer to certiorari in the latter circumstance as ‘non-jurisdictional certiorari’. Cases are usually run on affidavit evidence, but there are no special restrictions on the means of proving jurisdictional errors. Non-jurisdictional errors, however, must be errors of law, and they must appear on the face the ‘record’. The common law imposed substantial restrictions on that branch of certiorari. Administrators and tribunals are under no common law obligation to give reasons for their decisions, ⁵⁴ and any reasons that they do provide are not part of the ‘record’. ⁵⁵

Legislation extends the reach of non-jurisdictional certiorari in a number of ways. Statutory duties to give reasons are now ubiquitous, and legislation in two States deems reasons to be part of the record. ⁵⁶ In addition, quashing orders under ADJR and its equivalents are available for any error of law, with no requirement that they appear on the record or that they be jurisdictional. ⁵⁷

This thumbnail sketch of certiorari’s current status is uncontroversial, although Gageler J’s judgment in Probuild cast two major doubts on the availability of non-jurisdictional certiorari. ⁵⁸ The first of these was based in history, and the second proceeded from the view that non-jurisdictional certiorari poses an intellectual paradox. His Honour’s arguments were extended and complicated, and it is hoped that the following summary does them justice.

Certiorari was initially confined to the supervision of courts or, at least, justices of the peace and other judicial officers. In the 20^th century, its field of operation was extended to decisions or functions of a judicial or quasi-judicial nature, ⁵⁹ until even that restriction was dropped in consequence of the decision in Ridge v Baldwin to drop the same restriction which had to that point applied to confine the scope of the principles of natural justice. ⁶⁰ Gageler J had no argument with the extension of certiorari to administrators and tribunals, but only so far as the ground was jurisdictional. He said that until the early 1950s, non-jurisdictional certiorari had (for centuries) ⁶¹ issued only against inferior courts of record, ⁶² and even then, only if the judicial review court had concurrent jurisdiction with the inferior court, enabling the review court to resolve the matter. ⁶³ His Honour said that quashing a decision for non-jurisdictional error of law did not free the inferior court to re-decide, because its job was done — it was functus officio. ⁶⁴ Gageler J said that the English Court of Appeal had overturned too much history, when in R v Northumberland Compensation Appeal Tribunal; Ex parte Shaw, ⁶⁵ it had subjected administrators and tribunals to non-jurisdictional certiorari. His Honour hinted that the reason why Australian courts followed the Northumberland line was that they were accustomed to tugging their collective forelocks to their former masters in England. ⁶⁶

Gageler J’s attack on Northumberland went further than ancient history. His Honour also drew on subsequent decisions, which had made Northumberland entirely redundant in England (which now says that all errors of law are jurisdictional), ⁶⁷ and of reduced significance in Australia because such errors are arguably more likely to be jurisdictional when made by administrators and tribunals. ⁶⁸ The Australian position is a consequence of statutory interpretation, which provides the foundations for identifying jurisdictional (and therefore invalidating) errors. ⁶⁹ It all comes down to statutory interpretation, but if the legal error is non-jurisdictional, then that must mean that its author was acting within their authority, which for Gageler J makes its invalidation either ‘anomalous’ ⁷⁰ or an accident of history which might one day be abolished, but never extended. ⁷¹

Edelman J disagreed with Gageler J’s historical account. ⁷² His Honour accepted Northumberland’s claim to historical continuity. In his view, the writ had long gone to administrators and tribunals and had applied no consistent distinction between jurisdictional and non-jurisdictional legal errors until the 1840s, when a time-based or ‘commencement theory’ became prominent (albeit more honoured in the breach than the observance) ⁷³ until its formal abolition in England ⁷⁴ and Australia. ⁷⁵ His Honour also said that there was nothing anomalous about non-jurisdictional certiorari; it did not amount to legislative authority to go wrong, or a legalisation of illegalities. It is, more simply, a case of an error of law that is reviewable, albeit committed within the decision-maker’s authority. ⁷⁶

It is submitted that the historical record provides insufficient support for Gageler J’s history. Henderson ⁷⁷ and Rubinstein ⁷⁸ are at odds as to whether there was any settled distinction between jurisdictional and non-jurisdictional errors of law until the 1840s. ⁷⁹ Henderson thought that the courts manipulated the term ‘jurisdiction’ to suit their desired outcomes. Rubinstein thought that the distinction emerged as a device to ‘read down’ the effect of ‘no certiorari’ clauses. Edelman J was right to note that certiorari had long been issued against entities that we would not call ‘courts’, but those bodies were usually comprised of judges and justices of the peace.

These aspects of the differences between the two judgments at least remain debateable, but it is submitted that Gageler J’s elision of certiorari to hear with certiorari to quash is unsound. Certiorari had several uses; it seems to have started life simply as an order to ‘get me the file’. ⁸⁰ Certiorari to quash did not emerge until the 17^th century, ⁸¹ roughly a century after King’s Bench started using certiorari to hear. ⁸² The latter form of certiorari was one of several methods that King’s Bench had used in its competition for dominance over local and specialist forums. ⁸³ Certiorari to quash developed by analogy with the writ of error, and its distinctiveness from certiorari to hear had become quite clear by 1700. ⁸⁴ In reality, certiorari to quash was a common law appeal device, fashioned centuries before the statutory conferral of generalised appeal rights from the magistracy. Further, Gageler J’s view that the author of a decision vitiated by a non-jurisdictional legal error was functus officio introduces a novel concept of quashing, and is unsupported by the two cases cited in support. ⁸⁵ His Honour’s view that certiorari to quash for a non-jurisdictional legal error issues only if the reviewing court can hear the matter on the merits is similarly unsupported by the case that he cites. ⁸⁶

Policy Arguments for Keeping Non-Jurisdictional Certiorari

Gageler J said in Probuild that abolition of non-jurisdictional certiorari must await another day. In reality, however, that was only because no-one else signed up to his elision of non-jurisdictional certiorari to quash with certiorari to remove and hear. Were he to carry the day, then he will have achieved a de facto abolition. The question which now arises, therefore, is why others might agree with the abolitionist goal. Of all the answers on offer, history is the least convincing. The history is not only unclear, but it is a history of the common law, and there is no reason to suppose that common law understandings informing constitutional provisions should remain frozen in time. However, there are other arguments for abolition.

It is indeed difficult to explain non-jurisdictional certiorari in terms of the oft-stated functions of judicial review, because the decision-maker has not exceeded his or her power. The practical effect of non-jurisdictional certiorari is to serve as a common law appeal for a limited range of errors of law, and this flies in the face of the traditional mantra that appeals are the creature of statute. It is undeniably an accident of history, but one might argue that it is a play on words to call it ‘anomalous’. The charge is that it is anomalous because the impugned decision is valid and is therefore ‘authorised’ or has legal effect. However, there is neither a logical contradiction, nor a legal anomaly, in giving full force and effect to legally flawed decisions, whether those be decisions of a court or administrator. ⁸⁷

The position would become more complicated if the Act that conferred power on the administrative body not only gave full effect to any legally flawed decision that the body might produce but also provided that that body’s decisions on how its Act should be interpreted were to be taken as final for all purposes (including judicial review). There would in that case be a true anomaly, because Parliament lacks (it is submitted) the power to strip the courts of all powers of statutory interpretation. That, however, prompts a further question as to whether Parliament can strip the courts of some powers of statutory interpretation. In a sense, a clause ousting non-jurisdictional certiorari will usually achieve that outcome. These are arguments that might apply only to Commonwealth laws, because State legislatures are not bound by the Constitution’s separation of the judicial power. However, that way madness lies, and there is nothing to indicate that Gageler J (or any other Australian judge) would go down that path.

The judicial power of the Commonwealth is said to be quarantined from the legislative and Executive branches, but that quarantine is not complete. Some powers, for example, are said to have a chameleon quality, because their characterisation as ‘judicial’ or ‘executive’ depends upon the branch in which they are reposed. ⁸⁸ Australia’s retention of the concept of non-jurisdictional error of law could be seen as another way in which the High Court avoids some of the more unworkable consequences of an overly strict adherence to the separation of the judicial power. Indeed, that could be one of the reasons why Kirk had said that the concept of non-jurisdictional error of law had ‘utility’ for constitutional reasons. ⁸⁹

American constitutional lawyers are all aware of Marshall CJ’s foundational statement in Marbury v Madison: ‘It is, emphatically, the province and duty of the judicial department to say what the law is.’ ⁹⁰ Brennan J’s adoption of that statement into the Australian law of judicial review of administrative action is equally well-known, ⁹¹ but in that context, it does have the potential to mislead. No-one now disputes Marshall CJ’s proposition that it is ‘emphatically, the province and duty of the judicial department to say what the law is’. Nor is it doubted that to every question of law, there can be only one right answer. ⁹² However, it is neither the duty of the executive branch to refrain from saying what the law is, nor the province and duty of the judicial branch to have the last word on all question of legal interpretation. ⁹³

Current debates in the United States serve as a timely warning against adopting a stricter separation of the judicial power. American administrative law is currently in a ferment not seen since the 1930s. Everyone has an opinion — administrative law specialists, regulators, historians, constitutional lawyers, Republicans, Democrats, and even the New York Times. Central elements of the role of the judicial branch in administrative law are under question, with some doctrines under threat of legislative reversal. Some of administrative law’s principal operational devices are under fire. These include the very legitimacy of agency rule-making, ⁹⁴ let alone the processes for making rules. ⁹⁵ There are concerted efforts to reduce the sheer bulk of rules. ⁹⁶ Judicially created doctrines are also under fire. These include doctrines requiring judicial deference to interpretations by agencies not only of their own rules ⁹⁷ but also of their governing statutes. ⁹⁸ Bills to overturn these doctrines appear to have stalled at the time of writing, ⁹⁹ but the pressure remains. The hostile fire comes from various perspectives, such as the principles of statutory interpretation, debates about the relative institutional competence of agencies versus courts, whether agencies should be more electorally accountable (and if so, whether to the President or the Congress), and whether these and other operational doctrines of administrative law are much more than a cover for partisan politics. The debates are extremely heated. ¹⁰⁰ From this distance, at least, it would appear that those pushing for a stricter separation of powers are using that argument as the stalking horse for attacks on the legitimacy of the administrative state itself. ¹⁰¹ We should not go there.

Therefore, neither history nor the constitutional separation of powers doctrine is an appropriate platform for the abolitionist cause. Perhaps more promising is the modern-day mantra that all ‘appeals’ (at least, in crime and common law matters) are creatures of statute. ¹⁰² However, even that depends on how far back one is prepared to travel. Prior to federation, the common law had created a number of procedures that looked and functioned rather like our present-day ‘appeals’ and cases stated, usually limited to questions or errors of law, but sometimes allowing direct challenge to findings of fact. Examples included cases stated (both on certiorari and more generally), attaint of verdict, writ of error, judgment notwithstanding the verdict and even the simple method of adjourning a case for the consideration of Quarter Sessions or the Assize Court. The modern-day stance that ‘appeals’ are the creature of statute was unknown in those times. Isaacs J said that by the 1860s: ‘the primeval juristic meaning, if I may so term it, of the word “appeal” had undergone transformation’. ¹⁰³ Indeed, in 1765, Blackstone’s Commentaries had a chapter titled ‘Of Proceedings in the nature of Appeals’; ¹⁰⁴ its subject was common law appeals. The great institutional and procedural reforms of the 19^th century included the introduction of general rights of appeal in civil matters. However, generalised statutory rights of appeal in criminal matters occurred only in the early years of the 20^th century, although they had been urged for many years before that. ¹⁰⁵

It seems, therefore, that the abolitionist cause must argue directly from policy. There can be no doubt that applications for non-jurisdictional certiorari have been launched, and even granted on occasions, where the grounds are fairly trivial legal errors. In other words, non-jurisdictional certiorari can be dysfunctional. However, it can also be useful. The Parliaments of Victoria and New South Wales certainly thought so, because they expanded upon the common law definition of the ‘record’ which forms the focus of non-jurisdictional certiorari. ¹⁰⁶ Two of the reasons given in Public Service Board (NSW) v Osmond for refusing to create a common law duty that administrators give reasons for their decisions were that the topic had already attracted the attention of Parliament, and that that was the better option because it allowed for fine tuning between different contexts. ¹⁰⁷

From Certiorari’s History to Privative Clauses

The differences between Gageler and Edelman JJ are more than just a scholarly disagreement between two eminent judges with a keen interest in history. Probuild’s critical issue was whether the court should interpret an Act that had no ‘finality’ clause, let alone a privative clause aimed explicitly at certiorari, as ousting non-jurisdictional certiorari. The answer for all three Probuild judgments was ‘yes’, and all three acknowledged that to be novel. An unbroken line of cases before Probuild had said that an Act needs to be explicit if it is to oust certiorari, even for non-jurisdictional legal errors. In essence, Probuild’s question boiled down to whether those cases evidenced an unbending ‘manner and form’ requirement or (for six of the judges) whether the lesser form of certiorari could be ousted by implication. For the seventh judge, Gageler J, no implication was needed.

Australia theorises judicial review of administrative action as an extended (and ‘increasingly sophisticated’) ¹⁰⁸ exercise in statutory interpretation. That poses some difficulties for judicial review of non-statutory administrative action, ¹⁰⁹ which this article will not pursue. Aside from non-statutory power, and aside from review on constitutional grounds, the grounds for judicial review theoretically derive from legislation, as do the restrictions on those grounds. The noticeable expansion of the generic grounds of judicial review has been achieved by the deployment of interpretive presumptions — for example, presumptions that powers are granted on condition that they not be exercised unreasonably, irrationally, or in breach of natural justice.

Perhaps less remarked is the fact that statutory interpretation does not stop at the grounds; the judicial approach to privative clauses is informed by constitutional considerations, but operates for the most part in disguise, as ‘statutory interpretation’. Whether they be State or federal, privative clauses cannot oust judicial review for jurisdictional error, but they can oust non-jurisdictional certiorari. The standard-form federal privative clause purported to oust remedies available only for jurisdictional error, and nothing could have been easier than to invalidate it for conflicting with s 75(v), which invests the High Court with an entrenched jurisdiction to grant mandamus, prohibition and injunctions against Commonwealth officers. Wherever possible, however, the High Court has steadfastly rejected the invalidity route, in favour of a ‘reading down’ approach, which was originally ‘tortuous’, but now is less so. ¹¹⁰

In two separate judgments, six of Probuild’s seven judges took a statutory interpretation route, finding that the Act impliedly ousted non-jurisdictional certiorari. Gageler J was the seventh judge, and he had no need to take a statutory interpretation route, because in his view, non-jurisdictional certiorari was unavailable in any event. He said that it issues only from a court that can try the matter on its merits, to an inferior court of record. I have discussed this above, and need not revisit it here.

The Act in question in Probuild was the Building and Construction Industry Security of Payment Act 1999 (NSW), which has counterparts around Australia. The Act is currently under review, to make it more effective. ¹¹¹ Nevertheless, the overall aim of the scheme is accepted. The purpose of the legislation is to maintain progress payments (and therefore cash-flow) to building contractors, while not affecting the final determination of the rights or obligations of the parties, who can litigate in the normal course at a later date. Adjudicators drawn from a panel make rough-and-ready adjudications of payment disputes between principals and their contractors. The principal must pay any progress payments found to be due, and the adjudications can be registered in the appropriate court and enforced as judgments, albeit judgments that do not preclude either party from re-opening the entire matter in subsequent litigation. Any party unhappy with the decision can take it to court later, where the payment can be undone. The idea is that postponing fully-fledged litigation will more likely keep the construction process on foot, and the interim payment scheme will tend to counter-act the dominant financial power of the principal. It all depends upon compliance with extremely tight deadlines for the claim and the response, the adjudication and its registration and the interim payment, with the Act spelling out each step in some detail. Adjudicators themselves are under tight deadlines and will necessarily be determining both factual issues and questions of law regarding not just the Act’s requirements but also the underlying contract. Errors of law and fact are inevitable; they are the price paid for keeping the construction project on foot. Adjudicators’ decisions are reviewable for jurisdictional error. However, the courts are reluctant to characterise errors as jurisdictional; hence the Probuild dispute as to the availability of non-jurisdictional certiorari.

Probuild’s joint judgment drew its implication that the Act ousted non-jurisdictional certiorari from a total of nine factors, all of them relating to the Act’s overall purpose, its emphasis on speed (with a necessary consequence that there would be more errors), its reversal of financial power as between principal and contractor, the maintenance of cashflow and the interim nature of an adjudication (and therefore its reversibility in normal litigation in due course).

Edelman J reached his implication of an ouster of non-jurisdictional certiorari more directly. His Honour said that privative clauses had occasioned a ‘power struggle’ between the courts and Parliament, a struggle that had now reached a settlement or ‘détente’, in which Parliament may be assumed to have accepted the courts’ ‘narrow’ (meaning strict constructionist) approach to privative clauses. ¹¹² A presumption against ouster of non-jurisdictional certiorari remains, although its strength is weaker than in the case of jurisdictional ousters, and even more weak where (as in Probuild itself) vindication in normal court proceedings of the aggrieved party’s rights is delayed but not taken away. ¹¹³

Access to the Courts: What Needs Entrenching?

Section 75(v) of the Constitution entrenches the High Court’s jurisdiction to issue mandamus, prohibition and injunction against Commonwealth officers. The decision in Kirk declared that the State Supreme Courts have a similarly entrenched jurisdiction, ¹¹⁴ because their supervisory jurisdiction was one of their ‘defining characteristics’ as at 1901. ¹¹⁵ However, a number of constitutional puzzles remain. These include the extent (if at all) to which some of judicial review’s grounds are entrenched, or whether Parliament can use ‘no invalidity’ clauses to pre-emptively validate all breaches of the legislative requirements that attach to the exercise of executive power. The answers to that set of issues remain a work in progress, ¹¹⁶ but there can be no doubting the competence of Parliament to remove at least some of the review grounds typically implied into and conditioning statutory grants of power.

Judicial review’s generic grounds have a long history, and their reach and content are generally expanding. They are implied, and it usually takes very clear words before an Act is taken to have repelled the implication. Nevertheless, there is a growing line of cases suggesting another way of overcoming the implication.

Canada has long read privative clauses as indicating that what we in Australia would call the decision-maker’s ‘jurisdiction’ is more extensive than it might otherwise have been. ¹¹⁷ The Immigration Minister had argued for a similar approach in S157, ¹¹⁸ but without success. The court acknowledged the possibility of privative clauses tipping the balance in deciding whether a particular statutory requirement is directory or mandatory. ¹¹⁹ Nonetheless, it rejected such a role for the privative clause in the Migration Act 1958 (Cth), because it was not to be supposed that the clause there in question was meant to override the mandatory nature of the requirement for procedural fairness. More importantly, it was not credible to suppose that a single privative clause was meant to downgrade from mandatory to directory the nature of countless other requirements required by that Act. ¹²⁰

Edelman J’s route in Probuild to an implied ouster of non-jurisdictional certiorari has already been noted. Importantly, his Honour said that access to the courts was the critical issue, and Probuild’s Act had done no more than delay such access:

The rationale for the narrow approach to construction is protective of the reason for judicial review, namely, access to the courts to correct legal errors relating to a person’s rights. Where, as here, that access is generally preserved without much practical effect on rights then the rationale is not sufficiently engaged to overcome the inference that arises from ordinary principles of construction. ¹²¹

Although his Honour made no mention of it, the obvious analogy is with the High Court’s decision in Futuris. ¹²² That case upheld the validity of federal tax legislation that sought to channel all legal and factual disputes about tax assessments to the AAT and the Federal Court. Futuris said that constitutionally protected judicial review of tax assessments remained available, but only for decisions taken in bad faith. In effect, the provision of alternative ways of accessing the courts was a valid trade-off for almost all of judicial review’s grounds of review. ¹²³ I have used italics to emphasise that the Futuris trade-off was with judicial review’s grounds, not with judicial review’s existence. There is no reason to suppose that Futuris would have accepted a legislative scheme that purported to grant appeal rights and preserve judicial review, but limited in both cases to allegations of bad faith.

Futuris was not the first case to contemplate a swap — reduced judicial review in return for reasonably adequate statutory remedies, ¹²⁴ but nothing had come of those earlier cases, which had involved explicit privative clauses naming the remedies to be ousted. In view of the language of s 75(v), a purely functionalist assessment in terms of access to the courts might be a step too far in dealing with such privative clauses. ¹²⁵ However, it was possible in Futuris, whose Act combined ‘no invalidity’ and ‘conclusive evidence’ provisions that avoided a direct collision with the words of s 75(v).

Basten JA pushed the functionalism of Futuris a step further, in Kaldas. ¹²⁶ Mr Kaldas had sought declaratory and injunctive relief against the Acting Ombudsman, whose special report to Parliament had included some serious (and career-limiting) findings against him. His grounds included breaches of both rules of natural justice, and an argument that there was no power to report directly to Parliament. He had had ample time to litigate earlier. Indeed, he had complained and threatened litigation during a substantial part of the Ombudsman’s investigation. The Ombudsman Act 1974 (NSW) explicitly sought to immunise the Ombudsman from liability on any ground, including ‘want of jurisdiction’. While not a standard-form privative clause ousting named remedies, it was close enough. However, in functional terms, it need not have been as bad as it looked, because the Act created its own mechanism for judicial review. It allowed challenges to the Ombudsman’s ‘jurisdiction’, but Kaldas read this narrowly. The majority view was that it allowed challenges only for such jurisdictional issues as arise before commencement of the Ombudsman’s hearings. ¹²⁷ The alternative view was that it allowed for challenges only if they were made before the inquiry’s completion. ¹²⁸ It is difficult to see either approach as a functional equivalent to common law judicial review, allowing full access to the courts. A further complication was that the statutory mechanism for challenge needed court leave and a ‘substantial’ ground for alleging bad faith. Mr Kaldas did not allege bad faith.

Taking his cue from Futuris, Basten JA (with Macfarlan JA in agreement) read this rather complicated set of provisions as expanding the Ombudsman’s ‘jurisdiction’ to anything other than conduct taken in bad faith. ¹²⁹ Bathurst CJ disagreed on that point. ¹³⁰ With different majorities, in other words, Kaldas held that the statutory mechanism for challenge could only be for jurisdictional issues prior to the commencement of the Ombudsman’s inquiries, and that the only jurisdictional issue that could arise would be an allegation of bad faith. It is difficult to see how bad faith could arise at such an early point in the process. And if Bathurst CJ was right to say that the only restraint upon the Ombudsman’s jurisdiction was that he or she should act in good faith, it is difficult to understand how his Honour was able to avoid confronting the question of whether the ouster of all jurisdictional challenges could be read in such a way as to sever the bad parts from the good.

Futuris had indeed reduced the scope of judicial review to allegations of bad faith, but its trade-off was a statutory challenge scheme on all grounds — factual, legal, jurisdictional and non-jurisdictional. The Kaldas trade-off came nowhere near that, but Basten JA justified it by reference to history and a narrow view of the effect of the Ombudsman’s report. His bottom line was that although Ombudsmen’s reports can damage reputations, they carry no legal force — they seek only to urge the relevant public body to do things, such as apologise or pay compensation. ¹³¹ That had two consequences. First, the Act’s insubstantial trade-off was acceptable, because the Ombudsman’s report did Mr Kaldas no great (legal) harm. ¹³² Secondly, and Bathurst CJ’s reasons were substantially the same here, the Supreme Court’s entrenched judicial review is to be limited to the law as it obtained at 1901. The ‘bare’ declaration (at least in public law) was probably unknown back then, ¹³³ and Kirk did not mention it. ¹³⁴ Further, the prerogative writs in 1901 were based on a rights-protection model (although Kaldas did not use that terminology) — they issued only to litigants seeking either to protect their existing legal interests or to avoid the unlawful imposition of new legal obligations. ¹³⁵

Bathurst CJ said that certiorari was and remains (because of Ainsworth v Criminal Justice Commissison) ¹³⁶ limited to a rights-protection model, ¹³⁷ but that prohibition might not be so constrained. ¹³⁸ Both judgments accepted that bare declaratory relief is generally available in administrative law, and that it has long broken the constraints of the rights-protection model, but both said that the Ombudsman’s Act cut judicial review right back to an entrenched bare minimum, and it was at this point that legal history became so important.

Taking Kaldas Beyond Ombos

It would be a mistake to make too much of Probuild. Aside from its history wars, its innovation was to imply an ouster of non-jurisdictional certiorari, and in functional terms, that was eminently desirable. Kaldas is more important in two respects: its barely concealed antipathy to Kirk and its view that an extremely parsimonious statutory challenge mechanism served well enough as an ‘acceptable’ ¹³⁹ trade-off for traditional judicial review. In both cases, however, a strong regard for functionality is clearly evident. Challenges for any and every legal error by construction adjudicators would have crippled the Probuild scheme, and there is rarely a net social gain in tying up the Ombudsman in litigation.

There was, however, a much cheaper and faster way of dealing with Mr Kaldas. His grounds were extremely weak, and he was far too late. In other words, Kaldas was not about Mr Kaldas. It was about immunising the Ombudsman from the common law of judicial review. ¹⁴⁰ The collateral damage was either to undercut Kirk’s rationale of retaining relief for jurisdictional error (as frankly acknowledged by Bathurst CJ) ¹⁴¹ or to say (with Basten JA) that Kirk’s ‘principle’ was nothing more than to entrench a snapshot of doctrine frozen in time. ¹⁴²

Ombos are neither fish nor fowl. They have sweeping investigative powers. They can interrogate witnesses in camera and on oath or affirmation. They can require witnesses to keep their questioning secret, and Ombudsmen themselves are under strict secrecy obligations. They usually proceed informally, but if they hold hearings, these must be in private, although witnesses can bring their lawyers. The Kaldas Ombudsman could override the privilege against self-incrimination, although evidence thus obtained could not be used against the witness except in a prosecution for lying, and the Ombudsman thought that Mr Kaldas may have lied. Alongside these large powers, the Ombudsmen have no power to change things. Their Reports are immune from defamation law. While they can urge, shame, denounce or praise, they cannot make binding findings of fact or law, and they cannot make binding orders. They are a cross between an internal audit office and a complaints office, and their efficacy depends entirely on their power to persuade. Kaldas was explicitly influenced by the ‘weakness’ (in strictly legal terms) of Ombudsmen’s Reports.

An interesting comparison might be made with the legal status of Reports of the Independent Commission Against Corruption (ICAC), which for ordinary members of the public are as ‘weak’ in strictly legal terms. ICAC has fearsome investigatory powers, but its Reports impose no legal liability; nor do they amount to a conviction. They can, however, make findings of corrupt conduct, against public officials and against members of the public who were party to that conduct. An ICAC finding of corrupt conduct by a public officer can in itself suffice as the basis for disciplinary action on the part of the employing agency, although the officer is, of course, entitled to question the finding. ¹⁴³ As against the public officer, therefore, an adverse ICAC Report could be tested by certiorari, but members of the public ‘found’ to have been involved in that corruption will be unable to use that remedy. More importantly, however, Kaldas might present problems for those wanting to challenge an ICAC Report if that body’s immunities were to be amended to bring them into alignment with those enjoyed by the Kaldas Ombudsman.

ICAC’s staff are protected from any ‘action, liability, claim or demand’ for anything done in good faith, although ICAC itself does not have the same protection. ¹⁴⁴ In addition, no ‘criminal or civil liability’ attaches to any ‘person’ for good faith compliance or purported compliance ‘with any requirement’ of the Act (emphasis added). Those provisions would doubtless be read down so as not to affect judicial review. That approach was not credibly open in Kaldas, but if the ICAC protection was broadened simply to cover judicial review, would Kaldas then indicate that ICAC could commit jurisdictional error only when it acts in bad faith? Or would the amendment also need to add a statutory mechanism allowing some form of court challenge? If a statutory challenge mechanism were required, Kaldas poses further questions as to whether it could be limited to allegations of bad faith or, perhaps, limited to grounds alleging a want of jurisdiction on ICAC’s part to commence an inquiry or investigation.

Conclusion

The concerns of this article have to do with some of the reasons given in Probuild and Kaldas, rather than with their outcomes.

Non-jurisdictional certiorari can certainly be dysfunctional at times, but not always, and in any event, that sort of argument is more appropriately directed to legislatures, not the courts. The case for abolishing non-jurisdictional certiorari needs to focus on showing where it is dysfunctional, rather than searching for a grand answer, or relying on unconvincing history or semantical puzzles as to how legal errors might be ‘authorised’.

The policy reasons for limiting judicial review of Ombudsmen to substantial claims of bad faith have yet to appear, and it is submitted that the legal arguments have enormous potential for confusion. There is no functional reason for limiting the constitutional entrenchment of judicial review to the remedies and principles that applied at federation. The prerogative writs in 1901 were tied to the protection of existing legal rights, but they have developed since then to adjust to the emergence of the administrative state — that should not be undone in the name of history. It would encourage legislative drafters to judge-proof those parts of the administrative state that do not fit the 19^th century’s ‘rights protection model’. It could also lead to the emergence of parallel judicial review laws — one for the present, the other frozen in time.

The underlying reason for the constitutional entrenchment of judicial review of administrative action may well be to maintain access to the courts to correct unlawful government action. If so, then it might be possible to reduce judicial review’s common law grounds while at the same time granting full appeal rights under the relevant statute. However, it would be extraordinary to limit both judicial review and the statutory appeal rights to one ground only, namely, bad faith. We should be ‘entitled to expect more than good faith’. ¹⁴⁵