Melbourne University Law Review
|
|
Home
| Databases
| WorldLII
| Search
| Feedback
Melbourne University Law Review |
|
SIMON YOUNG[*]
This article explores how Australian courts have grappled with the challenges of changing context and a frenetic case load in their application of the rule against bias. In their efforts to keep this sacrosanct rule relevant and coherent they have employed three key tools of ‘calibration’: the ‘fair-minded lay observer’, the spectrum of standards, and a (re)emerging technique of sub-categorisation or ‘speciation’. The lay observer has wearied, becoming awkwardly indistinct in important contexts; the spectrum approach has enjoyed an expanding importance but now appears to have reached its high-water mark; however, the ‘speciation’ approach has shown its precision and is perhaps the key to the next generation of cases. This article re-maps the bias rule in Australia by reference to these three tools of calibration, thereby placing the accumulating critique of the
‘lay observer’ test into clearer context. It also offers some predictions on the law’s
future trajectory.
CONTENTS
The rule against bias has a uniquely long pedigree in Western legal thinking. It was broadly declared in the Magna Carta, the 800th anniversary of which recently passed, that ‘justices, constables, sheriffs [and] other officials’ would be drawn only from ‘men that know the law of the realm and are minded to keep it well’.[1] More specifically, in provisions dealing with the resolution of fine disputes by ‘twenty-five barons’ and the archbishop, it was declared that ‘if any of the twenty-five barons has been involved in a similar suit himself, his judgment shall be set aside, and someone else chosen and sworn in his place’.[2]
This revered principle, in its modern form, continues to be of fundamental importance in its broad ongoing application to the executive and judiciary in Australia. It is important to the quality of decisions being made, to the confidence and cooperativeness of individuals being affected, and ultimately to the coherency and security of the social order.[3] Yet despite this clarity of purpose and long lineage, this administrative law principle has been profoundly tested by contemporary context.
The modern challenges are many. Government in all forms (including the judiciary) continues to expand and diversify, to grapple with resource constraints, and to engage and interconnect more deeply with non-government players.[4] The political dimensions of decisions are more acute and more visible, potential conflicts of interest are more varied and less visible, and decision-making processes are more complex. Moreover, consumers of government services and litigants are more sophisticated in their expectations and savvier in their strategies — and indeed recusal processes are often publicly scrutinised in real time.[5]
There is no shortage of case law to test the old sacrosanct rule against bias. Although the pace has been more ferocious for the parallel ‘fair hearing’ limb of natural justice, both are busy fields of dispute. Importantly, both have felt the pull and pressure of migration litigation in recent years — which by virtue of its volume, focus and attendant controversy has been a remarkable force in modern Australian administrative law.[6]
Faced with this degree of contest, changing context, and the natural inertia of an old legal principle, Australian courts now appear to deftly employ three tools for finer calibration and adjustment in the application of the bias rule. One was developed many years ago as a malleable legal measure of acceptable risk and perception — the notional ‘fair-minded lay observer’. A second appeared to crystallise clearly in the course of the politicisation of migration decisions in the early 2000s — the ‘spectrum of standards’ approach, which tailors applicable standards to context. More recently, however, a third tool of calibration has started to emerge with some force — what might be termed the ‘speciation’ approach by which sub-categories of bias are identified and distinguished by the application of altered rules.
This article explores the current state of the rule against bias in Australia, with particular attention to the nature and interrelationship of these tools of calibration. Analysis reveals that the harried notional ‘lay observer’ has become awkwardly indistinct in important contexts. It also indicates that the ‘spectrum’ methodology has enjoyed an expanding importance but is now receding from its high-water mark. And it reveals that an emerging ‘speciation’ approach has shown its efficiency and precision and is perhaps the key to the next generation of bias cases. Under this latter approach, a deepening significance is attached to what were previously somewhat abstract sub-classifications of bias. It is too early to say whether this trend marks a paradigm shift, but there are signs that this categorisation approach is becoming the courts’ preferred tool for meeting difficult new arguments and adjusting old principle.
While findings of bias can resonate deeply through government and/or the legal profession, focused legal commentary on this topic is relatively rare.[7] More is written about the ubiquitous ‘fair hearing’ rule. Yet as noted above the case law on bias is also vibrant, and indeed from the contemporary foundations laid in Ebner v Official Trustee in Bankruptcy the law in this field has been edging towards ever-greater complexity.[8] Further, there has certainly been an accumulation of dissatisfaction with the unkempt ‘lay observer’
test — in Australia[9] and elsewhere.[10] All of this is worthy of close attention. This article is an attempt to re-map and reinterpret the bias rule in Australia by reference to the tools of ‘calibration’ that it identifies, in the process considering the problems of the ‘lay observer’ test in the context of the newer judicial strategies for dealing with changing context and difficult cases. It also offers some predictions on the future trajectory of the law in this field.
In the decision of Ebner the High Court rejected bias claims against judges who had shareholdings in public companies interested in litigation before them.[11] Importantly, the Court re-coined a generic ‘reasonable apprehension of bias’ test — in the process finally discarding the old automatic disqualification rule for ‘pecuniary’ bias. Under this test (building particularly upon the earlier decision in Webb v The Queen)[12] subject to principles of waiver and necessity, ‘a judge is disqualified if a fair-minded lay observer might reasonably apprehend that the judge might not bring an impartial mind to the resolution of the question’.[13] The Court explained that such cases involve a two-step inquiry: first identifying exactly what it is that might cause impartiality, and secondly seeking a logical connection between that and the ‘feared deviation’ from the course of deciding the matter on its merits.[14] Despite the original language, this test is readily applied beyond the judicial decision-making context.[15]
The Ebner refinements still left us with a pliable, broadly cast test. It is conventionally understood to be concerned with outward appearance and public confidence[16] — falling somewhere between the old automatic disqualification rule and a search for proof of actual bias (which remains a distinct claim). But how might the test be applied coherently across very different types of decision-making? How much knowledge and understanding — necessarily a subset of full awareness[17] — is to be attributed to the prefabricated lay observer? How much context is notionally included?[18] Is knowledge of circumstances not actually known to the decision-maker at the relevant time attributed to the lay observer?[19] Certainly it is clear that the test is in broad terms an objective one. We are regularly reminded that disqualification for bias is not triggered just by a party’s personal apprehension of partiality,[20] and certainly not just by their fear of an adverse result.[21] Nor, for that matter, is it reversed by a declaration of impartiality by the impugned decision-maker.[22] It has also been emphasised that the Ebner test is concerned with possibility,[23] rather than actuality or probability, and the double ‘might’ in the wording of the test (quoted above) has featured prominently in recent decisions[24] — although this has been blunted in some recent cases by a more exacting application of the Ebner second step (ie the search for a logical connection between the matter identified as a concern and the feared partiality).[25]
Critically, as alluded to above, there is ample room for contention about the knowledge and perception of the reasonable lay observer. As Kirby J has noted, often when a legal fiction is adopted ‘the law endeavours to avoid precision’.[26] In the recent High Court decision in Isbester v Knox City Council (concerning a council process leading to a dog-destruction order), it was reaffirmed in broad terms that ‘[t]he hypothetical fair-minded observer ... is to be taken to be aware of the nature of the decision and the context in which it was made as well as to have knowledge of the circumstances leading to the decision’.[27] Yet a more precise tuning will often be required in particular cases.[28] That precision can be elusive, and become somewhat subjective, and hence conclusions can be unpredictable. Small gradations in the understanding attributed to the lay observer can quickly determine the outcome of
a challenge.
Some cases raising these issues might require a measure of self-reflection by the courts. As Kirby J has noted, in contemporary Australia the fictitious bystander ‘is not necessarily a man nor necessarily of European ethnicity or other majority traits’.[29] Even with such social neutrality accommodated, perception is a tricky thing to predict. We have, however, been reminded that the lay observer is assumed not to make ‘snap judgments’,[30] is not ‘unduly sensitive or suspicious’,[31] and would recognise that some judges are naturally ‘more forthright’ than others.[32] It has also been said that the observer will know that a judge is capable of departing from earlier expressed opinions,[33] and can put things out of their mind (eg prejudicial matter not open
to them).[34]
Some notable pressure points have emerged in the contemporary application of this test. A selection of these are examined in this Part. They perhaps indicate that the volume, variety and complexity of modern challenges have fatigued the reasonable lay observer[35] to a point where alternative methods of calibration are now more actively sought — and are possibly more important in the future operation of the bias rule. In fact we may be nearing (or perhaps returning to)[36] the point of admission that in many circumstances the ‘lay observer’ test, despite the deliberate terminology, is in truth the law’s own sophisticated assessment[37] of what the system can bear.[38] Public perception does and should continue to play a role in that assessment — but perhaps not the dominant role that conventional statements of principle might suggest.
In a sense, a new and quite complex ethical standard has emerged, still
clothed — but barely — in a ‘lay observer’ test.
By way of first example, the reasonable lay-observer has been strained in the context of their scrutiny of adjudicative hearings. While judges are held to high standards,[39] the cases illustrate well that in this setting the principles are applied with a heavy pinch of pragmatism and quite sophisticated attention to relevant context. Bias claims resting upon exchanges or behaviour at a court hearing are rarely successful. And tribunal members are found to fall short of the standard applied to them only slightly more often than judges.[40]
It has long been accepted that a judge is not expected to sit in ‘Sphinx-like’ silence while arguments are presented and that a judge will often form and disclose tentative views (if only to draw a response or assistance from counsel).[41] Yet the contemporary challenges do make for colourful reading. Challenges have been unsuccessful in respect of continual irritation on the part of a judge;[42] fractious exchanges with a judge;[43] highly specific questioning by a judge[44] or a tribunal member;[45] probing questions and conspicuous reactions by a judge;[46] significant intervention by a judge in cross-examination[47] — even erroneous or undesirable intervention;[48] a judge’s declaration that claims were ‘fanciful and ridiculous’;[49] and indeed a judge’s description of a party’s evidence as ‘just preposterous’[50] (combined with fierce debate and interruptions).[51] This latter case illustrates the level of attention paid to setting.[52] The reviewing Court acknowledged the potentially cumulative nature of bias, but still excused the lower judge by reference to context,[53] counsel’s level of experience, and the fact that the strong views were expressed after evidence and argument.[54] Perhaps a high point of tolerance came in the recent Victorian decision of Cook v The Queen, where a bias challenge was unsuccessful in the context of what were said to be ‘needlessly argumentative and rude’ dealings by the judge with the defence counsel (that at times descended into ‘verbal abuse’) and comments to the effect that the jury would inevitably view the accused as ‘pathetic drug addicts’.[55]
Challenges of this nature are frequent, despite the lack of success. In one sense it is entirely appropriate that the reasonable observer is forgiving in this context. As legal observers well know, one would be hard pressed to find in Australia a judge who, even at their most irate, would decide a case other than on its merits. And legal observers recognise the significant provocation that often precedes judicial impatience.[56] But of course on a conventional understanding it is not legal observers that the Ebner rule is designed to reassure, and we have been told that the hypothetical observer should not be presumed to reject the possibility of prejudgment by a judge.[57]
The review courts must provide pragmatic answers to these challenges for the sake of the lower judges’ ability to manage proceedings and interact with parties — even more so in an age of high caseload pressures, unrepresented litigants[58] and bolder litigation strategy. However, this is one context that gives us reason to pause and consider whether this is all adequately explained by reference to the comfort of the reasonable lay observer, even one notionally loaded up with an understanding of the training and oath of judges and their likely intervention with opinions.[59] The lay observer must at least be cringing at some of the reported exchanges — and perhaps a truer contemporary explanation is that conduct at a hearing is measured by a more sophisticated legal assessment of what is acceptable.
A more technical and more significant challenge to the coherency of the reasonable observer test is found in the context of ‘multiple involvements’ in a matter by a decision-maker.[60] The cases on this appear to have been increasing in frequency, perhaps spurred on by two relatively recent High Court decisions (discussed below).
The issue of whether a matter being remitted should be returned to a differently constituted body has long been a live one (in both the court[61] and tribunal[62] contexts). And the 1983 decision in Livesey v New South Wales Bar Association[63] (concerning findings on credibility in connected matters) has of course long provided an example of how difficulties can arise less directly in
the context of overlapping or related proceedings.[64] However, in this judicial context (and indeed the tribunal context) the issues are now made much
more complex by contemporary case management techniques and the frequency of interlocutory orders,[65] and indeed by the procedural and evidential peculiarities that might attend particular decision-making frameworks.[66] The reasonable lay observer standard becomes a difficult one to apply in such circumstances.
The recent High Court decisions on point ultimately illustrate that the exact nature of the earlier involvement, and of any findings made, will be critical.[67] In British American Tobacco Australia Services Limited v Laurie, the central issue was that a judge of the Dust Diseases Tribunal of New South Wales had made an interlocutory finding (in a separate case involving different parties) that the appellant (‘BATAS’) had a fraudulent policy of destroying potentially adverse evidence.[68] A similar allegation was made against BATAS in this later case, before the same judge, and this prompted a claim of reasonable apprehension of bias (in the form of prejudgment). In the High Court, the parties’ arguments focused on the judge’s express qualification (in the earlier matter) that the finding was only interlocutory and that the question remained ‘a live issue for the trial’ as contrary evidence had not really been led at that point.[69] Nevertheless, a majority of the High Court upheld the challenge, deciding that the express qualification was not enough to remove the difficulty given the clarity, strength and seriousness of the earlier finding.[70]
A different result was reached in the High Court decision of Michael Wilson & Partners Ltd v Nicholls (‘Wilson’),[71] which emerged from litigation over conduct following a professional split. One party had made several ex parte applications in the matter. The judge’s later participation in the trial was challenged on the basis of reasonable apprehension of bias. The Court concluded that the interlocutory orders obtained (with accompanying non-disclosure orders) did not found a reasonable apprehension of prejudgment of issues to be decided at trial — even if the making or duration of the non-disclosure orders was in error.[72] Their Honours emphasised that the judge had made no determination of any issue to be decided at trial, no finding about the reliability of any party or witness, nor any choice between competing versions of events.[73]
Clearly, as has long been acknowledged, the exact nature of the earlier involvement and of any earlier findings will be a key issue.[74] The difficulty might therefore seem to be navigable, however the distinctions drawn in many of the contemporary cases and the complexity of the processes involved might be confounding for the lay observer.[75] Moreover, there is a potentially significant practical problem here for both courts and many tribunal-type bodies. While in some instances initial case management and mediation might be kept away from the final substantive decision-makers, this may
not always be desirable, possible or affordable. Indeed the efficiency and substantive benefits of some continuity in the management of a matter are now well recognised.[76]
The courts have limited scope to respond to these modern practicalities with the lay observer test. In some instances, it might be possible to acknowledge the formal segmentation of a matter. The existing rule may still readily accommodate the incremental resolution of one contest or the early disposition of some aspects of a case.[77] Beyond this, courts might attempt to further load the fictional lay observer with procedural understanding so as to allow more prior involvement. Past decisions have noted that it is appropriate to attribute some understanding of ordinary judicial practice, which is not frozen in time.[78] A less obvious possibility is a modernisation of the under-explored exception to the bias rule that we loosely term ‘necessity’ — under which persons can act in situations where they are the only ones able or authorised to do so.[79] Notions of practicability, reasonableness and overall public confidence have been called upon in previous necessity cases.[80] Ultimately, however, all of these mooted responses perhaps lead us to the realisation that this is another context in which we might question whether the revered lay observer can still coherently lead the way. There appears to be a more complex and more exacting contextual balancing to be undertaken in determining the appropriateness of multiple involvements.
Another matter that has been contentious in recent years, and that perhaps awaits further close exploration, is the extent to which a challenged judge’s own reasons on the recusal application, or ultimate reasons in the substantive matter, are part of the context to be assessed by the reasonable lay observer and hence relevant to the challenge. The question can also, of course, arise beyond the judicial context. Once again, the cases reveal some strain in the application of the lay observer test.
In British American Tobacco, the High Court majority ventured into the dispute between the parties as to whether the judge’s own reasons for refusing to disqualify himself in the second matter were ‘assumed knowledge’ of the lay observer. Heydon, Kiefel and Bell JJ took the view, contrary to Basten JA’s view below,[81] that these ‘add[ed] nothing of moment to the material on which the hypothetical observer’s assessment is to be made’.[82] Their Honours acknowledged that later comments of a judge might help to explain earlier questionable comments, but emphasised that a later assertion of impartiality per se was not important: it is assumed that a judge who feels compromised in the requisite way will not sit, so the judge who does decide to sit when challenged for apprehended prejudgment will necessarily have been confident of their impartiality.[83] Here, it was said that in the event that the earlier finding gave rise to a reasonable apprehension of prejudgment the judge’s subsequent characterisation of the nature of the finding could not serve to allay that apprehension.[84] The basic proposition that a judge’s own later assertion of impartiality cannot be allowed to decide the matter on a review is a logical one — they of course are not the ‘fair-minded lay observer’ upon whose perception the test turns. However, there would appear to be a thin division between assertions of impartiality and comments that may help to explain the earlier impugned conduct.[85]
In the Wilson decision, Gummow ACJ, Hayne, Crennan and Bell JJ were critical of the attempt to rely on aspects of the final trial judgment in establishing a reasonable apprehension of bias, and (once again) critical of
Basten JA’s preparedness to pursue such a line in the decision below.[86] Their Honours considered that a search of the final reasons for some confirmation of a reasonable apprehension, or for some ‘crystallisation’ of the apprehended bias into actuality, heads for the ‘fallacious argument’ that because one side lost (or because there was an appealable error) the judge was biased — and indeed impermissibly blurs actual and apprehended bias.[87]
The conspicuous differences in the approaches of the High Court and the New South Wales Court of Appeal suggests that there is more to be explored in respect of these issues. The High Court’s thinking, particularly from Wilson, has been duly reflected in some later Federal Court decisions,[88] and has since been taken to the tribunal and indeed the ministerial decision-making contexts.[89] Yet it might be premature to attribute too small a role in this context to a decision-maker’s final reasons and own responses on a bias challenge — or indeed to any incremental correction (implicit or explicit) of earlier statements.[90] After all, the courts have frequently emphasised in earlier reasonable apprehension cases the need for attention to broad context, the reality that bias can be a cumulative concept, and the fact that comments in a final decision and indeed other post-decisional conduct may be relevant.[91] Basten JA, whose more searching and flexible approach in the New South Wales Court of Appeal was rejected by the High Court in both British American Tobacco and Wilson, has himself on the critical issue retreated only to the position that it will be ‘rare’ for apprehended bias to be made out by reference to the final judgment.[92]
Ultimately, on the issue of recusal reasons, there would appear to be a fine line between mere late assertions of impartiality (perhaps unhelpful if unsubstantiated) and elucidation of earlier impugned comments or actions.[93] And as regards the relevance of final reasons, whilst strong language may indicate nothing more than strong legitimate conclusions[94] and final errors should not too readily be attributed to partiality, again there may be a fine line between these and incremental evidence supporting a reasonable lay observer’s suspicion of one-sidedness. Moreover, given the ever-increasing accommodation of context in the application of the lay observer test, it might seem somewhat artificial to remove the decision-maker’s own response to the bias challenge, and their final reasons on the substantive matter, from the purview of the notional observer. A conscientious lay observer might see the former to be very relevant to their assessment, and possibly also the latter. The sophisticated debate underlying these decisions, turning as it does upon some fine legal distinctions, seems once again to have drifted quite a distance from an inquiry into the likely instincts of the hypothetical observer.
Another difficult question that has re-emerged recently, testing the mettle of the notional lay observer, is whether the manifest bias of just one member of a decision-making panel is terminal to the whole process.[95] In years past, the review courts have on several occasions found the presence of one affected member too great a risk to be acceptable to the reasonable observer.[96] In Stollery v Greyhound Racing Control Board, the High Court found apprehended bias where the respondent board had deliberated on a complaint in the presence of the original complainant.[97] The line of authority considered in Stollery was approved of in the prominent Queensland decision of Carruthers v Connolly.[98] In that case, one of two commissioners jointly conducting an inquiry was found to be affected by apprehended bias — particularly by reason of their prior provision of advice, public statements and conduct in hearings. Importantly for present purposes, it was held that the second commissioner also could not continue — one reason being that his association and joint work over a substantial period with the disqualified commissioner would give rise to a reasonable apprehension of there being an unacceptable risk of partiality.[99]
These ostensible precedents for a strict approach, closely considered, in fact seem to foreshadow that the issue of bias in multi-member contexts must be handled with close and able attention to the specific context. The important High Court decision in Hot Holdings Pty Ltd v Creasy suggests, for example, that a contemporary court may be slow to act upon the peripheral involvement of a person with a potential financial interest.[100] And importantly, this broad issue split the New South Wales Court of Appeal in the key 2008 decision of McGovern v Ku-ring-gai Council.[101] In this case the issue arose in the context of possible prejudgment in local council decision-making.
Basten JA considered that even though the votes of the two impugned councillors were not decisive, if they were disqualifiable their participation during deliberations and voting may have tainted the proceedings and vitiated the decision.[102] However, Spigelman CJ rejected the general applicability of this ‘rotten apple in the barrel’ type test and applied a ‘but for’ analysis to the scenario — noting that even if the two councillors had not voted, the result would have been the same.[103]
In the 2015 decision in Isbester v Knox City Council, the High Court appeared to approve of the caution applied in early cases such as Stollery, but seemed also to suggest that it was appropriate to ask if the impugned decision-maker played a ‘material part’ in the decision.[104] This perhaps falls somewhere between Basten JA’s ‘bad apple’ approach and Spigelman CJ’s ‘but for’ approach. Yet the reference to this issue was brief in Isbester. Ultimately, it might be inevitable and appropriate, given the range and complexity of decision-making contexts now facing the review courts, that this issue will turn upon close consideration of the precise situation — the issues, the background, the process, the relationships — and a sense as to the likely (or reasonably apprehended) influence of the impugned panel member upon their colleagues.[105] In contemporary contexts (eg high-level or complex panel structures) a considerable amount of instinct and experience might go into any such assessment, and once again it is a difficult assessment to notionally devolve to the reasonable lay observer.
The critique offered here of the ‘reasonable observer’ test, via these examples of evident strain upon its contemporary operation, is in some respects unproductive. Even if we acknowledge the increasingly animated elephant in the room — namely, the legally expert tempering of the hypothetical lay observer’s instincts — what alternative test could be proposed? It may well be that a truly generic test is no longer possible. The purpose of this article is not to offer a new test, but rather a new explanation of what is actually happening. New tools of calibration have emerged and are emerging in the application of the bias test — born, it seems, of the combination of challenging new contexts and the fatigue of the ‘reasonable observer’.
The two modern methodologies which are now to be examined — termed here the ‘speciation’ and ‘spectrum’ approaches — both emerged clearly in the contemporary Australian context in the frenetic case law of the early 2000s, with thin conceptual roots in the earlier jurisprudence. They developed at different paces, but with interlocking stories. A suitable place to start is the notion of speciation, which in a sense has deeper roots and is closer to the heart of the topic.
The modern ‘speciation’ of bias began in reverse. As noted above in passing, the important High Court decision in Ebner finally put to rest the traditional sub-classification of bias into pecuniary and non-pecuniary bias.[106] It will be recalled that for the former a strict test had previously applied: once the decision-maker was shown to have a direct pecuniary interest, a conclusive presumption of bias arose resulting in automatic disqualification.[107] This approach had been increasingly beleaguered by definitional problems and inconvenience, for example in contexts where judges had insignificant interests in large corporate litigants,[108] and the principle was eroded through the Australian case law of the 1990s.[109] The Ebner decision effectively merged the tests for pecuniary and non-pecuniary bias in Australia, leaving us with the ostensibly generic ‘reasonable apprehension of bias’ test. While Ebner itself effected some convergence of relevant legal principle, its long-term impact appears to have been very different. The discarding of the old pecuniary/non-pecuniary distinction left space for renewed attention to more contemporary classifications of bias.
There remained, of course, a well-established distinction between apprehended and actual bias. This has been periodically reaffirmed in the contemporary cases, and with growing attention to the very different nature of the two inquiries.[110] Whilst claims of actual bias were encouraged for a time by the mechanics of the migration laws,[111] successful claims are now extremely rare — particularly in more formal decision-making contexts. Even at lower levels, administrative law’s ‘smoking gun’ is not often found in the workings of contemporary, well-organised government bodies.
Beyond the apprehended/actual distinction, more precise taxonomies have been gaining clarity and acceptance. It is suggested, for example, that in broad terms bias may emerge from interests, conduct, association or knowledge of extraneous information.[112] The precise characterisation of alleged bias will not always be easy,[113] particularly given the growing prevalence of self-represented claimants[114] and the fact that bias can be cumulative.[115] Moreover, it has been emphasised that any classification such as that noted above can only be a basic frame of reference, rather than a strict and comprehensive taxonomy.[116] Yet the categorisation of bias is becoming more prominent, and indeed in the heavy traffic of cases new sub-classifications appear to be emerging.[117]
The growing attention to the categorisation of bias is in part just a product of the patterns made visible by a steady accumulation of cases. However, it is argued in this article that there is something deeper at play in this trend. The courts are seeking new tractability in an old legal principle — in the face of a fast changing landscape and a yardstick ‘reasonable lay observer’ that is losing coherency. It may well be that this speciation of bias is now the settled trajectory of the Australian jurisprudence, and importantly, as will be seen, it has very real significance. The next challenge then lies in finding clarity in the division between categories, amidst the vast range of scenarios presented to the courts, so that the accumulated difficulties and perhaps subjectivities of previous approaches are not simply replicated in a new form.
Before we pursue the most important contemporary examples of the ‘speciation’ of bias, and their significance, it is necessary to deviate to the intersecting story of the ‘spectrum’ approach. In recent years the Australian case law has steadily and conspicuously built on the idea that decision-making context is critical in the application of the bias rules. The idea that there is a ‘spectrum’ of standards to be applied has been keenly deployed in various contexts on many occasions. This was initially the courts’ most visible reaction, amidst the rush of modern case law, to the need for a greater responsiveness and coherency in the bias principles. It effectively tempered the sensitivity of the reasonable lay observer by reference to the courts’ own hierarchy of decision-making contexts.
The spectrum approach appears to have firmly crystallised in the context of the politicisation of migration decisions in the late 1990s to early 2000s. In the 2001 decision of Minister for Immigration and Multicultural Affairs v Jia, Gleeson CJ and Gummow J particularly emphasised that in considering a charge of actual prejudgment (arising here from a Minister’s comments on radio and in a letter to the Administrative Appeals Tribunal) the nature of the decision-making process and the identity of the decision-maker may be critical,[118] and that the Minister’s conduct in this case had to be viewed in light of his being an elected official — accountable to the public and Parliament and entitled to be forthright and open about his portfolio.[119] And later, in a surprisingly brief consideration and rejection of the alternative ‘reasonable apprehension’ claim, their Honours again emphasised the error of applying to the Minister the standards of detachment applicable to judicial officers or jurors.[120]
Soon after Jia, this ‘spectrum’ approach was also being carefully applied to tribunal members. In Re Minister for Immigration and Multicultural Affairs; Ex parte Epeabaka, the High Court considered a challenge based on a Refugee Review Tribunal member’s statement of personal opinion on the internet about his position and role.[121] The joint majority emphasised that conduct which may lead to a reasonable apprehension of bias in respect of a judge might not have the same result in respect of a tribunal member.[122] The variable standard was again emphasised, in the tribunal context, in the recent decision of AZAEY v Minister for Immigration and Border Protection, which this time concerned a challenge based on interruptions, raised voices
and sarcasm.[123]
Looking beyond the migration context, the ministerial perspective was discussed a second time by the High Court in the 2002 decision of Hot Holdings Pty Ltd v Creasy.[124] In that case (concerning a Minister’s decision on an exploration licence) it was emphasised that Ministers (in particular) might in some cases properly have regard to a wide range of considerations — including highly political ones — and that it is wrong to apply in every case the standard of independence applied to a judge.[125] And in the years following, this approach has been taken to various decision-making settings. As a notable example, the South Australian decision in Watson v South Australia confirmed the need for care in applying bias principles in the context of state Cabinet advice to the Governor on a parole decision.[126] The important New South Wales decision of McGovern also emphasised the variable operation of the basic bias test across different contexts, itself according some leniency to local councillors.[127] And the importance of context was again emphasised by the New South Wales Court of Appeal in Duncan v Ipp, a case concerning allegations against the Independent Commission Against Corruption.[128]
The High Court returned to this issue, in broad terms, in the 2015 Isbester decision, emphasising that the application of the Ebner principles to decision-makers other than judges must necessarily recognise and accommodate differences between court proceedings and other kinds of decision-making.[129] Importantly, however, as will be explained below, this decision indicates that the high-water mark has perhaps now been reached for the ‘spectrum of standards’ approach.
While the speciation approach was slower to emerge as a significant force in the contemporary case law, it did reveal its potential efficiency at an early stage. In the very case that moulded the ‘spectrum of standards’ approach into an organised qualification on the reasonable apprehension test, the High Court also overlaid it with a more significant clarification. It identified the bias in issue — ‘prejudgment’ — as a separate sub-species of bias and refined the rules that apply to it (though this has been the subject of lingering confusion). In this case, Jia, the High Court considered principally an allegation of actual bias that rested upon the Minister’s statements (prior to the making of decisions).[130] Gleeson CJ and Gummow J emphasised that for bias in the form of prejudgment, establishing a ‘tendency of mind’ is not sufficient; there must be a mind incapable of alteration, as natural justice does not require the absence of any predisposition or inclination.[131] The narrowness of this explanation of prejudgment was a conspicuous fine-tuning of the rules, or at least a pointed clarification.[132] In another case shortly after, Kirby J was accordingly at pains to emphasise that a defective or illogical approach to evidence or even irrationality in reasoning might create an impression of confusion, carelessness or incompetence, but does not easily lead to a finding of bias of this type.[133]
Unusually, the primary focus in both of these High Court cases was on actual bias — a point which left some apparent confusion as to whether this narrow approach applied to prejudgment generally (ie beyond the actual bias context).[134] This confusion has lingered, in part by reason of vacillation in later prejudgment cases between the basic language of Ebner (the lack of an ‘impartial mind’) and the language of Jia (a mind ‘incapable of alteration’).[135] At an abstract level, the Jia wording might be viewed simply as an elaboration or interpretation of Ebner for the purposes of its application in the specific context of prejudgment.[136] Yet in practical terms, and in the setting of a doctrine so reliant on the lay observer’s perspective, the variation and hence the confusion following Jia is significant.[137] The distinction between the two standards is well illustrated by the recent decision in ALA15 v Minister for Immigration and Border Protection, where it was noted that raw statistics on a judge’s previous decisions were not enough to found a prejudgment claim because a mere propensity of thinking is not enough.[138]
In McGovern, Spigelman CJ attempted some valuable clarification on the critical issue. He emphasised that the correct approach now for reasonable apprehension of bias in the form of prejudgment is to ask whether an independent observer might reasonably apprehend that the decision-maker might not be open to persuasion (in line with the narrower language of Jia).[139] Whilst there continues to be some vacillation between and juxtaposition of terminology, the approach in McGovern appears now to be largely accepted
as correct.[140]
The key point then, in practical terms, is that in an important respect it is harder to establish ‘prejudgment’ than other species of bias. This can be defended as a logical variation given that predisposition (beyond that driven by interest, association or access to extraneous material — the other broad categories of bias) is a natural and proper product of human experience and should be presented and tested rather than denied. More importantly, from the perspective of the arguments in this article, the courts have in this context found the calibration they needed in a ‘speciation’ type approach — ie by carving out a sub-category of bias with a more particularly worded test. The tightening of the applicable test firmly directs the reasonable lay observer and indeed in Jia and McGovern rendered the accompanying ‘spectrum of standards’ somewhat redundant.
Another conspicuous emphasis upon the distinctions between different types of bias came in Isbester.[141] The High Court considered facts approximating the scenario of one person effectively acting as both ‘prosecutor’ and ‘judge’ — and appeared to approve the categorisation of this as ‘incompatibility’.[142] The case concerned a dog-destruction order by a local council committee — open to it by reason that the owner had been convicted of an offence under the legislation.[143] One committee member had substantial involvement in the prosecution as well as in the destruction decision. The Court held that there was a reasonable apprehension of bias notwithstanding that the committee’s decision was in a sense a second step on a different issue (ie penalty) — particularly because much of the same evidence would be relevant and no clear line could be drawn between the member’s different involvements.[144]
Importantly for present purposes, the Court appeared to accommodate this notion of ‘incompatibility’ within the category of ‘interest’ bias by explaining that ‘interest’ for these purposes did not require some material or other benefit but might merely involve vindication of a person’s earlier opinion and role.[145] In the course of its reasoning the Court carefully distinguished this situation from ‘prejudgment’ (which logically frees the assessment from the narrower Jia version of the lay observer’s inquiry). It was also noted that the Ebner second step (the search for ‘connection’ with feared impartiality) was ‘necessarily ... satisfied’ or ‘obvious’ in incompatibility scenarios[146] — and indeed that the standard applied in this context would in fact be similar to that applied in high-level decision-making.
The last point is significant for present purposes. However, a few other implications should first be noted. First, there might in future be some fine lines to be drawn between ‘incompatibility’ bias (involving an ‘interest’ — including vindication of an earlier view) and other types of ‘pre-involvement’ (which will presumably remain under the prejudgment rules). The hazy boundary here is now significant by reason of the several apparent differences in the approaches to be applied. Secondly, the discarding of the Ebner second step might logically also seem apposite for cases of ‘association’ bias (as has been already been suggested)[147] and even for other categories of ‘interest’ bias.[148]
Most interesting, from the perspective of this article, is Isbester’s visible retreat from the ‘spectrum’ approach — at least in the case of what we might now term ‘incompatibility’ (ie acting as both prosecutor and judge). A key issue before the Court was whether the rules on this point were to be less stringently applied in a non-judicial context. The Court essentially responded ‘no’. Most particularly, it was said at one point that identifying a decision as ‘quasi-judicial’ (or not) is distracting in such cases.[149] At another point their Honours clearly indicated that ‘incompatibility’ operates similarly for judges or ‘member[s] of some other decision-making body’.[150]
It might be argued that this carries no significant change in direction — that exceptions to the prominent spectrum approach were simply hidden for a time by the case law’s focus upon prejudgment-type claims (and the lack of an unusual case to test its perimeter). In any event, we must now ask whether the ‘spectrum’ approach, so influential in the last two decades, should also be discarded or doubted in other contexts. It may be, for example, that this flexible approach is now unlikely to be applied in any other cases of ‘interest’ bias — both the older pecuniary interest claims and more contemporary conflict of interest scenarios. If the question is what the reasonable lay observer would tolerate, it might seem logical that they would be similarly unforgiving in ‘interest’ scenarios whatever the level of decision-maker. Similar arguments might be made as regards ‘association’ bias cases and
even ‘extraneous information’ cases. It is conceivable that ultimately the spectrum methodology might only enjoy firm footing in the ‘prejudgment’ category of bias, which might itself now be at risk of some incursion by the notion of ‘incompatibility’ given possible overlaps in the context of multiple involvements.
In Isbester, the speciation methodology excluded the spectrum approach. In cases such as Jia and McGovern, it rendered the spectrum analysis somewhat redundant. The prediction in this article is that speciation might become the courts’ preferred tool for meeting new arguments and adjusting old principle. This is based on the thinking that, first, the reasoning in Isbester has logically set up a further retreat from the spectrum approach, and second, the speciation methodology might seem to be more predictable and precise — not least because it can in a sense actually refine the applicable test (rather than just the sensitivity of the lay observer).
The rule against bias provides the quintessential example of context-driven legal evolution. The testing of old principle in this field of Australian administrative law has come from many directions — from the expansion, diversification and increased engagement of government; from the growing complexity of decision-making processes; from the sharpening political dimensions of decisions; from the variegation of potentially competing ‘interests’; and from the growing sophistication of the public in terms of both understanding
and expectation.
In broad terms, the purpose of this article has been to explore how the courts have grappled with these challenges, through the frenetic case load of recent years, in their efforts to keep this sacrosanct legal rule relevant and coherent. More specifically, it has examined how the doctrine has evolved and might continue to evolve via the use of three key tools of ‘calibration’ in
its application.
The ongoing deployment of the ‘reasonable lay observer’, as a measure of risk and perception, reminds us of the deep importance of public confidence in this context. However, close analysis of some difficult applications of the standard confirm that the lay observer, so prominently wielded as the yardstick since Webb and Ebner, is somewhat worn and wearied — and presents an increasingly indistinct path through a complex contemporary array of challenges. The courts have felt this and have stretched the imputed knowledge and understanding of the lay observer as far as possible to house the increasingly sophisticated analysis and balancing needed in the cases. Yet importantly, the courts have now also been looking elsewhere for responsiveness and modern coherency in the doctrine, and have been seeking rational bases for their necessary intrusion and extrapolation in the lay observer’s notional assessment of difficult contemporary scenarios.
Most prominently to date, the courts have applied a ‘spectrum of standards’ approach, which in a sense allows them to moderate the sensitivity of the notional lay observer by reference to a hierarchy of decision-making contexts. However, it appears that after a decade and a half of keen application this methodology has now reached its high-water mark. It was clearly rejected as a basis for applying a more lenient approach to ‘incompatibility’ bias in Isbester. It may yet retreat further — conceivably back to its heartland in the ‘prejudgment’ cases.
Isbester reveals that the retreat from the ‘spectrum’ approach perhaps reflects a new trajectory in the jurisprudence, one built around the sub-categorisation or ‘speciation’ of bias. From a quieter start, this methodology is now gaining momentum and the significance of the distinctions drawn is certainly deepening. ‘Prejudgment’ and ‘incompatibility’ are now governed by somewhat tailor-made principles, and there are signs that more categories might receive similarly specific attention in the near future. Even the basic distinction between ‘actual’ and ‘apprehended’ bias appears to be sharpening, as the High Court insists on a clear division of inquiries in the lower courts.[151]
From a broader vantage point, how might we assess this recent history of evolution in the rule against bias? This study would seem to reveal that while the challenges of modern context are unavoidable in this field of law, the principled guidance offered by the new tools of ‘calibration’ can help to make the tiring lay observer test more predictable and sustainable — in a sense they allow the courts to carefully tailor the lay observer’s brief to help them navigate the contemporary contests. But should the lay observer test be sustained, given that it now clearly costumes a more sophisticated analysis? The conclusion here is still ‘yes’. While it does remain — even somewhat faded and ill-fitting — the notion of public perception is not forgotten[152] and hence the increasingly sophisticated court decisions are unlikely to stray too far from their critical function of reassuring litigants and the public at large that the system is ‘fair’.[153]
There are in fact a number of core purposes that might be attributed to the rule against bias (as noted at the outset of this article). In broad terms, these include its role in sustaining the quality of decision-making, its shoring up of the confidence and cooperativeness of individuals being affected and ultimately its contribution to maintaining the coherency, propriety and security of the social order. The future balancing of the various methodologies identified in this paper — the three tools of ‘calibration’ — might benefit from some renewed deliberation on these fundamental purposes and the priority accorded to each. Added to this, of course, are issues of workability. While the ‘speciation’ approach to bias might appear to be gaining momentum, and might appear to offer some new precision, there are challenges lining this path. In the first place, the jurisprudential history illustrates that the range of possible bias scenarios is potentially large. Secondly, growing intricacy and variability in the application of the rules might in some ways unsteady the all-important public confidence that they serve. And most importantly, as noted in the course of this article, finding clarity in the division between categories might not be an easy task. The hazy distinctions inherent in this field of law were at the heart of both the original Ebner decision and indeed the recent High Court decision in Isbester.
What must be conceded at this juncture is that the rules on bias are growing increasingly complex, despite the ostensible simplification in Ebner that ushered in the modern era of case law. This phenomenon is perhaps an inherent trait of contemporary common law process, despite the best efforts of judges to organise and streamline. Often the periodic attempts to simplify just return us to the beginning of another journey. Certainly in the natural justice context, the sheer volume and variety of arguments presented to the courts produce a legal entropy that is hard to curtail. Moreover, the idea of ‘fairness’ has proven itself over many years to be at once both a fundamentally simple and intractably complex concept.
[*] BA, LLB (Hons) (Qld), AMusA, LLM (QUT), PhD (Dist) (UWA); Professor of Law and Justice, University of Southern Queensland; Adjunct Professor, School of Law, The University of Western Australia; External Fellow, Centre for Public, International and Comparative Law, The University of Queensland. The author thanks Ann-Louise Shrapnel for initial research assistance, and Bill Lane and the anonymous referees for their insightful comments.
[1] GRC Davis, Magna Carta (British Library, 1977) 29.
[2] Ibid 30. For brief overviews of the history of the rule, see John Tarrant, Disqualification for Bias (Federation Press, 2012) ch 2; Lord Woolf et al, De Smith’s Judicial Review (Sweet & Maxwell, 7th ed, 2013) 536–8 [10-008]–[10-011].
[3] See, eg, Assistant Commissioner Condon v Pompano Pty Ltd (2013) 252 CLR 38, 107 [186] (Gageler J). See generally Matthew Groves, ‘The Evolution and Entrenchment of Natural Justice’ in Matthew Groves (ed), Modern Administrative Law in Australia: Concepts and Context (Cambridge University Press, 2014) 205.
[4] On the issue of greater engagement by the judiciary, see, eg, the recent commentary on public comment by judges: Susan Bartie and John Gava, ‘Some Problems with Extrajudicial Writing’ [2012] SydLawRw 30; (2012) 34 Sydney Law Review 637; Chris Finn, ‘Extrajudicial Speech and the Prejudgment Rule: A Reply to Bartie and Gava’ [2013] AdelLawRw 16; (2014) 34 Adelaide Law Review 267; Matthew Groves, ‘Public Statements by Judges and the Bias Rule’ (2014) 40 Monash Law Review 115; Jasmin Moran, ‘Courting Controversy: The Problems Caused by Extrajudicial Speech and Writing’ (2015) 46 Victoria University of Wellington Law Review 453.
[5] For emerging critique of traditional recusal processes beyond the scope of this article, see, eg, Grant Hammond, Judicial Recusal: Principles, Process and Problems (Hart, 2009); Abimbola A Olowofoyeku, ‘Bias in Collegiate Courts’ (2016) 65 International and Comparative Law Quarterly 895; Gabrielle Appleby and Stephen McDonald, ‘Pride and Prejudice: A Case for Reform of Judicial Recusal Procedure’ (2017) 20 Legal Ethics 89.
[6] See Stephen Gageler, ‘Impact of Migration Law on the Development of Australian Administrative Law’ (2010) 17 Australian Journal of Administrative Law 92.
[7] For detailed and enduring overviews, in addition to those already cited above, see Robin Creyke, John McMillan and Mark Smyth, Control of Government Action: Text, Cases and Commentary (LexisNexis Butterworths, 4th ed, 2015) ch 10; Mark Aronson, Matthew Groves and Greg Weeks, Judicial Review of Administrative Action and Government Liability
(Lawbook, 6th ed, 2017), ch 9.
[8] [2000] HCA 63; (2000) 205 CLR 337.
[9] For previous critique of the state of the ‘reasonable observer’ test, see, eg, John Griffiths, ‘Apprehended Bias in Australian Administrative Law’ (2010) 38 Federal Law Review 353; Matthew Groves, ‘The Imaginary Observer of the Bias Rule’ (2012) 19 Australian Journal
of Administrative Law 188; Anna Olijnyk, ‘Apprehended Bias: A Public Critique of the
Fair-Minded Lay Observer’, AUSPUBLAW (Blog Post, 3 September 2015) <http://auspublaw.org/2015/09/apprehended-bias/> , archived at <https://perma.cc/YEU3-H8VN>; Justice Debbie Mortimer, ‘Whose Apprehension of Bias?’ [2016] AIAdminLawF 14; (2016) 84 AIAL Forum 45.
[10] See, eg, Simon Atrill, ‘Who Is the “Fair-Minded and Informed Observer”? Bias after Magill’ (2003) 62 Cambridge Law Journal 279; Abimbola A Olowofoyeku, ‘Bias and the Informed Observer: A Call for a Return to Gough’ (2009) 68 Cambridge Law Journal 388.
[12] [1994] HCA 30; (1994) 181 CLR 41.
[13] Ebner (n 8) 344 [6] (Gleeson CJ, McHugh, Gummow and Hayne JJ). As to the development of the concept of the ‘lay observer’ (and terminology employed) in the years before Ebner, see Groves, ‘The Imaginary Observer of the Bias Rule’ (n 9) 118–19.
[15] For recent confirmation, see Isbester v Knox City Council [2015] HCA 20; (2015) 255 CLR 135, 146 [22] (Kiefel, Bell, Keane and Nettle JJ).
[16] See Johnson v Johnson [2000] HCA 48; (2000) 201 CLR 488, 493 [12] (Gleeson CJ, Gaudron, McHugh, Gummow and Hayne JJ).
[17] British American Tobacco Australia Services Ltd v Laurie (2011) 242 CLR 283, 300 [33] (French CJ).
[18] See generally Griffiths (n 9); Groves ‘The Imaginary Observer of the Bias Rule’ (n 9).
[19] See, eg, Lawrie v Lawler [2016] NTCA 3; (2016) 168 NTR 1, 22 [106]; Day v Woolworths Ltd [2016] QCA 321. But see Islam v Minister for Immigration and Citizenship [2009] FCA 1526; (2009) 51 AAR 147, 155 [36].
[20] See, eg, Stone v Moore [2015] SASC 46; (2015) 122 SASR 54, 76–7 [51]–[54]; ALA15 v Minister for Immigration and Border Protection [2016] FCAFC 30, [28]; Kennedy v Secretary, Department of Industry [No 2] [2016] FCA 746; Ferdinands v South Australia [2016] FCA 1268, [36]; SZUEP v Minister for Immigration and Border Protection [2017] FCAFC 94, [12].
[21] See, eg, Cabcharge Australia Ltd v Australian Competition and Consumer Commission [2010] FCAFC 111, [32]; ResMed Ltd v Australian Manufacturing Workers’ Union [2015] FCAFC 106; (2015) 232 FCR 152, 158–9 [32]–[33].
[22] See, eg, British American Tobacco (n 17) 330–1 [137] (Heydon, Kiefel and Bell JJ).
[23] See, eg, Duncan v Ipp [2013] NSWCA 189; (2013) 304 ALR 359, 369 [147].
[24] See, eg, Eastman v The Queen [2015] ACTCA 24; (2015) 295 FLR 426; Trombone Investments Pty Ltd v TBT (Victoria) Pty Ltd [2016] VSCA 108; Crossman v Sheahan [2016] NSWCA 200; (2016) 115 ACSR 130; Golden v V’landys [2016] NSWCA 300; (2016) 339 ALR 610; Melbourne City Investments Pty Ltd v UGL Ltd [2017] VSCA 128; Kuek v Phillips [2017] VSC 332, [77]–[78].
[25] See, eg, Golden (n 24) 623 [89]. Note in this respect Basten JA’s apparent resistance to the ‘diminution of the double “might” test’: Crossman (n 24) 140 [33].
[27] Isbester (n 15) 146 [23] (Kiefel, Bell, Keane and Nettle JJ) (citations omitted).
[28] See, eg, Weinstein v Medical Practitioners Board of Victoria [2008] VSCA 193; (2008) 21 VR 29, 39 [26] (understanding of the character of an investigatory panel’s function?); O’Sullivan v Medical Tribunal of New South Wales [2009] NSWCA 374, [40]–[43] (knowledge of the nature of extraneous documents, the nature of the proceedings and the composition and powers of
the tribunal?); British American Tobacco (n 17) 330 [134] (knowledge of special evidentiary provisions?).
[29] Johnson (n 16) 508 [52]. It has been pointed out that the observer should be considered as a single amalgam representing society — one that cannot be distilled to ‘particular personal qualities’: Groves, ‘The Imaginary Observer of the Bias Rule’ (n 9) 190.
[30] AB v DPP (NSW) [2016] NSWCA 73, [21].
[31] Johnson (n 16) 509 [53] (Kirby J).
[33] GlaxoSmithKline Australia Pty Ltd v Reckitt Benckiser Healthcare (UK) Ltd [2013] FCAFC 150, [38] (‘GSK’).
[34] Zhai v Luo [2015] FCAFC 144, [38]. But see Eastman (n 24) 432 [31].
[35] For earlier expressed concerns, see, eg, Smits v Roach (2006) 227 CLR 423, 457 [96] (Kirby J). See generally Groves, ‘The Imaginary Observer of the Bias Rule’ (n 9) 199–202.
[36] For discussion of the pre-Webb and Ebner position, see Groves, ‘The Imaginary Observer of the Bias Rule’ (n 9) 188–9.
[37] See British American Tobacco (n 17) 306–7 [48] (French CJ).
[38] See the call for explicit attention to other policy objectives in Atrill (n 10).
[39] See the discussion of the ‘spectrum of standards’ in Part III(B).
[40] See, eg, Re Refugee Review Tribunal; Ex parte H [2001] HCA 28; (2001) 179 ALR 425; SZRUI v Minister for Immigration, Multicultural Affairs and Citizenship [2013] FCAFC 80; Ikosidekas v Karkanis [2015] VSCA 121; MZZLO v Minister for Immigration and Border Protection [2015] FCA 963; MZZLO v Minister for Immigration and Border Protection [No 2] [2016] FCA 356; (2016) 246 FCR 111.
[41] GSK (n 33) [36]. See also R v Watson; Ex parte Armstrong [1976] HCA 39; (1976) 136 CLR 248, 264; Concrete Pty Ltd v Parramatta Design & Developments Pty Ltd [2006] HCA 55; (2006) 229 CLR 577, 610 [112]; Scott v DPP (NSW) [2015] NSWCA 60, [27]–[37]; Ryan v Vizovitis [2017] ACTCA 3, [87].
[42] Michael v Western Australia [2007] WASCA 100. See also Elsafty Enterprises Pty Ltd v Gold Coast City Council [2011] QCA 84.
[43] Reznitsky v DPP (NSW) [2015] NSWCA 338. See also George v Fletcher (2012) 10 ABC(NS) 301, 333 [158]–[159] (words including ‘be quiet. Just shut up’ insufficient to raise a reasonable apprehension of bias when taken in context); Waddington v Dandenong Magistrates’ Court [2014] VSCA 12, [19]–[23] (‘abruptness’ found to be warranted in the circumstances).
[44] Zorbas v Sidiropoulous [No 2] [2009] NSWCA 197, [38]–[46]. See also George (n 43) 329–30 [140]–[144]; Clyne v New South Wales [2012] NSWCA 265.
[45] SZJBD v Minister for Immigration and Citizenship [2009] FCAFC 106; (2009) 179 FCR 109. See also Tarrant v Australian Securities and Investments Commission [2015] FCAFC 8; (2015) 317 ALR 328.
[46] AXQ15 v Minister for Immigration and Border Protection [2016] FCAFC 73.
[47] Ball v McInerney [2014] NSWCA 331; R v T, WA (2014) 118 SASR 382.
[48] Lancaster v The Queen [2014] VSCA 333; (2014) 44 VR 820.
[50] Spanos v Lazaris [2008] NSWCA 74, [43].
[51] For a rare example of a successful challenge, relating to statements in the primary hearing and reasons that the appellant’s claim was a ‘bag of chips’, a ‘try on’, a ‘baseless trifle’ and ‘doomed’, and that the claim depended on a ‘rubbish’ proposition and was an ‘obvious abuse of process’, see Hinton v Alpha Westmead Private Hospital [2016] FCAFC 107; (2016) 242 FCR 1, 10 [25].
[52] Spanos (n 50) [43]. See also Reznitsky (n 43).
[53] As to the significance of context in an assessment of critical comments by a judge, see also Barakat v Goritsas [No 2] [2012] NSWCA 36.
[54] Spanos (n 50) [41]–[48]. See also Downey v Acting District Court Judge Boulton [No 5] [2010] NSWCA 240; (2010) 272 ALR 705. But contrast AJH Lawyers Pty Ltd v Careri [2011] VSCA 425; (2011) 34 VR 236.
[55] [2016] VSCA 174, [84], [99], [102].
[56] See De Alwis v Western Australia [No 2] [2015] WASCA 42, [89] (McLure P).
[57] British American Tobacco (n 17) 333 [144] (Heydon, Kiefel and Bell JJ).
[58] See, eg, De Alwis [No 2] (n 56) [71]; Ferdinands (n 20).
[59] Johnson (n 16) 492–3 [12]–[13] (Gleeson CJ, Gaudron, McHugh, Gummow and Hayne JJ). See also Szaintop Homes Pty Ltd v Krok [2012] VSCA 176; (2012) 36 VR 56, 64–5 [28]–[29]; Bahonko v Moorfields Community [2012] VSCA 89; (2012) 34 VR 409, 422 [32]; Duncan (n 23) 397 [155]–[156]; Stone (n 20) 80–1 [63].
[60] The focus here is on multiple involvements in a decision-making capacity. On the issue of earlier conflicting roles, see Part III(C)(2). On the issue of earlier provision of advice on point, see, eg, Gabrielle Appleby and Stephen McDonald, ‘Disqualification of Judges and Pre-Judicial Advice’ (2015) 43 Federal Law Review 201.
[61] See, eg, Goodwin v Commissioner of Police [2012] NSWCA 379; Pavlovic v Universal Music Australia Pty Ltd [No 2] [2016] NSWCA 31; Hinton (n 51).
[62] See, eg, MZZXM v Minister for Immigration and Border Protection [2016] FCA 405;
MZAEU v Minister for Immigration and Border Protection [2016] FCAFC 100; (2016) 70 AAR 22.
[63] [1983] HCA 17; (1983) 151 CLR 288.
[64] See also Islam (n 19) (tribunal member involved both in the issue of a listening device — on suspicion of criminal conduct — and in the later refusal of a visa to the relevant person on the basis of criminal conduct); Deputy Commissioner of Taxation v Chemical Trustee Ltd
[No 9] [2015] FCA 1178; Frugtniet v Tax Practitioners Board [2015] FCA 1066; (2015) 67 AAR 336; Brown Brothers Waste Contractors Pty Ltd v Pittwater Council [2015] NSWCA 215; (2015) 90 NSWLR 717; Picos v Servcorp Ltd [2015] FCA 344; Amos v Wiltshire [2015] QCA 44; R v Kay; Ex parte A-G (Qld) [2016] QCA 269; Crossman (n 24).
[65] See, eg, Priestley v Godwin [2008] HCA 59; (2008) 251 ALR 612; George (n 43); Bob Jane Corporation Pty
Ltd v ACN 149 801 141 Pty Ltd [2015] FCA 1343; Berry v Commissioner of Taxation [2015] FCA 1244; (2015) 149 ALD 270; Zhai (n 34); Asden Developments Pty Ltd (in liq) v Dinoris [2015] FCA 729; Shaw v Australian Pump Industries Pty Ltd [2015] FCA 547; Ikosidekas (n 40); Yule v Irwin [2016] SASC 30; Melbourne City Investments Pty Ltd v UGL Ltd [2017] VSCA 128; Asden Developments Pty Ltd (in liq) v Dinoris [2017] FCAFC 117. See generally Anna Olijnyk, ‘Apprehended Bias and Interlocutory Judgments’ [2013] SydLawRw 30; (2013) 35 Sydney Law Review 761.
[66] See, eg, British American Tobacco (n 17) 330–1 [134] (Heydon, Kiefel and Bell JJ); Reece v Webber [2011] FCAFC 33; (2011) 192 FCR 254, 270–5 [43]–[60].
[67] For a clear example of previous involvement on the very same issues, see Maher v Adult Guardian [2011] QCA 225. For separate involvement on materially different matters, see Lashansky v Legal Practice Board [2011] WASCA 42. See also Berg v DPP (Qld) [2015] QCA 196; [2016] 2 Qd R 248; Amos (n 64). As to the distinct situation of previous statements of opinion on legal issues, see ResMed (n 21).
[68] British American Tobacco (n 17).
[69] Ibid 295 [21] (French CJ).
[70] Ibid 333 [145] (Heydon, Kiefel and Bell JJ). Contrast the dissenting judgment of French CJ, with its emphasis on the interlocutory nature of the first finding, the significant time lapse and the strength of the express qualifications: at 308 [50].
[71] (2011) 244 CLR 427.
[72] Ibid 447 [70]. As to the insufficiency of mere error, see also Reid v Commercial Club (Albury) Ltd [2014] NSWCA 98.
[73] Wilson (n 71) 447–8 [69]–[73].
[74] See, eg, Watson (n 41) 264–6; Livesey (n 63) 300.
[75] See Crossman (n 24) 140 [33]. In British American Tobacco (n 17) 328–30 [130]–[135], there was much argument before the High Court on whether the New South Wales Court of Appeal had attributed too much technical knowledge to the lay observer, particularly as regards the special evidential rules governing proceedings of the type in question.
[76] For an overview of the arguments for a broader approach, see Olijnyk, ‘Apprehended Bias and Interlocutory Judgments’ (n 65) and the critique and defence of Cadbury Schweppes Pty Ltd v Darrell Lea Chocolate Shops Pty Ltd [No 2] [2009] FCAFC 8; (2009) 174 FCR 175 in Groves, ‘The Imaginary Observer of the Bias Rule’ (n 9) 193–4.
[77] See, eg, Re Finance Sector Union of Australia; Ex parte Illaton Pty Ltd (1992) 107 ALR
581 (‘Illaton’).
[78] Johnson (n 16) 493 [13] (Gleeson CJ, Gaudron, McHugh, Gummow and Hayne JJ). See also Szaintop (n 59) 64–5 [28]–[29]; Bahonko (n 59) 422 [32]; Duncan (n 23) 381 [64].
[79] See generally Builders’ Registration Board of Queensland v Rauber (1983) 47 ALR 55.
[80] See, eg, Illaton (n 77); Ebner (n 8) 359 [64]–[65].
[81] British American Tobacco Australia Services Ltd v Laurie [2009] NSWCA 414 (‘British American Tobacco (NSWCA)’).
[82] British American Tobacco (n 17) 331 [138] (Heydon, Kiefel and Bell JJ).
[83] Ibid 330–1 [137]. See also Orleans Investments Pty Ltd v MindShare Communications Ltd [2009] NSWCA 40; (2009) 254 ALR 81, 94 [36]; CUR24 v DPP (NSW) [2012] NSWCA 65; (2012) 83 NSWLR 385, 395–6 [39]; Barakat (n 53) [62]–[63].
[84] British American Tobacco (n 17) 330–1 [136]–[138].
[85] Note also Basten JA’s concern that to disregard judges’ own assertions of impartiality might lead to perceptions that challengers are ‘manipulating the system’ — not attempting ‘to avoid a prejudiced mind, but [simply] an adverse result’: British American Tobacco (NSWCA)
(n 81) [140]. See also Barakat (n 53) [58]–[64] (Basten JA, Young JA and Sackville AJA agreeing); Kay (n 64) [30]–[34].
[86] Wilson (n 71) 446–7 [67]–[68], referring to Nicholls v Michael Wilson & Partners Ltd (2010) 243 FLR 177, 200 [91]. See also Goodwin (n 61) [15]; Bagshaw v DPP (NSW) [2016] NSWCA 340, [62].
[87] Wilson (n 71) 446–7 [67].
[88] Zhang v Minister for Immigration and Border Protection [2015] FCA 935, [31]–[38]; Asden Developments Pty Ltd (in liq) v Dinoris [2017] FCAFC 117, [48]–[49].
[89] MZZXM (n 62) [123]–[131]; Zaburoni v Minister for Immigration and Border Protection [2017] FCA 654, [73].
[90] As to the latter, see Shaw (n 65) [37]–[39] (mitigating comments at later listings considered); Kuek v Phillips [2017] VSC 332, [74].
[91] See generally Vakauta v Kelly [1989] HCA 44; (1989) 167 CLR 568, 573; Re Minister for Immigration and Multicultural Affairs; Ex parte Epeabaka [2001] HCA 23; (2001) 206 CLR 128, 139 [29]; Islam (n 19) 156–9 [41]–[55]; MZZLO (n 40) [38].
[92] Re Henry; JL v Secretary, Department of Family and Community Services [2015] NSWCA 89, [235]. See also MZAEU (n 62) 41–2 [45]. But see Young v King [2016] NSWCA 282,
[20]–[24] (Basten JA, Gleeson JA agreeing).
[93] For a case where a judge’s handling of recusal applications was seen to heighten the
fair-minded observer’s concerns about their conduct in the substantive matter, see AJH Lawyers (n 54).
[94] See, eg, Minister for Immigration and Citizenship v SZJSS [2010] HCA 48; (2010) 243 CLR 164, 178 [44]; Zhai (n 34) [32].
[95] See generally Olowofoyeku, ‘Bias in Collegiate Courts’ (n 5).
[96] See McGovern v Ku-ring-gai Council [2008] NSWCA 209; (2008) 72 NSWLR 504, 553–7 [238]–[250] (Campbell JA).
[97] [1972] HCA 53; (1972) 128 CLR 509, 519 (Barwick CJ), 520 (Menzies J), 526 (Gibbs J). See also Re Macquarie University; Ex parte Ong (1989) 17 NSWLR 113. Cf Laws v Australian Broadcasting Tribunal [1990] HCA 31; (1990) 170 CLR 70.
[99] Ibid 392–3.
[100] [2002] HCA 51; (2002) 210 CLR 438.
[102] Ibid 524 [100]–[103].
[103] Ibid 510 [31], 513 [45]–[46], 515 [62].
[104] Isbester (n 15) 153 [48] (Kiefel, Bell, Keane and Nettle JJ); cf at 158 [65] (Gageler J).
[105] In Marrickville Metro Shopping Centre Pty Ltd v Marrickville Council [2010] NSWCA 145; (2010) 174 LGERA 67, 104–5 [165]–[169], Tobias JA suggests that there might be a distinction drawn for these purposes between legislative-type functions and judicial-type ones. See also Isbester (n 15) 156 [61] (Gageler J).
[107] See, eg, Dimes v Proprietors of Grand Junction Canal [1852] EngR 789; (1852) 3 HL Cas 759; 10 ER 301.
[108] See generally Aronson, Groves and Weeks (n 7) 674–6 [9.140].
[109] See, eg, Webb (n 12); Dovade Pty Ltd v Westpac Banking Group [1999] NSWCA 113; (1999) 46 NSWLR 168; Clenae Pty Ltd v Australia & New Zealand Banking Group Ltd [1999] VSCA 35; [1999] 2 VR 573.
[110] See Wilson (n 71) 437–8 [33].
[111] See Re Minister for Immigration and Multicultural Affairs; Ex parte Applicant S20/2002 [2003] HCA 30; (2003) 198 ALR 59, 81 [98] (‘Applicant S20’).
[112] See, eg, British American Tobacco (n 17) 302 [38] (French CJ), citing Webb (n 12) 74
(Deane J).
[113] See, eg, Bahonko (n 59) 419 [23]; PLP v Legal Services Commissioner [2014] VSCA 253,
[35]–[39].
[114] See, eg, Underwood v Gaudron [2015] NSWCA 269; (2015) 324 ALR 641; GJ v AS [No 2] (2015) 299 FLR 13.
[115] See, eg, Carruthers (n 98) 376.
[117] See Part III(C).
[118] (2001) 205 CLR 507, 539 [102]. See also Illaton (n 77); Matthew Groves, ‘The Rule against Bias’ in Matthew Groves and HP Lee (eds), Australian Administrative Law: Fundamentals, Principles and Doctrines (Cambridge University Press, 2007) 316, 318.
[120] Ibid 539 [102]. See also Fraser v Minister for Immigration and Border Protection [2015] FCAFC 48; (2015) 145 ALD 337; Zaburoni v Minister for Immigration and Border Protection [2017] FCA 654. For a detailed account of the history of the Jia litigation and critique of the finding that there was no bias made out, see Alan Freckelton, ‘The Benefit of Law, the Devil and the Jia Litigation’ (2015) 23 Australian Journal of Administrative Law 37.
[122] Ibid 138 [27].
[123] [2015] FCAFC 193; (2015) 238 FCR 341, 353 [48]. See also SZRUI (n 40) [25] (Flick J), [93] (Robertson J).
[125] Ibid 455 [50].
[126] [2010] SASCFC 69; (2010) 278 ALR 168.
[127] McGovern (n 96) 514 [51], 519 [80]. See also Marrickville (n 105) 104–5 [165]–[169].
[128] Duncan (n 23) 397 [152]–[156]. See also Carruthers (n 98) (commission of inquiry); Murlan Consulting Pty Ltd v Ku-ring-gai Municipal Council [2009] NSWCA 300; (2009) 170 LGERA 162 (commissioners and judges distinguished). For broader applications of the ‘spectrum’, see, eg, Greyhound Racing NSW v Cessnock & District Agricultural Association Inc [2006] NSWCA 333 (statutory corporation overseeing racing); Carbotech-Australia Pty Ltd v Yates [2008] NSWSC 540 (referee providing report); Colonial Range Pty Ltd v Akritidis [2017] VSC 337 (private building surveyor carrying out statutory functions).
[129] Isbester (n 15) 146–8 [22]–[28] (Kiefel, Bell, Keane and Nettle JJ).
[131] Ibid 531–2 [71]–[72]. See also, again with particular emphasis on the ministerial context and the nature of the discretion involved, Fraser v Minister for Immigration and Border Protection [2015] FCAFC 48; (2015) 145 ALD 337.
[132] Cf the earlier comments in R v Commonwealth Conciliation and Arbitration Commission;
Ex parte The Angliss Group (1969) 122 CLR 546, 555; Vakauta (n 91) 570–1; Helljay Investments Pty Ltd v Deputy Commissioner of Taxation [1999] HCA 56; (1999) 166 ALR 302, 307 [12]. See generally Finn (n 4).
[133] Applicant S20 (n 111) 82 [101].
[134] The general application of the approach was, however, quite clearly implied in the comments of Kirby J in Applicant S20: ibid 81–2 [98]–[101].
[135] For recent examples, see Mogan Holdings Pty Ltd v Harrison [2011] VSCA 202; Chin v Legal Practice Board (WA) [2011] WASCA 110; Bahonko (n 59); George (n 43); Duncan (n 23); B v DPP (NSW) [2014] NSWCA 232; Reid v Commercial Club (Albury) Ltd [2014] NSWCA 98; De Alwis v Western Australia [No 4] [2015] WASCA 43; Berry (n 65); Brown Brothers (n 64); Ikosidekas (n 40); Scott (n 41); Mahmoud v AG (NSW) [2016] NSWCA 263; Crossman (n 24); Ryan (n 41).
[136] See, eg, the implication in McGovern (n 96) 517 [72] (Basten JA).
[137] See Marrickville (n 105) 100–1 [146]–[153].
[139] McGovern (n 96) 508–9 [15]–[18]; see also at 517 [72] (Basten JA).
[140] See, eg, British American Tobacco (n 17) 322 [104]; CUR24 (n 83) 394–5 [36]; SZRUI (n 40) [29]; Zhai (n 34) [20]; Minister for Immigration and Border Protection v SZTJF [2015] FCA 1052; (2015) 149 ALD 552, 569 [78]; Brown Brothers (n 64) 749 [137]; Zaburoni v Minister for Immigration and Border Protection [2017] FCA 654, [67].
[142] Ibid 149 [34] (Kiefel, Bell, Keane and Nettle JJ). Distinguish the situation of a prosecutor wishing to oppose an appeal: Viscariello v Legal Profession Conduct Commissioner [2015] SASC 132, [65]–[67].
[143] Domestic Animals Act 1994 (Vic) ss 29(4), 84P(e).
[144] Isbester (n 15) 151 [40]–[42] (Kiefel, Bell, Keane and Nettle JJ).
[145] Ibid 152 [46]. See also Golden (n 24) 626–7 [111]–[119].
[146] Isbester (n 15) 152 [47], 153 [49] (Kiefel, Bell, Keane and Nettle JJ). See also Golden (n 24) 622–3 [86]–[90], 625–6 [106]–[110], where the Ebner second step appeared only to be pursued in an alternative formulation of the bias argument.
[147] See Murlan Consulting (n 128) 178–9 [62] (Basten JA, Macfarlan JA agreeing); Spencer v Bamber [2012] NSWCA 274, [15] (Basten JA). But see A-G (NT) v DPP (NT) (2013) 166 NTR 1, 8 [28]–[29]; Hurd v Zomojo Pty Ltd [2015] FCAFC 147; (2015) 110 ACSR 38, 50 [62]; Mueller v Que Capital Pty Ltd [No 2] [2016] WASCA 157, [32]–[33].
[148] See, eg, Day v Sanders [2015] NSWCA 324; (2015) 90 NSWLR 764, 784 [92] (Basten JA). But see the apparent implication in Isbester (n 15) 152 [47]. See generally Griffiths (n 9).
[149] Isbester (n 15) 148–9 [31] (Kiefel, Bell, Keane and Nettle JJ).
[150] Ibid 149 [34].
[151] See, eg, Wilson (n 71) 446 [67]. See also the reasoning behind the Court’s rejection of any requirement of ‘personal animus’ in a reasonable apprehension challenge in Golden (n 24)
625 [108].
[152] See British American Tobacco (n 17) 306–7 [48] (French CJ). See also Groves, ‘The Imaginary Observer of the Bias Rule’ (n 9) 202.
[153] See Johnson (n 16) 507–8 [52] (Kirby J). See also R v Sussex Justices; Ex parte McCarthy [1924] 1 KB 256, 259; British American Tobacco (n 17) 331–2 [139]–[140] (Heydon, Kiefel and Bell JJ).
AustLII:
Copyright Policy
|
Disclaimers
|
Privacy Policy
|
Feedback
URL: http://www.austlii.edu.au/au/journals/MelbULawRw/2017/42.html