ńňđ. 13
(âńĺăî 19)



utive government.59 In Department of Immigration and Ethnic Affairs v Ram,60
Hill J expressed his reservations on the efficacy of the 1995 joint statement
achieving its intended effect. Hill J doubted that Mason CJ and Deane J had:
. . . contemplated a case where at the time of ratification, Australia had expressed to
the world and to its people its intention to be bound by a treaty protecting the rights
of children, but subsequently one or more Ministers made statements suggesting that
they at least had decided otherwise.

Although the High Court has not had the opportunity to consider the legal effect
of either joint statement, the fact that no judge in Lam suggested that the decision
in Teoh had been rendered nugatory by the executive government’s actions indi-
cates that the High Court continues to regard Teoh as arguable, even if members
of the Court have doubts about the correctness of the majority’s approach.

The future impact of international treaties in Australian
administrative law
While most of the High Court’s decision in Lam examines the concept of legiti-
mate expectations (discussed in Implications for procedural fairness and the role of

legitimate expectations – see page 307–8), members of the High Court could not
resist the opportunity to deal with the status of international treaties in Australian
administrative law, despite the fact that no treaty obligation was in dispute. In
their joint judgment, McHugh and Gummow JJ reinforced the accepted view
that ratification of an international treaty does not ‘confer rights or impose liabil-
ities’ upon citizens.61 They acknowledged that Australia’s international obliga-
tions may impact on statutory interpretation, but otherwise their Honours had a
restricted view of the relationship between international law and domestic law.
For example, they signalled a retreat from the majority judgment in Teoh stating
that it was a ‘curiosity’ that although treaties were not to be accorded the status
of relevant considerations, Teoh would result in international obligations being
‘mandatory relevant considerations for that species of judicial review concerned
with procedural fairness’.62
Hayne J confined himself to suggesting that the consequences of the Teoh
principle with respect to the ratification of international treaties may need to
be reconsidered in the future.63 Callinan J prefaced his discussion of the role of
treaties in administrative law by stating that the applicant’s assertion there was
no need to rely upon Teoh was ‘not entirely’ convincing.64 His Honour indicated
his distaste for the majority’s treatment of international conventions in adminis-
trative law in Teoh by referring to the ‘elevation of an Executive ratification of an
un-enacted Convention to almost the level of a concrete legal right or at least a
springboard therefor’.65 Callinan J’s comments in Lam are of little surprise given
his suggestion in the earlier case of Sanders v Snell that the impact of Teoh may be
limited to matters where ‘any civilised person would hold expectations, whether
referable to a United Nations Convention or otherwise’.66 This comment is in
accordance with Gaudron J’s suggestion in Teoh that the status of the children as
Australian citizens was more significant than ratification of the CRC.
Such observations indicate that the prospect for future developments in the
relationship between international treaties and Australian administrative law
may be limited. While the comments were obiter, as is suggested by Ruddle and
Nicholes, the ‘decision in Teoh remains good law at present, albeit with an uncer-
tain future’.67 More broadly, the discussion of Teoh in Lam highlights that (with
a few exceptions) the Australian judiciary displays a number of ‘anxieties about
international law’.68 Charlesworth, Chiam, Hovell and Williams have listed these
anxieties as ‘the preservation of the separation of powers . . . the fear of opening the
floodgates to litigation; the sense that the use of international norms will cause
instability in the Australian legal system; and the idea that international law is
essentially un-Australian’.69 To some extent these anxieties are understandable –
if international treaties are given too prominent a place in domestic case law then
the judiciary will be accused of ‘backdoor’ incorporation.70
In Lam, the first anxiety listed by the four commentators was manifested in the
judgments of McHugh and Gummow JJ and Callinan J through their discussion
of the proper roles of the Executive, the Parliament and the courts. McHugh
and Gummow JJ pointed out that Teoh involved a treaty that had been ratified

by the executive, but had not been separately implemented by the parliament
pursuant to the external affairs power in s51(29) of the Constitution.71 Their
Honours stated that it was the task of the judiciary ‘to declare and enforce the
limits of the power conferred by statute upon administrative decision makers,
but not, by reference to the conduct of external affairs, to supplement the criteria
for the exercise of that power’.72 Callinan J indicated that the High Court in
Teoh had elevated ‘the Executive above the parliament’ in giving effect to the
CRC.73 By invoking the separation of powers, these judges have suggested that
the majority in Teoh interfered with the constitutional arrangements with respect
to international obligations.
Prior to the High Court’s decision in Lam, concern had been expressed in other
circles about the respective roles of Parliament and the Executive in relation to
Australia’s involvement in international treaties. In 1995, the Senate Legal and
Constitutional References Committee tabled its report on the treaty-making pro-
cess in Australia, Trick or Treaty? Commonwealth Power to Make and Implement
Treaties. A number of witnesses to the inquiry made it clear that the decision in
Teoh was an important consideration in their belief that parliamentary partic-
ipation in the treaty process should be increased.74 The Committee concluded
that the High Court’s decision confirmed the influence of treaties in Australian
domestic law which, in turn, indicated the need for increased parliamentary
involvement prior to ratification.75 This Report resulted in a number of reforms,
including the establishment of the Joint Standing Committee on Treaties. This
Committee was established to examine proposed treaty actions that have been
tabled in the Commonwealth Parliament and to report on whether Australia
should undertake steps with respect to treaties that would bind Australia in inter-
national law, including whether it should ratify new treaties. Viewed in the light
of this inquiry, the decision in Teoh has played a part outside the administrative
law context and has been utilised in debates concerning the roles of the three
branches of government in entering into and implementing Australia’s interna-
tional obligations.

Implications for procedural fairness and the role
of legitimate expectations

The above discussion demonstrates that Lam has foreshadowed a retreat by the
High Court from further developments in the utilisation of Australia’s ratification
of international treaties as a means of illuminating the statutory discretions of
administrative decision makers. Moving away from the impact of international
treaties on administrative law to deal with the principles of procedural fairness,
there are also indications in the judgments that the concept of legitimate expec-
tations, and the notion of ‘unfairness’ for the purposes of procedural fairness as
a ground of review, may have been altered as a result of the decision. In Kioa,
Mason J held that:76

. . . it is a fundamental rule of the common law doctrine of natural justice expressed
in traditional terms that, generally speaking, when an order is to be made which will
deprive a person of some right or interest or the legitimate expectation of a benefit, he is
entitled to know the case sought to be made against him and to be given an opportunity
of replying to it.

Mason J’s formulation of the principle of procedural fairness in Kioa has been
preferred over Brennan J’s reliance on the construction of the statutory power
under review.77 The High Court decisions in Teoh and Lam give different emphases
to the role of legitimate expectations in procedural fairness, with the judges in
Lam ultimately preferring a broader use of the concept of unfairness. This section
will examine the discussion of legitimate expectations and unfairness in Teoh and
Lam and the way in which the Lam decision has been applied in subsequent High
Court and Federal Court cases.

The concept of legitimate expectations
In administrative law, the concept of a legitimate expectation has been employed
for a number of different purposes.78 First, it has been used to impose a duty to
observe procedural fairness. Secondly, it may clarify the content of the duty, that
is, whether notice is required or a hearing will be imposed in the circumstances
of a particular case.79 This is illustrated by Gaudron J’s comment in Haoucher v
Minister for Immigration and Ethnic Affairs,80 where her Honour stated that:

The notion of legitimate expectations is one to which resort may be had at two distinct
stages of an inquiry as to whether there has been a breach of the rules of natural
justice. It may serve to reveal whether the subject matter of the decision is such that
the decision-making process is attended with a requirement that the person affected
be given an opportunity to put his or her case . . . . On the other hand, it may serve
to reveal what, by way of natural justice or procedural fairness, was required in the
circumstances of the particular case.

In Teoh, there was no dispute that the applicant was entitled to procedural
fairness.81 The majority of the High Court held that Australia’s ratification of
the CRC gave the applicant a legitimate expectation that administrative decision
makers would make the ‘best interests of the child’ a ‘primary consideration’, as
required by Article 3(1). Thus, it was a decision which resulted in the content of
procedural fairness being increased for the particular applicant. Mason CJ and
Deane J were careful to point out that the legitimate expectation did not require
the decision maker to act in a particular way as that would be ‘tantamount to
treating it as a rule of law’.82 Procedural fairness only required that if a deci-
sion maker ‘proposes to make a decision inconsistent with a legitimate expecta-
tion . . . the person affected should be given notice and an adequate opportunity
of presenting a case against the taking of such a course’.83
While the majority of the Court in Teoh drew upon the language of legitimate
expectations to find that the applicant had been denied procedural fairness,

McHugh J cast doubt on the continuing applicability of the concept, questioning
‘whether the doctrine of legitimate expectations still has a useful role to play’.84
In his view, since the rules of procedural fairness are ‘presumptively applicable
to administrative and similar decisions made by public tribunals and officials’,
there is no need for legitimate expectations.85 This is not the first time that a
High Court judge has expressed dissatisfaction with the concept of legitimate
expectations. In Kioa,86 Brennan J questioned the utility of legitimate expecta-
tions in determining the threshold question as to whether procedural fairness
should be implied into a decision-making process. His Honour described legit-
imate expectations as a ‘seed’ which had been planted by Lord Denning MR in
Schmidt v Secretary of State for Home Affairs87 and had ‘grown luxuriantly in
the literature of administrative law’.88 Brennan J preferred the term ‘interests’
as being a more accurate descriptor than legitimate expectations.89 A number
of writers agree with this position, arguing that the notion of an interest in the
threshold test is broad enough to encompass matters that fall within the ambit of
procedural fairness.90 In Lam, McHugh J found greater support for his desire to
limit or abandon the concept of legitimate expectations altogether amongst other
members of the Court. For example, Callinan J described the phrase ‘legitimate
expectation’ as ‘unfortunate’ and the concept as a ‘fiction’.91 Hayne J suggested
that the use of legitimate expectations in Teoh ‘raised more questions than it
answers’.92 In their joint judgment, McHugh and Gummow JJ explicitly confined
legitimate expectations to the content of procedural fairness.93 The judgments
evidence two concerns about the development of legitimate expectations. The
first relates to the question whether an expectation has to be held by a particu-
lar individual in order for it to be regarded as ‘legitimate’. The second concern
draws upon the merits/legality distinction in administrative law and deals with
the issue as to whether a legitimate expectation can give rise to substantive as
well as procedural benefits. Each of these issues will be considered in turn.

Subjective versus objective expectations
In Australian administrative law, it is accepted that in applying the concept of
legitimate expectations it is not necessary to demonstrate that an individual
actually held an expectation, rather it is enough to show that it was reasonable
in the circumstances. In Teoh, Mason CJ and Deane J stated that an applicant
need not personally entertain the expectation, ‘it is enough that the expectation
is reasonable in the sense that there are adequate materials to support it’.94 In
Haoucher, McHugh J stated that a legitimate expectation could be founded on
a detailed policy statement made by the responsible minister in the House of
Representatives as to the considerations that would govern the exercise of a
statutory power (in that case, the power to deport).95 Such a legitimate expecta-
tion was not based on a subjective belief, but rather on the reasonableness of the
expectation. In Lam, McHugh and Gummow JJ contrasted a legitimate expecta-
tion arising from ratification of a convention to the policy statement in Haoucher.
In their view, Teoh could not stand beside Haoucher.96 While their Honours did

not require that a person must actually hold an expectation, they suggested that
ratification of a convention was beyond the range of expectations that could be
regarded as legitimate. In their view, in determining the circumstances in which a
legitimate expectation may arise it was wrong ‘to treat the question of the extent
to which such matters impinge upon the popular consciousness as beside the
point’.97 On the facts, Gleeson CJ agreed that Lam did not possess a subjective
expectation as a result of the letter dated 7 November 2000 and he was unable to
accept that ‘it would have been reasonable to expect the Department to write to
the applicant if for any reason there was a change of plan about contacting Ms
Callinan J was scathing of the idea that a legitimate expectation could arise
(as was the case in Teoh) where an individual has no knowledge of the existence
of an international treaty, and therefore had no actual expectation ‘legitimate or
otherwise’.99 In Callinan J’s view, if legitimate expectations were to remain part
of Australian law then they should be confined to situations where ‘there is an
actual expectation’ or where ‘a reasonable inference is available that the party
turned his or her mind consciously to the matter in circumstances only in which
that person was likely to have done so, he or she would reasonably have believed
and expected that certain procedures would be followed’.100 Although Callinan
J was critical of the idea that a legitimate expectation may arise where it is not
held by an individual, he did not, nor did any other member of the High Court,
rule that an expectation must be subjectively held. However, it does appear that if
legitimate expectations are to survive post-Lam then the range of circumstances
in which it may be said that they are reasonably held will be limited. The extent
to which this has occurred in practice will be examined in the context of the
post-Lam jurisprudence. As is suggested by Dyer, ‘if it happens that we use the
term [legitimate expectations] less, it might well be because the courts give more
attention to the different kinds of expectations and their significance for fair

Substantive versus procedural outcomes
Australian courts have continually reinforced the principle that a legitimate
expectation will only result in procedural rather than substantive outcomes for an
applicant. In Attorney-General (NSW) v Quin,102 Mason CJ expressed the position
as follows:

In the cases in this Court in which a legitimate expectation has been held entitled to
protection, protection has taken the form of procedural protection, by insisting that
the decision maker apply the rules of natural justice. In none of the cases was the
individual held to be entitled to substantive protection in the form of an order requiring
the decision maker to exercise his or her discretion in a particular way.

In accordance with this reasoning, in Teoh the Court refused to hold that
the administrator must comply with the provisions of the CRC. The Australian
focus on procedure rather than on substance can be contrasted to the position in

England where the courts have extended the concept of legitimate expectations
to deliver substantive outcomes and not just procedural benefits. In R v North and
East Devon Health Authority; Ex parte Coughlan,103 the English Court of Appeal
held that an assurance given by a Health Authority that a nursing home would
remain the home for life of a severely disabled patient, Mrs Coughlan, was binding
in the absence of an overriding public interest. Although counsel for the appli-
cant in Lam did not argue for substantive protection on the basis of Coughlan,
the English Court of Appeal’s decision was given prominence in the judgments of
Gleeson CJ and McHugh and Gummow JJ. Gleeson CJ suggested that the appli-
cant’s arguments in Lam came ‘very near’ to suggesting a substantive outcome
rather than a procedural benefit.104 In considering Coughlan, McHugh and Gum-
mow JJ explicitly rejected the operation of substantive protection in Australia on
the basis that it would impinge on the merits of the case and thus be beyond the
proper constitutional role of the judiciary in determining proceedings for judi-
cial review of administrative action.105 These statements echo comments made
by Brennan J in Kioa where he emphasised that procedural fairness is not con-
cerned with the merits of the case, but with the procedure that is to be observed
in the exercise of a power.106 While the comments do not add anything new to an
understanding of the principles of procedural fairness in Australia, they indicate
that many judges in the High Court are dissatisfied with the existing concept
of legitimate expectations and are looking for ways to diminish its significance.
In this context, Burmester believes that the Court in Lam ‘was sending a strong
message that in Australia there were not to be any substantive rights associated
with the doctrine’.107

The meaning of unfairness
It would appear that that one way in which the High Court in Lam sought to
downplay the place of legitimate expectations was to give greater weight to the
meaning of ‘unfairness’. Gleeson CJ stated that ‘what must be demonstrated is
unfairness, not merely a departure from a representation . . . The ultimate ques-
tion remains whether there has been unfairness, not whether an expectation has
been disappointed’.108 For Gleeson CJ, the essential question was one of ‘prac-
tical injustice’.109 Hayne J also suggested that if the procedures were fair, then
‘reference to expectations, legitimate or not, is unhelpful, even distracting’.110
Questions of fairness generally go towards determining the content of procedu-
ral fairness – that is, the procedural steps that are required of the decision maker
in a particular case.111 Flexibility appears to be a key factor, with Brennan J in
Kioa describing the principles of procedural fairness as having ‘a flexible quality
which, chameleon-like, evokes a different response from the repository of statu-
tory power according to the circumstances in which the repository is to exercise
the power’.112 In the same case, Gibbs CJ stated that the ‘rules of natural justice
are flexible, requiring fairness in all the circumstances’.113 Mason J also equated
flexibility with fairness when he held that ‘the statutory power must be exercised

fairly’. In Mason J’s view, this question was to be determined ‘in accordance with
procedures that are fair to the individual considered in the light of the statutory
requirements’.114 The crucial factor for the High Court in Lam was the inability
of the applicant to demonstrate that he had been deprived of an opportunity to
put material to the Department, or that there was additional material that would
have influenced the Department to decide the case differently.115 In determining
whether there had been unfairness, some members of the High Court employed
a concept that bordered on the private law notion of detrimental reliance.116 For
example, Gleeson CJ found that the applicant had not suffered any detriment as
the Department had not made a statement of intention that had resulted in Lam
failing to put material before the Department.117 McHugh and Gummow JJ also
accepted that there was no evidence to support a conclusion that the applicant
had relied upon the letter of 7 November 2000 and consequently failed to submit
information to the Department on the best interests of the children.118
Lacey has criticised the use of detrimental reliance in Lam as an improper
transposition of private law concepts to the public law arena. In her view, it
draws the courts into a consideration of the fairness of the outcome to a particular
applicant.119 This brings the Court to a consideration of the merits, a matter which
is outside the purview of judicial review. Lacey argues that an approach which
concentrates on the outcome may also encourage poor administrative decision-
making as it suggests that a determination of whether a procedure is unfair
will depend on whether there was any detriment to a particular individual.120
Dyer acknowledges that some members of the High Court used the language of
detrimental reliance, but believes that the Court concentrated on the question
whether the breach affected the applicant’s right to be heard, rather than whether
the breach affected the outcome of the applicant’s case.121

Unfairness and ‘practical injustice’ post-Lam
Lacey’s concern that the courts will be drawn into considering the merits of a
case does not appear to have been borne out by subsequent cases in the High
Court and the Full Court of the Federal Court. Since Lam, the High Court has
considered the concept of unfairness in Applicant NAFF of 2002 v Minister for
Immigration and Multicultural and Indigenous Affairs,122 a case in which the Court
unanimously declared that there had been a breach of procedural fairness. The
applicant in Applicant NAFF of 2002 was denied a protection visa and applied for
review in the Refugee Review Tribunal (RRT). At the close of the hearing the
RRT stated that it would write to the applicant to clarify inconsistencies in his
evidence and to enable him to place further information before the RRT. Contrary
to this statement, the RRT did not write to the applicant and instead published
its decision in which it rejected the applicant’s claim. In its reasons, the RRT
stated that it did not find the applicant a credible witness. As is noted in the joint
judgment of McHugh, Gummow, Callinan and Heydon JJ, the RRT member’s
concluding remarks indicate that she believed that the applicant’s case would be

assisted by further explanation.123 The High Court held that the deprival of the
opportunity to answer further questions was a breach of procedural fairness.124
The joint judgment paid careful attention to the words of the RRT member and
the inference that could be drawn from these words that fairness required further
steps to be taken.125
In his separate judgment, Kirby J referred to Gleeson CJ’s remarks in Lam where
he indicated that unfairness would easily be demonstrated where ‘a decision
maker informs a person affected that he or she will hear further argument upon a
certain point, and then delivers a decision without doing so’.126 Applicant NAFF
of 2002 was just such a case. Kirby J accepted that the applicant not only had to
prove that the representation was made, but that unfairness resulted.127 However,
he denied that an applicant ‘must always prove . . . what the party would have
done had the procedural defect not occurred’.128 For Kirby J ‘it was sufficient for
the applicant . . . to establish that a procedural default had occurred which was
not immaterial but might have affected the outcome of the proceedings’.129 Such
a conclusion obviated the need for the Court to consider the merits of the case.130
The Federal Court has also had occasion to consider Gleeson CJ’s notion of
‘practical injustice’. The Full Court of the Federal Court in WACO v Minister for
Immigration and Multicultural Affairs131 stated that Gleeson CJ’s comments in
Lam should not be read as requiring an applicant to demonstrate to the court that
there would have been a different outcome. For example, it would be ‘unfair’ for
a decision maker to make a finding about an issue without giving an individual
notice that the issue is the subject of dispute and without giving the persona
adversely affected an opportunity to be heard.132 If the ‘possibility exists’ that
an applicant might be able to call evidence that ‘could affect the outcome’, it
is not necessary for the applicant to prove that the evidence ‘would affect the
outcome’.133 In WACO the applicant contended (amongst other claims) that the
RRT had denied him procedural fairness when it failed to indicate that it did not
believe that two supporting letters he produced were authentic and therefore did
not give him an opportunity to call evidence to buttress his claim that the letters
were genuine.134 The Full Court accepted that this was ‘unfair’.135
In Minister for Immigration and Multicultural and Indigenous Affairs v SZFDJ,136
a refugee applicant argued that he had been denied the opportunity to comment
on information relied upon by the RRT in deciding against his refugee application.
The Full Court of the Federal Court purported to distinguish Lam in stating that ‘in
the present case, unlike that in [Lam], the demonstration of a positive unfairness
flowing from the accepted denial of natural justice does not require affirmative
evidence that, but for the denial, the visa applicant would have taken, or refrained
from taking, a particular course’.137 What is interesting about this statement is that
the Full Court appeared to assume that Lam was authority for the proposition that
‘positive unfairness’ needed to be demonstrated. The impact of this interpretation
is somewhat mitigated by the Full Court’s conclusion in SZFDJ that it was ‘not
necessary to conclude affirmatively’ that the RRT would have arrived at a different
result but for the denial of procedural fairness.138 Together with Applicant NAFF

of 2002, these two decisions demonstrate that in applying Lam the courts have
not required an applicant to show that the result would have been different, only,
in the words of the Full Court of the Federal Court, that it ‘could’ have been
A single judge of the Federal Court has also considered the concept of practical
injustice in a case involving a dumping notice issued pursuant to the Customs Act
1901 (Cth). In Expo-Trade Pty Ltd v Minister for Justice and Customs the applicant
contended that the method of calculation adopted for the non-injurious price
in a report by the Customs Department alleging that the applicant was liable
for dumping duty on goods imported into Australia did not match the methods
indicated in a Customs Manual.139 Expo-Trade Pty Ltd argued that the failure to
tell it that there would be a departure from the policy in the Manual constituted
a denial of procedural fairness as it was unable to make submissions on the
method of calculation.140 Citing Gleeson CJ’s decision in Lam, Moore J found
that no ‘practical injustice’ could be demonstrated by the corporation as:141

There was no evidence to suggest that Expo-Trade acted on the basis that what was said
in the Manual would be given effect to and refrained from advancing material (which
it can now point to) or making submissions because of an erroneous assumption about
how Customs would go about considering relevant matters.

This statement appears to give great weight to the place of detrimental reliance
in determining whether procedural fairness has been breached. Dyer has high-
lighted that Lam does not go as far as Moore J indicates, particularly as McHugh
and Gummow JJ’s comments demonstrate that they agree with the way in which
Haoucher dealt with an announced policy.142 However, it does reinforce the con-
clusion that where applicants cannot indicate to a court that there is any addi-
tional material that could result in a different decision, or that a procedure has
been fundamentally ‘unfair’, the concept of legitimate expectations will not add
to their case. This view is supported by the majority’s decision in Le v Minister for
Immigration and Multicultural and Indigenous Affairs, where it was held that ‘a
mere failure to follow the stated procedure does not, of itself, amount to a denial
of procedural unfairness . . . [t]he applicant has not established that there was
any practical unfairness’.143

In the 2006 case of Royal Women’s Hospital v Medical Practitioners Board of Vic-
toria, Maxwell P of the Court of Appeal of the Victorian Supreme Court stated

Teoh might well be decided differently by the present High Court. But the occasion for
that re-examination of Teoh has not yet arrived, and the legitimate expectation test
continues to be applied in the courts.
In any case, the question of legitimate expectation represents only one part of what
was said in Teoh. The other propositions to which I have referred about the relevance
of international human rights law are still good law, and continue to be applied.

In that case, Maxwell P relied upon Teoh as support for a number of propositions
concerning the relationship between international law and Australian law.145
Thus, despite the doubt cast on the majority’s decision in Teoh by the High Court
in Lam, the case is still being cited with approval in a number of contexts. In
particular, the fact that the Departmental officer in Lam referred to Article 3
of the CRC in a letter to the applicant, and no judge doubted the relevance of
the children’s best interests, indicates the extent to which the case has perme-
ated the decision-making process. Maxwell P’s statement also suggests that the
doctrine of legitimate expectations is pertinent post-Lam. Commenting on the
decision in Lam, French has written that ‘notwithstanding the doubts expressed
by some members of the High Court, the doctrine of legitimate expectations
remains practical and relevant’.146 But the continuing relevance of legitimate
expectations must be tempered by the High Court’s reluctance to contemplate
further developments in the doctrine and its preference for taking an approach
which pays closer examination to the concept of unfairness. It is yet to be seen
whether a legitimate expectation based on Australia’s ratification of an interna-
tional convention (apart from the CRC) will be successfully argued in the future,
or whether some other type of executive action will be taken to found a legitimate
expectation. Dyer best sums up the current position by stating that ‘it appears that
a majority of the High Court is now unlikely to accept loose usage of the concept of
‘legitimate expectation’. Rather, the court is likely to expect explanation as to the
nature, basis, doctrinal justification and affect of alleged expectations’.147 These
comments are supported by the courts’ subsequent reliance on the concepts of
‘unfairness’ and ‘practical injustice’ in applying Lam.
Standing back from the details of the status of international law in Australian
administrative law and the doctrine of legitimate expectations, one of the most
interesting aspects of the decision in Lam is the underlying concern displayed
by the judges for the separation of powers and the proper functions of the three
branches of government. This is evidenced in their treatment of two issues. First,
the judges’ obiter discussion of the role of international treaties in generating
a legitimate expectation is explicitly based on the need to ensure that unincor-
porated treaties, ratified by the executive but not separately implemented by
parliament, are not given an ‘elevated’ status in domestic law. The judges’ fun-
damental concern appears to be that the Teoh decision could interfere with the
proper roles of the parliament and the executive with respect to international
obligations. Secondly, the invocation of the Constitution and the description of
the proper role of judicial as distinct from merits review in the discussion of
substantive expectations also evidences a desire to ensure that the appropriate
sphere of the judiciary is preserved and the merits/legality distinction is upheld.
The extent to which the Court dealt with both issues, despite the fact that neither
point was in contention, is an indication of the depth of feeling regarding the
need to maintain the proper constitutional functions of the executive, the parlia-
ment and the judiciary.
The rule against bias
Matthew Groves

Callinan and Heydon JJ recently observed that ‘unfairness can spring not only
from a denial of an opportunity to present a case, but from denial of an opportunity
to consider it’. They explained that the failure to fairly consider a case could arise
‘not only from obstruction . . . of its presentation but also from self-disablement by
the [decision maker] from giving consideration to that presentation by permitting
bias to affect its mind’.1 Their Honours were right to suggest that the two pillars
of natural justice – the hearing rule and the rule against bias – each emanate from
notions of fairness, but each fosters fairness in different ways. The hearing rule
governs the right to know and be heard. It is the source of principles governing
the information that should be provided to a person who may be affected by a
potential decision, when the information should be provided and how the person
should be able to put his or her views to the decision maker. The bias rule provides
an important complementary right, which is to ensure that the decision maker
to whom the hearing rule applies is impartial and can approach a decision with
an open mind.
This chapter examines the test governing allegations of bias and the apparently
objective standard by which a court asks whether a well-informed observer would
reasonably apprehend that a decision maker might not bring an impartial mind
to a decision. It also considers the three exceptions to the bias rule – necessity,
waiver and statutory modification. But first it is useful to explain the distinction
between actual and apprehended bias and also the basis upon which the rule
against bias rests.

Actual and apprehended bias
There are two forms of bias – actual and apprehended bias. A claim of actual
bias requires proof that the decision maker approached the issues with a closed


mind or had prejudged the matter and could not be swayed by the evidence. An
allegation of actual bias essentially invites a court to make an adverse finding
against a decision maker in a personal sense because it involves a claim that the
decision maker was biased or had prejudged the matter. The courts are naturally
reluctant to make findings of this nature and subject a claim of actual bias to a
high, arguably almost impossible, standard of proof.2 It is not enough to establish
that the decision maker held preliminary or tentative views of a strong nature.
Short of an admission of guilt from the decision maker, or, more likely, an open
and public expression of bias, this standard is difficult to satisfy.3 A claim of
apprehended bias is much easier to establish because it requires a finding that
a fair minded and reasonably well informed observer would conclude that the
decision maker did not approach the issue with an open mind.4
Several consequences attach to the different requirements for actual and
apprehended bias. Actual bias is assessed on a subjective basis, while appre-
hended bias is assessed objectively. A court that upholds a claim of apprehended
bias is not required to make an adverse finding against the decision maker. It can
make the more palatable finding that a reasonable observer, but not necessarily
the court, might conclude that the decision maker was not impartial and go no
further. There are many instances in which courts have stressed that a claim of
apprehended bias is not upheld lightly,5 but there is no doubt that the standard
applicable to apprehended bias is not nearly so strict as that which applies to
actual bias. Although a successful claim of either form of bias is sufficient to
set aside a decision, there is little reason for a party to attempt the more onerous
and confronting requirements that are raised in a claim of actual bias.6 Kirby J
acknowledged this when he conceded that a party who sought to raise a claim
of actual rather than apprehended bias would be ‘foolish . . . to assume a heav-
ier obligation when proof of bias from the perceptions of reasonable observers
would suffice to obtain relief’.7 This admission highlights the overlap between the
two forms of bias. A case that might support a claim of actual bias may often be
argued upon the ground of apprehended bias simply because the latter is easier
to establish.8

What values underpin the rule against bias?

The core principles of the rule against bias are fairness and impartiality. The prin-
ciple of fairness requires that the parties be treated in an even-handed manner.
An important aspect of fairness in this sense is that each party be allowed to par-
ticipate in the decision-making process in a manner that does not grant an unfair
advantage to the other.9 The principle of impartiality requires a decision maker to
approach a matter without predispositions of a character or strength that prevent
the decision maker from reaching a decision contrary to those predispositions
in an appropriate case. On this view, the principle of impartiality is not abso-
lute. It requires a decision maker to be open to persuasion but not necessarily

to have a blank mind or be devoid of all preconceptions or other possible
There is a related issue that underpins the rule against bias, at least as it
applies to judges, which emanates from the constitutional requirements that
attend the position of judicial officers in our system of government. Some judges
have suggested that the requirements of Chapter III of the federal Constitution
might provide a constitutional basis for a requirement of impartiality on the part
of judges.11 Although this view has not yet gained strong currency in the High
Court, it could be argued to represent a natural connection between the common
law (from which the rule against bias developed) and the constitution (from
which the requirements of the separation of powers have developed). If the rule
against bias was accepted to be ultimately founded in Chapter III of the Constitu-
tion it could not be abrogated by legislation, at least in its application to federal
Any constitutional foundation for the rule against bias would be quite limited
in its application. It would not extend to federal administrative officials, who exer-
cise power under Chapter II of the Constitution and are, therefore, not subject
to the requirements of impartiality that attend the exercise of powers by judges
acting under Chapter III. If the rule against bias was accepted to be a requirement
of Chapter III of the Constitution it might extend to state judges by virtue of the
principle established in Kable’s case.13 That principle prohibits state parliaments
from investing in state courts non-judicial powers of a kind which are incompati-
ble with the exercise by them of the judicial powers of the Commonwealth which
may be invested in them pursuant to s77 of the Constitution. At present, however,
the rule against bias is clearly not a requirement to which this principle might
It should be noted that the requirement of judicial impartiality may also be
founded in the common law.15 It is clear that the rule against bias operates to
preserve the administration of justice by removing the influence of factors that
might distract a judge from performing his or her task in an objective manner.16
On this view, the rule helps to maintain actual impartiality of judges but it also
bolsters their perceived impartiality by assuring observers that judges remain
free of possible influence.17

The demise of the rule of automatic
disqualification – the rise of a context sensitive
rule against bias

The rule against bias was long dominated by the principle of automatic disqualifi-
cation which can be traced to Dimes v Grand Canal Junction.18 That case involved a
long running dispute that ended in favour of the Canal Company. It later emerged
that the Lord Chancellor of England, who had presided over this decision, held a
substantial holding in the Canal Company that would have been rendered almost

worthless if the Canal Company had lost. The decision of the House of Lords to
overturn the ruling of the Lord Chancellor was long regarded to have established
a rule of automatic disqualification for pecuniary interest.19
The decision of the House of Lords in Pinochet (No 2)20 provides a useful
starting point to trace the demise of the automatic disqualification rule even
though that case confirmed and extended the reach of the principle in England.
The Pinochet cases involved extradition proceedings against the former military
ruler of Chile. In Pinochet (No 1),21 a narrow majority of the House of Lords held
Pinochet was eligible for surrender, but this finding was challenged in Pinochet
(No 2)22 on the basis that a member of the majority of the first case (Lord Hoff-
mann) should be disqualified because of his association with Amnesty Interna-
tional. Amnesty had long lobbied for the extradition and trial of Pinochet and
was granted leave to intervene in the first case. Lord Hoffmann was an unpaid
director of the charitable arm of Amnesty which was established to raise funds
for Amnesty. He had no pecuniary interest in the proceeding and no direct legal
relationship to Amnesty, but the House of Lords held that his connections to
Amnesty required his disqualification.
The House of Lords could have rested its decision on a narrower basis and
confined its reasoning to the unusual coalescence of factors in the case at hand,
but it adopted the more general view that the principle of automatic disqual-
ification should not be limited to pecuniary interests. The Lords held that the
rationale of automatic disqualification, which required that no person should be
the judge or his or her own cause, applied equally to the promotion of a cause.23
Although the House of Lords did not clearly explain which interests might attract
this extended application of the rule of automatic disqualification, the general
tenor in which this principle was applied to the interest of Lord Hoffmann sug-
gests that the strength of the interest and its relevance to the case at hand will be
important. Accordingly, any social cause or similar interest might trigger auto-
matic disqualification if the view is strongly held by the decision maker and clearly
relevant to the case at hand. The key, on this view, is the effect that an interest may
The High Court took quite a different path in Ebner v Official Trustee.24
Ebner involved two co-joined applications, in each of which it was argued that
the holding by a trial judge of shares in a bank that was a party to proceed-
ings before the judge required the judge’s automatic disqualification by rea-
son of pecuniary interest. A majority of the High Court rejected the rule of
automatic disqualification for pecuniary interest, partly because it was said to
have evolved from a mistaken interpretation of earlier authority.25 More impor-
tantly, it also rejected the conceptually broader notion that pecuniary interests
should be the subject of a separate rule governing their effect for the purposes
of bias.26 Gleeson CJ, McHugh, Gummow, Hayne and Callinan JJ adopted a sim-
ilar view to the House of Lords in Pinochet (No 2) by accepting the pecuniary
and other interests should be treated in the same manner. According to their

As a matter of principle, in considering whether circumstances are incompatible with
the appearance of impartiality, there is no reason to limit the concept of interest to finan-
cial interest, and there may be cases where an indirect interest is at least as destructive
of the appearance of impartiality.27

Their Honours also held that the question of whether a pecuniary or other
interest might give rise to a reasonable apprehension of bias should be determined
according to a two-fold test. They explained that as follows:

First, it requires the identification of what is said might lead a judge (or juror) to decide
a case other than on its legal and factual merits. The second step is no less important.
There must be an articulation of the logical connection between the matter and the
feared deviation from the course of deciding the case on its merits. The bare assertion
that a judge (or juror) has an ‘interest’ in litigation, or an interest in a party to it, will be
of no assistance until the nature of the interest, and the asserted connection with the
possibility of departure from impartial decision-making, is articulated. Only then can
the reasonableness of the asserted apprehension of bias be assessed.28

The reasoning in these passages invites several comments. First, the High Court
has clearly adopted a single standard for bias. While the test itself is uniform, its
outcome is not. Disqualification is no longer automatic and instead depends
on the circumstances of each case. Secondly, the requirement to articulate the
connection between the alleged source of bias and the potential departure from
impartiality highlights the need to explain the effect of the interest in question
upon the decision maker. A party must now essentially ‘join the dots’ for a court
before it may be satisfied of an apprehension of bias.29 Thirdly, the requirement
to explain the effect of a source of alleged bias raises the standard of the test,
at least in practical terms. An applicant cannot simply point to an interest, but
must now take the further step of explaining the effect of that interest.30 This
protean test does not distinguish between either decision makers or interests. If
the basic elements of the test remain constant, the main variable is the context
in which an allegation of bias is raised. In the area of bias, therefore, context is
everything.31 The majority in Ebner hinted at this issue when it explained that
the application of the rule against bias to ‘decision makers outside the judicial
system must sometimes recognise and accommodate differences between court
proceedings and other kinds of decision-making’.32
The extent of those differences was subsequently highlighted in the case of
Jia.33 That case concerned the effect of statements by a minister in a talkback
radio program. The minister explained what might lead him to conclude a person
was not of good character for the purposes of migration legislation. Mr Jia had
been convicted of several offences involving the type of behaviour the minister
had discussed. Mr Jia’s case came before the minister, who made an adverse deter-
mination. The immediate question for the High Court was whether the minister’s
statements gave rise to a reasonable apprehension of bias. The deeper question
was the extent to which the minister’s political activities, which required him
to provide media interviews and discuss issues within his ministerial portfolio,

could be taken into account in fashioning the requirements of the rule against
Gleeson CJ and Gummow J, with whom Hayne J agreed, accepted that the
principles applicable to judges should be modified when applied to the minister.34
They reasoned that the decision by Parliament to vest a discretionary power in the
hands of a minister suggested that the power was intended to be exercised in light
of the minister’s many responsibilities, which included accountability to both the
Parliament and the electorate. It was possible that a minister who discharged
these responsibilities would make statements that could offend the rule against
bias, at least if it was applied with the same requirements that extend to judicial
officers. Gleeson CJ and Gummow J were clearly mindful of this possibility. They

As the circumstances of the radio interview demonstrate, the Minister himself can be
drawn into public debate about his powers. He might equally well have been asked
questions about the cases in Parliament. The position of the Minister is substantially
different from that of a judge, or quasi-judicial officers, adjudicating in adversarial
litigation. It would be wrong to apply to his conduct the standards of detachment which
apply to judicial officer or jurors. There is no reason to conclude that the legislature
intended to impose such standards upon the Minister, and every reason to conclude

Callinan J was perhaps even more sensitive to the position of a minister. He
accepted that the minister was essentially required to ‘wear two hats’; one as
a member of the federal executive who was required to engage heavily in the
political process, the other as a decision maker to whom a discretionary power
was entrusted. Callinan J held that the views advocated by the minister in one
capacity could not necessarily be imputed to his thinking in another capacity.36
It is unclear whether his Honour would accept that other decision makers should
have such leeway in their different role. Perhaps the better view is that ministers
represent one extreme of the variable content of the rule against bias. That view
seems to accord with what Callinan J saw as the obvious ‘difference between a
Minister and a judge, and indeed, a Tribunal or member’. According to his Honour,
the consequence of that distinction was that:

The role of none of these is identical with the role of others. And different considerations
requiring the application of different rules in relation to each of them are involved in
a judgment whether one of them is affected by disqualifying bias. The Minister is, it
should be noted, in a different position from a Tribunal . . . 37

The reasoning of the High Court in Jia confirms the context sensitive nature
of the rule against bias. There also appears to be general support within the
High Court to group decision makers into broadly similar categories, such as
tribunal members and ministers and also, one assumes, administrative officials.
But subsequent decisions of the High Court have provided little elaboration on
precisely what the rule against bias might mean for those different categories. It
is notable that the court has not used the several later cases concerning alleged

bias in tribunal decision-making to explain more clearly how those requirements
might apply to tribunal members. In these cases, which are discussed below,
the court has either pointedly declined to explain the standards it might expect
from a tribunal member,38 or confined itself to more general remarks about the
character of inquisitorial tribunal proceedings.39
The Court did, however, provide some further guidance in Hot Holdings v
Creasy.40 McHugh J suggested that while the actual test for bias remained

. . . its content may often be different. What is to be expected of a judge in judicial
proceedings or a decision maker in quasi-judicial proceedings will often be different
from what is expected of a person making a purely administrative decision.41

Gaudron, Gummow and Hayne JJ stressed the particular position occupied by
ministers,42 which impliedly accepts the more general contextual approach that
McHugh J accepted more explicitly. But no member of the majority explained in
detail the standards expected of either the departmental officers who advised
the minister or the minister himself.
The demise of the principle of automatic disqualification has left the law gov-
erning bias in a state of flux. The effect of financial and all other interests is
now determined according to one test. The crucial element of that test is the
requirement to articulate the effect of an interest in a very precise way, namely,
to explain how the alleged interest might cause a decision maker to lose his or her
impartiality. While all interests may be treated in the same manner, decision mak-
ers are not. The High Court has signalled that the rules devised for judicial bias
will be modified in their application to other decision makers but is yet to clearly
articulate the nature or extent of those modifications. That issue clearly awaits
further guidance from the court, but in the meantime it seems that principles
governing judicial officers will continue to provide the main point of reference.

Real likelihood, real danger or reasonable
likelihood of bias?

The qualities of the fair minded observer, by whose judgment the rule against
bias is determined, can only be fully understood by reference to the earlier tests
that were rejected or refined. For some time the test for apprehended bias veered
between one of ‘reasonable suspicion’ or ‘real likelihood’. The House of Lords
discarded the test of ‘real likelihood’ in favour of one of ‘real danger’ in R v
Gough.43 The Lords reasoned that the test of ‘real danger’ rather than ‘likelihood’
would orient the test of bias to ‘possibility rather than probability’.44 It also held
that that test could be applied by the court after consideration of any relevant
evidence, even if not available to the decision maker at the time. The High Court of
Australia retained the ‘real likelihood’ test in Webb v R,45 largely because it enabled
the court to focus on the perception of the public, rather than the court, on the
effect of the alleged source of bias.

The difference between the two approaches was not so much about words as
the position from which bias was assessed. The ‘real danger’ test as applied in
Gough clearly made the question of bias one for the court to decide, according
to the court’s view of the case. The ‘real likelihood’ test devised in Webb – sub-
sequently refined to one of a ‘reasonable apprehension’ as held by a fair minded
lay-observer – also left the question of bias ultimately to the court, but required
that the circumstances of the case be assessed in an objective manner.46 The dif-
ference between the two is smaller than might first appear. It is not a difference
between the view of the court and that of a bystander, but rather one between
the view of the court in a purely subjective sense and view of the court on what is
the view of the bystander (which introduces some level of objective judgment).
The semantic differences between each test should not obscure the point that
the perceptions of an individual judge presiding over a claim of bias will always
be of great importance no matter what test applies.47 Kirby J attempted to strike
a balance in Johnson v Johnson48 when he suggested that a ‘reasonable mem-
ber of the public is neither complacent nor unduly sensitive or suspicious’. The
attraction of this approach is its emphasis placed on the ‘reasonable’ nature of
any assessment made of the facts by the fictional well informed observer. But sub-
sequent cases make clear that most judges will endow the reasonable observer
with all the information placed before a review court, no matter how detailed. In
such circumstances, it is almost impossible to adopt the arms’ length assessment
favoured by Kirby J.
Hot Holdings v Creasy49 is an example. The crucial question in that case was
whether a minister’s decision to grant a mining licence to Hot Holdings was influ-
enced by the interest that two administrative officers held in that company. One
officer held shares in the Creasy company, while another had a close relative
with a large holding in the company. The minister was unaware of these inter-
ests but it was argued that his decision was essentially infected by the interests
of the two officers. A majority of the High Court rejected the claim because the
role of the officers in the decision-making process was too ‘peripheral’ to have
affected the minister’s decision. Kirby J dissented strongly, principally because
he disagreed with the majority on whether the conduct complained of could
provide the basis of a reasonable apprehension of bias. Kirby J thought that the
majority drew fine distinctions on the facts before the court which were anti-
thetical to the objective nature of the test for bias.50 He complained that the
reasonable observer upon which the bias test was based would have:

. . . neither the time nor the inclination to evaluate detailed evidence and protestations as
have been made in this case. He or she, as a lay-person, simply sees a ministerial minute
in which two senior departmental officials participated without declaring personal or
familial interests known to each other.51

This criticism highlights a paradox that arises in any judicial assessment of
what a fair minded and reasonably well informed observer might conclude in
any given case. In an application for judicial review the parties normally pro-
vide considerable information about the particular circumstances of the case,

which appears consistent with the touchstone of the ‘reasonably well informed
observer’. But the objective position of the reasonably well informed observer
diminishes as the information attributed to this observer (or the appellate judges
who step into the shoes of the observer) increases.52 The more that the observer
is deemed to know about the circumstances of the decision and the knowledge
held by the decision maker at the time of deciding, the more likely it is that the
observer’s reasoning will mirror that of the decision maker rather than a per-
son placed at an appreciable (and more objective) distance from the case. The
preparedness of courts to receive detailed evidence of the actual knowledge of
decision makers means that the ‘reasonably well informed observer’ may in fact
become a ‘reasonably well informed decision maker acting with the benefit of
hindsight and much more information’. More particularly, the continued will-
ingness of courts to accept very detailed evidence of the circumstances about
a decision claimed to be affected by bias undermines the objective basis upon
which the test to determine this issue is supposed to rest.

What behaviour might give rise to a
complaint of bias?

The facts that might give rise to a claim of bias are clearly open but existing cases
provide a useful list of what might normally support a claim of bias. In Webb v
R,53 Deane J explained that the areas encompassed by the rule against bias fell
into four main categories, which were separate but overlapping:

The first is disqualification by interest, that is to say, cases where some direct or indirect
interest in the proceedings, whether pecuniary or otherwise, gives rise to a reasonable
apprehension of prejudice, partiality or prejudgment. The second is disqualification by
conduct, including published statements . . . The third category is disqualification by
association. It will often overlap the first and consist of cases where the apprehension of
prejudgment or other bias results from some direct or indirect relationship, experience
or contact with a person or persons interested in, or otherwise involved in, the pro-
ceedings. The fourth is disqualification by extraneous information. It will commonly
overlap the third . . . 54

Much attention has focussed on the second and third categories of conduct
and association, particularly after the decision in Pinochet (No 2)55 when parties
became much more inquisitive about the associations and activities of judges.
There was even some suggestion that cynical parties were seeking to ‘judge
shop’, by gathering information about judges who might be unfavourable to a
case and using that information to exclude the judge on the ground of bias. The
English Court of Appeal moved to stamp out this practice in Locobail (UK) Ltd v
Bayfield Properties Ltd,56 by issuing a rigid set of principles governing the factors
commonly invoked in a claim of bias. The Court of Appeal held that the age, gen-
der, national or ethnic origin, religion, sexual orientation, social class or wealth

of a judge could not normally support a claim of bias. It even doubted whether
these factors might ever support a claim of bias. A further category of factors that
would ‘hardly ever’ support a claim of bias included judges’ educational, employ-
ment or social background, or members of social, charitable bodies, or previous
statements made in lectures, articles, textbooks or prior judicial decisions, or
the fact that the judge had previously acted as counsel on behalf of a party. The
factors the court accepted would ‘ordinarily’ support a claim of bias included a
close familial relationship or personal friendship with, or animosity to, someone
closely associated with the case.
Several comments can be made about this apparently clear cut list. One is that
the Court of Appeal was clearly anxious to prevent parties from using the rule
against bias as a means of ‘judge shopping’.57 When viewed from this perspective,
it seems clear that the approach in Locobail is essentially a pragmatic one that
lacks a coherent or guiding principle.58 Despite that caution, the problems that
could flow from any attempt by parties to ‘judge shop’ could also arise in other
forms of decision-making, so the Locobail list might be of wider use even though
the Court of Appeal stressed that its reasoning was confined to judges. Adminis-
trative decision-making by either tribunals or bureaucrats would be undermined
in the same manner as judicial decision-making if the rule against bias could be
deployed too easily.
Although decision of the Court of Appeal in Locobail carried a dogmatic tone,
there must always be exceptions.59 For example, some judges have been dis-
qualified for bias because of their extra-judicial writings,60 and others by their
earlier judicial decisions.61 The decisive factor is usually the strength with which
these earlier statements were made, which suggests that the bias test will not
be breached by a judge who has expressed negative opinions or made adverse
findings but it will be when the judge has done so in an unduly strong fashion.
This focus on the earlier decisions and articles written by decision makers can be
applied readily to judges, whose decisions and articles are freely available to the
public, but one can question how it might be applied to other decision makers.
Most decision makers do not issue detailed reasons for decisions or publish in
the scholarly journals, so it is much more difficult to gain access to the previous
opinions of non-judicial decision makers.
There are other reasons to doubt whether the decisions on judicial bias can and
should be applied to administrative decision-making without some modification.
The judicial mode of decision-making focuses on adversarial decision-making
which is usually conducted in oral proceedings, with evidence gathered and issues
defined almost entirely by the parties and presented in a competitive manner
by the parties in a public hearing. The role of the judge within this contest is
essentially that of an umpire.62 Decisions on the requirements of the rule against
bias in this context can only be understood as reinforcing that impartial and often
passive judicial role.
Administrative proceedings may contain several key differences. One is that
most administrative decision-making does not involve a hearing. The decision

maker will instead simply work in an office, read over material relevant to the
decision, make a decision and communicate that decision by letter to the person
or people directly affected by the decision. Is it important that justice is ‘seen’
to be done when no-one may be looking? The short answer is ‘yes’. The reasons
are both instrumental and non-instrumental.63 The instrumental reason is that
a strict adherence to the rule against bias will ensure that decision makers act
according to the facts before them rather than the extraneous and often irrelevant
factors that the rule against bias can operate to cast aside. A non-instrumental
reason is that the exercise of power according to a rule that promotes fairness
and impartiality fosters values such as respect and fair treatment, and is likely
to enhance respect for and acceptance of administrative action. Another key
difference is that administrative decision-making often involves only one party.
The decision-making process in such instances cannot be reduced to a competition
of the type associated with adversarial litigation. Even if there is a second party,
such as a respondent agency in external review proceedings, the parties are not
necessarily adversaries. Sometimes a respondent decision maker may adopt a
neutral view, while in other times it may adopt a more adversarial one because
it believes the applicant is not entitled to gain a favourable decision.
An important consequence of these various differences is that the nature of
judicial proceedings is such that intervention by, and involvement of, a judge
is not normally required and, when it does occur, is likely to provoke query or
challenge from the parties. It is for this reason that most of the decisions on
judicial bias are instances where judges have moved outside the carefully struc-
tured scheme within which judicial decision-making occurs.64 Procedures for
administrative decision-making are almost always less structured, so the active
involvement of decision makers in procedural issues is more likely. It is also more
desirable because it can provide guidance, clarity and assistance to the parties
when necessary. On this view, the relatively passive decision-making role that
occupies a central role in the doctrine of judicial bias has no place in administra-
tive decision-making.

Exceptions to the rule against bias

The rule against bias is subject to three exceptions. The exceptions of waiver
and statutory modification, which also apply to the hearing rule, have a similar
conceptual basis in their application to the rule against bias, namely that the rule
may be modified or abrogated in whole or in part by either the consent of a party
or by clear legislation to that effect. The rule against bias is also subject to the
further exception of necessity, which enables some decisions that might offend
the rule to be preserved essentially on pragmatic grounds.

A party who becomes aware of facts that could support an objection on the
ground of bias but fails to raise the issue in a timely manner may be found to have

waived the rule against bias.65 This apparently simple principle can give rise
to many practical problems. First, the requirement to make a timely objection
presupposes that a suitable and convenient time for objection will arise. In some
instances it will be difficult to identify the point at which there is sufficient basis
for an objection. A claim of bias might be based on the cumulative effect of
many small issues, such as continued interruptions and comments from a judge,
the effect of which cannot be understood until viewed in hindsight and after
a decision is delivered.66 In addition, a party who objects during the hearing
risks offending the court or having the objection considered before all possible
supporting evidence arises.67
Secondly, an objection is often difficult to make. Many judges have conceded
that even the most skilled advocate can hesitate to pursue a claim of bias.68
A party must explain to the judge why he or she might not be seen as fair and
impartial. The party must, however, remain mindful that the judge may reject the
application and continue to preside. The question is how a party may press a claim
of bias with sufficient strength to explain the claim, while preserving a necessary
level of courtesy with the judge. The inherent tension in these issues presents
a great challenge to experienced advocates and one that most unrepresented
parties would find impossible to manage.
Finally, it should be noted that waiver may be express or implied. Express
waiver normally presents little difficulty. A party who clearly disavows any right
to pursue a question of bias that is raised during a hearing has virtually no prospect
of pursuing the issue on appeal or review. Implied waiver may be more difficult if
there is uncertainty as to whether and when a party had sufficient knowledge of
the facts that might support a claim of bias. It is clear that the consent of a party to
some procedures that might otherwise offend the rule against bias do not neces-
sarily amount to a wholesale waiver of the rule.69 Different considerations arise
if an objection is not pursued for tactical reasons. Such decisions will constitute
an implied waiver of the rule against bias and, according to the rules governing
adversarial litigation, the tactical decision of a lawyer will almost always bind
the client.70

Statutory abrogation or modification
The rule against bias may also be excluded or modified by statute. Legislative
attempts to wholly exclude the rule against bias are rare. The decision of the High
Court in the Epeabaka case71 suggests that such legislation must be expressed in
very clear terms to be effective. That case concerned legislation that sought to
preclude judicial review for a breach of ‘the rules of natural justice’ that occurred
in connection with decision-making, but expressly allowed for review on the
ground that the decision was affected by actual bias. The High Court rejected
the suggestion that the detailed procedures provided evidence of an intention to
exclude the operation of the rule against bias. More particularly, the court did
not accept that an implied intention to exclude or limit review for apprehended
bias could be found when that issue was not clearly addressed. This conclusion

suggests that legislation which purports to exclude judicial review for either
actual or apprehended bias will almost certainly not be taken to have intended
to exclude the other form of bias.
The most common means by which the rule is modified is by legislation that
invests a decision maker with functions which would otherwise offend the rule
against bias if performed by a single decision maker.72 The rule against bias will
yield to the extent required by any such arrangements but may still be invoked
against the use of procedures or statements made by the decision maker which
would offend the rule against bias and are not clearly sanctioned by the Act.
On this view, for example, a court might give effect to legislation that expressly
enabled a decision maker to exercise both prosecutorial and adjudicative func-
tions but it would not allow that officer to make biased statements that were
neither permitted nor required for the proper administration of the statute under
which the officer acted.73

Although necessity has long been recognised as an exception to the rule against
bias the cases do not provide a coherent body of doctrine on this exception. In
Metropolitan Fire Brigade and Emergency Services Board v Churchill,74 Gillard
J explained that a question of necessity was usually decided by reference to a
range of factors, including the nature and degree of the bias, the qualifications
and experience of the decision maker, the conduct of both the parties and the
decision maker, the existence of appeal rights and the public interest. While each
of these factors may be influential, the question for the court is usually a more
general one of whether the circumstances of a case provide some reason for the
court to uphold a decision that would otherwise be invalid by reason of the rule
against bias.
Some cases suggest that necessity will operate to enable a decision maker
to perform its essential functions.75 On this view, the rule against bias cannot
operate so as to frustrate the proper operation of the statute from which the
decision maker draws power.76 The application of this principle is sometimes very
clear. If, for example, there is only one decision maker who may decide a matter
or discharge a function, the necessity exception may be invoked to enable that
decision maker to act even though the circumstances of the case might support a
claim of bias.77 But the exception is much harder to invoke if the decision maker
could have avoided the circumstances that gave rise to a claim of bias78 or another
decision maker may act.79

So can the principles governing the rule against bias be stated with any precision?
Probably not, but that may not be a great drawback. The hearing rule provides
some indication why. Flexibility is the great strength of the hearing rule. It allows
the requirements of fairness to be moulded to the circumstances of each case
and this has fostered the modern growth of the hearing rule. But this flexibility

comes at a price. The hearing rule has long been criticised as uncertain, imprecise,
difficult to predict and a principle that can only be sure when pronounced upon
by court at some later date. Now that the rule of automatic disqualification has
been discarded, the bias rule can be criticised for the same lack of certainty, but
should it be? The rejection of the hard and fast approach of automatic disquali-
fication in favour of a more flexible approach is only the latest stage of the bias
rule. If the evolution of the hearing rule is any guide, the adoption of a more
flexible and context based approach will inevitably lead the courts away from
their focus on the traits of judicial hearings (which are almost always oral and
adversarial) to administrative decision-making (which is almost never oral and
rarely adversarial). The courts will inevitably provide greater guidance on exactly
what they expect of administrative officials, while still preserving a good dose of
flexibility, and that will be no bad thing.
Jurisdictional error without the tears
Mark Aronson

‘Jurisdictional error’ is a term which has puzzled many people.1 Kirby J thinks it
is useless at best, and retrograde at worst:2

In England, the former distinction between jurisdictional and non-jurisdictional error,
once of great significance in cases concerned with the prerogative writs, has now been
abandoned. The precise scope of error classified as ‘jurisdictional’ was always uncertain.
In contemporary Australian law, the boundary between error regarded as ‘jurisdictional’
and error viewed as ‘non-jurisdictional’ is, to say the least, often extremely difficult to
. . . Once it is appreciated that the [constitutional] writs . . . are distinct, are not
confined to their historical provenance, have high constitutional purposes in Australia
and may adapt over time within the limits of their essential characteristics, the old
insistence upon preserving the chimerical distinction between jurisdictional and non-
jurisdictional error of law might be interred, without tears, in Australia as has happened

The term once had a strict meaning, back in the mid-sixteenth century. How-
ever, its vast expansion of scope and meaning over the course of the twentieth
century means that it now conveys no meaning whatsoever as to how it was
reached, but only as to what will be likely to follow, if it has indeed been reached.
Specifically, there are numerous different pathways towards ‘jurisdictional error’,
but usually only one route away from it. In other words, it is both a conclusory
term3 and a point of departure.
There was more to Kirby J’s protest than semantics. His Honour was sug-
gesting that continued use of the term cannot hide the need to develop higher-
level principles of good administration, and this chapter will look also at that


Different contexts, different meanings?

‘Jurisdictional error’ is a term used in several contexts. It might be helpful first
to identify these, because one of the reasons why some people despair at ever
giving it real meaning is their view that the term might carry a different meaning
for each context.
First, there are the remedial contexts. It is trite law that a jurisdictional error
of one sort or another is a prerequisite to obtaining the common law’s prerogative
writs of prohibition and mandamus, or orders in the nature of those writs. The
same remedies are part of the High Court’s entrenched original (as opposed to
appellate) jurisdiction under s75(v) of the Constitution, although in that context
they are called ‘constitutional writs’. As with their common law models, the
constitutional writs also require a showing of jurisdictional error. However, some
members of the court may have been hinting that fewer things might count as
jurisdictional error for the constitutional writs.4 Although jurisdictional error is
usually a prerequisite for the common law’s certiorari, the common law has long
extended that particular remedy beyond jurisdictional error if the challenged
decision displays a non-jurisdictional error of law upon the face of the decision
maker’s record.5
Decisions interpreting privative clauses provide another context. For example,
s474 of the Migration Act 1958 (Cth) tried to oust judicial review of any ‘deci-
sion . . . made . . . under this Act’. However, the High Court interpreted that as not
applying to decisions flawed by a jurisdictional error. This was because jurisdic-
tionally flawed decisions are not ‘decisions made under this Act’. In strict legal
theory, they are not ‘decisions’ but purported decisions.6 ‘Jurisdictional error’ is
therefore a handy way around most privative clauses, whose drafting usually
confines their protection to ‘decisions’, not ‘decisions or purported decisions’.
Some state privative clauses have been stronger than their federal counter-
parts, on the theory that they need observe no constitutional restraints. One state
minister proudly described one such clause7 as ‘one of the most complete priva-
tive clauses on record’,8 because it protects decisions or purported decisions.9
The High Court recently decided three cases involving that clause, all on appeal
from the New South Wales Court of Appeal.10 The majority thought that the
extension of the privative clause’s coverage to purported decisions added noth-
ing to its meaning, because Australia’s typically strong clauses were all drafted
on the assumption that the decision had not been made according to law.11 Per-
haps so, said the Court of Appeal in a subsequent case, adding that the legislative
removal of that particular extension was nevertheless significant in signalling
Parliament’s intention to soften the clause.12 Kirby J thought that a clause pro-
tecting purported decisions would not work if there were ‘fundamental affronts
to jurisdiction’, whilst Heydon J seemed to accept the effectiveness of the priva-
tive clause without qualification. Strictly speaking, the majority view was obiter,
but it does signal that whether future privative clauses be federal or state, their

drafters should shift focus. Instead of taking away remedies or talking of pur-
ported decisions, they should spell out which errors (breach of natural justice,
breaches of specified sections of the Act and so forth) will no longer nullify the
protected decision.
Collateral attack provides yet another context for using the terminology of
jurisdictional error. The reasoning is analogous to the privative clause cases. For
example, because a jurisdictionally flawed surveillance warrant for a listening
device is no warrant in law, there is an inherent weakness in a prosecution case
which is reliant on audiotapes obtained ‘under’ the warrant. The taped evidence
will have been obtained unlawfully, and an accused can raise that issue with the
judge at the criminal trial, without having first to get a judicial review decision
from another judge to quash the warrant.13
‘Jurisdictional error’ is also used in cases discussing when administrators or
tribunals can re-make their decisions. Some powers can be exercised from time to
time, but others can be exercised only once in relation to any specific circumstance
or individual. For example, an appellant before a Migration Review Tribunal gets
a hearing followed by a decision one way or the other. Whether the appellant
has won or lost, the matter is out of the Tribunal’s hands once the decision
is completed. There is no room for second thoughts. Second thoughts on the
merits, that is. But second thoughts will be first thoughts in law if the decision
was invalid. For example, a clerical error by the Tribunal’s registry staff may have
been the reason for the appellant’s non-attendance at the first hearing. If so,
then there was no ‘decision’, because there was a breach of both the common
law requirement of natural justice and the statutory requirements to afford an
opportunity for a hearing. As there was no decision, it was perfectly proper for
the Tribunal to reconvene and conduct a fresh hearing on the footing that the
first hearing could be ignored.14
The Administrative Decisions (Judicial Review) Act 1977 (ADJR Act) and its
state and territory equivalents provide yet another context for ‘jurisdictional
error’. Two of the grounds of review found in the Commonwealth’s ADJR Act
and its territory and state equivalents have been taken as either encompassing
or even just re-stating the ‘jurisdictional error’ idea.15 These are first, ‘that the
person who purported to make the decision did not have jurisdiction to make the
decision’, and secondly, ‘that the decision was not authorized by the enactment
in pursuance of which it was purported to be made’.
Professors Creyke and McMillan appear to treat each of the above usages of
‘jurisdictional error’ as a conceptually distinct context, and suggest that ‘jurisdic-
tional error’ carries different (but overlapping) meanings in each context.16 They
also offer ‘jurisdictional facts’ as another context, although one might well argue
that it is no more ‘separate’ (albeit a good deal more dysfunctional) than any of
the other ways of committing an error that is jurisdictional.
Other contexts can be imagined, but they would be no more productive. For
example, judges are immune from common law liability in tort for things done
in office unless they acted in ‘excess of jurisdiction’. In that context, however,

not even bias or bad faith will amount to such an excess.17 One could engage
here in a semantical debate as to comparisons between ‘excess of jurisdiction’
and ‘jurisdictional error’.18 However, the real reason why we should not fret as to
whether this is yet another context with yet another meaning is that the words are
only labels. The contexts are so different that the semantical similarities should
not be taken seriously. Bias and bad faith are clearly jurisdictional errors in the
context of inquiring whether a challenged decision is a nullity, but it does not
necessarily follow that the same criteria should be used when stating a judge’s
liability to defend tort actions for damages, or to pay damages if those actions
were to succeed.
Believing as they apparently do that jurisdictional error has so many mean-
ings, it is no surprise that Creyke and McMillan suggest that it ‘has become an
overloaded distinction’.19 That would be mild criticism indeed if the term really
bore so many different meanings, but subject to qualifications relating to the
special position of the High Court, this essay begs to differ.
The fact that ‘jurisdictional error’ is conclusory is not necessarily an argument
for abandoning it. Conclusory terms can be useful. They can help us to group
concepts together when we think that they share something in common. In the
case of the many and varied ways of committing ‘jurisdictional errors’, it is argued
that what they share is not so much the quality of the errors, as their legal conse-
quence. In essence, there are many sorts of jurisdictional errors but usually only
one legal consequence, which is that they make the relevant act or decision null
and void. Where nullity is important and where one has been able to establish it
by proving the commission of an error which has nullity as a consequence, there
is no harm and much convenience in characterising that error as jurisdictional.
If jurisdictional error is conclusory, it cannot now be an organising concept
for judicial review, but if nullity is its usual consequence, then that might well
be (or once have been) an organising concept. Nullity’s primary indicia are the
ability to treat decisions as if they had not been made, and to treat conduct as if
it had no legal authority, which explains the basis of collateral attack, and how
the courts outflank most privative clauses.
A word of warning is necessary, however, because it should not be imagined
that nullity is an absolute affair. It is a bundle of legal consequences which are not
meant to apply to some circumstances, and which the courts sometimes ‘disapply’
in other circumstances. Rights to appeal to the Administrative Appeals Tribunal,
for example, apply even where the original decisions were nullities.20 Similarly,
the ADJR Act allows for judicial review of invalid decisions, even though its rights
are all framed in terms of challenges to decisions of an administrative character
made under enactments.21 There is nothing illogical in this. Parliament can grant
appeal or judicial review rights as it sees fit, and it would be perverse to give a more
narrow construction to its Acts in these two contexts. There are also situations
where the court will not treat a decision as a nullity even though there has been
a jurisdictional error. There is no point, for example, in declaring a decision to
be null and void if the only challengers lacked standing. Similarly, a challenger

who had standing, but who applied too late, might well be denied relief, with the
result that the ‘decision’ remains in effect.

The High Court’s special position

The High Court’s judicial review jurisdiction is peculiar in a number of respects.
Relevant to this chapter, its constitutional writ jurisdiction is not bounded by
‘jurisdictional error’ in quite the same way as in the state Supreme Courts exer-
cising their prerogative writ jurisdiction.
The High Court’s constitutional writ jurisdiction is available to stop Common-
wealth officers enforcing unconstitutional legislation, and to stop them breaching
any constitutional limits to the Commonwealth’s executive power as granted by
s61 of the Constitution. The writs are available even though one would not nor-
mally talk of such officers acting in ‘excess of jurisdiction’ in those situations.22
The constitutional writ jurisdiction lies against Commonwealth officers, a
term which includes the judges of both the inferior federal courts (such as the
Federal Magistrates Court) and the superior federal courts (such as the Fed-
eral Court and the Family Court).23 In one sense, the presence of constitutional
restraints upon all Australian courts means that this country has no equivalent
of the English court of ‘unlimited jurisdiction’. The state and territory Supreme
Courts are nevertheless treated as being immune from judicial review, as if their
jurisdictions were unlimited. The requirement for ‘jurisdictional error’ can there-
fore explain Supreme Court immunity.24 However, the constitutional writs’ cov-
erage of federal judges beneath the High Court cannot be convincingly explained
on the basis of an application of or extrapolation from the general law’s concept
of ‘jurisdictional error’. The Federal Court, for example, always has jurisdiction to
determine its jurisdiction, and its decisions in that regard are legally effective and
remain so unless and until they are reversed by the High Court, which says that
it can do this in either its appellate or original jurisdiction.25 This aspect of the
High Court’s constitutional writ jurisdiction therefore goes further than ‘excess
of jurisdiction’, if by that one means to involve the concept of nullity.26 On that
assumption, it is explicable only as an essential feature of the High Court’s role in
enforcing all limitations on the exercise of power by the Commonwealth Parlia-
ment’s statutory creations. Because the High Court’s judicial review jurisdiction
is discretionary, the court has been able to state that it will usually decline to
exercise its constitutional writ jurisdiction against federal judicial officers where
the applicant could and should be using a statutory appeal mechanism.27
Certiorari is not directly entrenched in the High Court’s original jurisdiction,
but it frequently issues as ancillary to the court’s constitutional writ jurisdiction.
Considerable doubts have been expressed as to whether ancillary certiorari is
available in the High Court for non-jurisdictional errors of law.28
It should also be noted that some members of the High Court may have been
hinting that the term might have a narrower meaning in the constitutional writ

context.29 At this stage, one can only guess at their concerns, but these probably
include a wish to entrench as little as possible of the grounds of review, leaving it
ultimately to the mercy of the parliamentary drafter as to whether the legislature
meant30 a decision maker’s breach of a statutory requirement to result in nullity.31
Kirby J has protested that giving a narrower meaning to jurisdictional error in
the context of the constitutional writs ‘would add another layer of obscurity
to what are already elusive distinctions’.32 There is an obvious attraction to his
Honour’s appeal to doctrinal simplicity, but guessing once more, one suspects that
there is more to his protest than that. Kirby J believes in expanding the grounds
of judicial review, whether that be via the constitutional writs, the common
law’s prohibition or mandamus, or certiorari or injunction. His Honour sees an
anomaly in certiorari and injunction being available for non-jurisdictional error
of law whilst mandamus and prohibition are not. His preferred way out of that
anomaly would be to broaden the reach of mandamus and prohibition.33 On the
other hand, Hayne and Callinan JJ have suggested the exact opposite, namely a
retraction of certiorari and injunction, at least in the exercise of the High Court’s
original jurisdiction.34

The modern catalogue of jurisdictional errors

The grounds of judicial review frequently overlap and frequently chase the same
or largely the same goals. More importantly for present purposes, they are fre-
quently expressed in ways which are either entirely circular or open-ended. It
is as if the judges want to avoid being pinned down. Consider the High Court’s
classic definition of ‘jurisdictional error’ in Craig v South Australia.35 It is too
long to quote, but it is suggested that this is because it was carefully crafted to
say very little indeed. What follows is a fair summary of Craig’s catalogue of a
decision maker’s jurisdictional errors. It uses only some of its original wording,
but to help the reader who is familiar with that original, it adheres to Craig’s
expositional sequence. Craig listed six errors as jurisdictional. In the same order
as they appeared in Craig, these were:

(1) A mistaken assertion or denial of the very existence of jurisdiction.
(2) A misapprehension or disregard of the nature or limits of the decision maker’s
functions or powers.
(3) Acting wholly or partly outside the general area of the decision maker’s jurisdic-
tion, by entertaining issues or making the types of decisions or orders which are
forbidden under any circumstances (e.g. a civil court trying a criminal charge).


ńňđ. 13
(âńĺăî 19)