ńňđ. 11
(âńĺăî 19)




meant to be an expanded form of its common law counterpart, albeit within the
confines established by s5(3).111
The ‘harmonisation’ approach suggested by Mason CJ in Bond seems to have
been generally accepted, at least in the sense of not being the subject of any
detailed judicial analysis in subsequent cases. It was briefly referred to by the
High Court in Minister for Immigration v Rajamanikkam,112 and whilst it did not
receive unequivocal endorsement, nor was it expressly rejected.113 The lack of
specific analysis in Rajamanikkam was not, however, surprising given that the
case did not directly concern the two ADJR Act grounds but dealt with analogous
judicial review provisions of the Migration Act. In this respect, whilst the Migration
Act ‘no evidence’ ground was identical to its ADJR Act counterpart,114 the ‘error
of law’ ground115 was significantly different, thus ruling out the need to address
the assumed relationship between the two ADJR Act grounds put forward in
Some difficulties remain with the Bond ‘harmonisation’ approach. First, the
demarcation it makes between the ‘error of law’ and ‘no evidence’ grounds
assumes that the common law position in Australia concerning ‘no evidence’
is readily identifiable as at the date of enactment of the ADJR Act.116 Secondly,
it also seems to assume that the substantive content of the ADJR Act grounds
of review remains immutably fixed, reflecting the law as it stood at the date of
enactment of the ADJR Act. Yet, there are persuasive arguments for interpreting
the provisions containing the ADJR Act grounds of review117 in an ambulatory
manner, so that their scope and content are capable of varying over time, in tune
with common law judicial development.118
Finally, whilst Rajamanikkam did not expressly disapprove of the Bond ‘har-
monisation’ approach, comments made by Gaudron and McHugh JJ seem to
imply an important qualification. In their view, a single finding may involve a
reviewable error under both the ‘error of law’ and the ‘no evidence’ grounds of
the ADJR Act.119 If that is so, then the rationale for maintaining a separate field of
operation for each ground becomes less convincing. As well as this, if common law
development in Australia follows the direction of English law on ‘no evidence’,
then the specific statutory ‘no evidence’ ground may become less relevant.

Section 5(3)(a) – the first ‘no evidence’ limb
The first limb of the s5(1)(h) ‘no evidence’ ground applies where the relevant
decision-making power is structured in such as way as to mean that the decision
maker was ‘required by law’ to reach a decision only if ‘a particular matter was
established’. Given that the ADJR Act only operates in relation to the exercise of
statutory decision-making powers, ‘required by law’ will generally mean that the
statute creating the decision-making power must specify the ‘particular matter’
which must be established.
In that respect, it is not surprising that the first limb of the rule has drawn
analogies with the common law ‘jurisdictional fact doctrine’. For instance, in

Western Television Ltd v Australian Broadcasting Tribunal,120 Pincus J said that
s5(3)(a) applied to legislation which, either expressly or by implication, provides
that the making of decision ‘A’ depends upon the establishment of matter ‘B’.121
Similarly, in Television Capricornia Pty Ltd v Australian Broadcasting Tribunal,122
Wilcox J stated that that s5(3)(a) applies ‘where the establishment of a particular
fact is a pre-condition in law to the decision’.123
The analogy is well illustrated by the reasoning and result in each of these
cases. In Western Television Ltd, the legislation required a tribunal to grant a tele-
vision broadcasting licence to the ‘most suitable applicant’.124 As there was no
statutory definition of this term, the tribunal, itself, identified certain charac-
teristics it considered relevant to suitability, one of which was the ‘shareholding
stability’ of the companies competing for the licence. According to the reasoning
of the tribunal, shareholding stability was necessary to ensure that the success-
ful contender would not fall under different control subsequent to having been
granted a licence. The tribunal then proceeded on the assumption that corporate
shareholders were more likely to ‘sell out’ than individual shareholders. On that
basis, the licence was awarded to the applicant’s rival, whose shares were held
by a higher proportion of individual, as opposed to corporate, shareholders.
In an application for judicial review under the ADJR Act, the applicant asserted
that in terms of s5(3)(a), there was ‘no evidence’ to support the tribunal’s assump-
tion about corporate shareholders. However, Pincus J pointed out that whilst
shareholding stability was a relevant aspect of the statutory requirement of suit-
ability, it was not, in itself, a criterion which the legislation expressly required
the tribunal to establish in order to reach a valid decision.125
In a similar fashion, the relevant statute in Television Capricornia Pty Ltd spec-
ified that the ‘financial, technical and management capabilities’ of an applicant
had to be considered by the tribunal in reaching a decision to grant a television
broadcasting licence. However, the legislation did not specify that these criteria
were exclusive, thus allowing the tribunal to consider other relevant matters. On
that basis, the court ruled out any reliance on s5(3)(a) of the ADJR Act that there
was no evidence to support the tribunal’s conclusion that the successful applicant
had adequate ‘financial etc. capabilities’. As Wilcox J explained, satisfaction of
the ‘financial etc. capabilities’ requirements was not, as a matter of law, specified
by the legislation as a necessary precondition to a decision to grant a licence.126
Rulings in cases of this nature indicate a rather confined role for the first limb
of the statutory ‘no evidence’ rule. Many statutory decision-making powers are
cast in terms which require the decision maker to consider or have regard to
specified matters in reaching a decision. Yet the rule seems to require that the
‘particular matter’ relied on by an applicant for judicial review is expressly and
specifically identifiable in the statute as a requirement and secondly, that the
statute describes it in terms which show that its existence is clearly a condition
precedent to the valid exercise of the decision-making power. In other words, it
will not be sufficient to simply show that, in light of the statute, the matter is, by
law, a relevant consideration in reaching a decision.127

The first limb of the statutory ‘no evidence’ ground also refers to the need
to show that there was no ‘evidence or other material’ from which the decision
maker could reasonably be satisfied that the ‘particular matter’ was established.
As indicated earlier, given that most administrative decision makers, including
tribunals, are not bound by the rules of evidence, the term ‘other material’ allows
the rule to operate in relation to matter which would not be admissible in accor-
dance with the rules of evidence.128 Nonetheless, there must be limits on the
kind of material to which the provision can relate. Mason CJ in Bond said that
the ground allows for review of a finding of fact where ‘there is no probative
evidence to support it’,129 a threshold which is consistent with the general rule
as to the kind of material the law expects administrative decision makers to rely
on in reaching decisions.130
Finally, the wording of the first limb has been described as lessening the burden
under common law of having to demonstrate a complete absence of evidence to
support the decision.131 This stems from the fact that it simply requires an absence
of evidence or material from which the decision maker ‘could reasonably be
satisfied that the particular matter was established’. At the same time, however,
as Aronson, Dyer and Groves point out, an odd feature of the rule is that it
is confined to the decision maker’s reasoning process, irrespective of whether
or not the decision would have been different if other material was before the
decision maker.132

Section 5(3)(b) – the second ‘no evidence’ limb
The second limb of the ‘no evidence’ rule in s5(3)(b) of the ADJR Act operates
where a decision is based on the existence of ‘a particular fact’ in circumstances
where an applicant can establish that it ‘did not exist’. As the wording indicates,
there are at least three issues involved in its application, the first two of which
are inter-related.
The first concerns the manner in which the ‘particular fact’ is identified, rais-
ing a question as to the level of specificity or abstraction involved in the process
of identification. In a broad sense, it is generally understood that the law distin-
guishes between ‘primary’ or ‘evidential’ facts – which are those observable by
witnesses or proved by testimony – and ‘secondary’ or ‘ultimate’ facts – which
are those inferred by a process reasoning from the existence of ‘primary’ or ‘evi-
dential’ facts.133 Given that ‘primary’ or ‘evidential’ facts do not ordinarily attract
immediate legal consequences,134 there seems to be an assumption that, for the
most part, the second limb of the ‘no evidence’ rule requires identification of a
‘secondary’ or ‘ultimate fact’ as the ‘particular fact.135 This view is also consistent
with the approach taken by the majority on Azzopardi v Tasman UEB Industries
Limited136 that errors of law are unlikely to arise in the process of determining
primary facts.137
Secondary or ultimate facts may be constituted by a wrong assumption by
a decision maker about the implications of a primary fact. This is illustrated

by Curragh Queensland Mining Ltd v Daniel.138 In that case, the applicant had
entered into a contract to supply a quantity of coal by a specified date and had
imported mining equipment in order to meet its obligations under the contract. An
import tariff concession was available where, according to the statute, no suitably
equivalent locally manufactured product was reasonably available. In applying
for a tariff concession, the applicant argued that the locally manufactured product
could not move the quantities of coal in the time frame necessary to enable it to
meet its contractual obligations. However its application was rejected on the basis
that it could have arranged the terms of its contract for the supply of coal in such
a way as to allow for the use of a locally manufactured product.
According to the Full Federal Court, the decision to refuse a tariff concession
was based on a fact which was shown to be incorrect, namely, that the appli-
cant could have negotiated a later delivery date for the coal than that stipulated
in its contract. In other words, the ‘particular fact’ was the (incorrect) assump-
tion by the decision maker about the legal liability of the applicant under the
The second and inter-related issue is that the decision must be shown to be
‘based on’ the particular fact. In Bond, Mason CJ139 had said that a decision
is ‘based on’ a particular fact where that fact is ‘critical to the making of the
decision’,140 a description which the court subsequently approved of in Minister
for Immigration and Multicultural Affairs v Rajamanikkam.141 This means that the
‘particular fact’ must attain a high level of significance in the process of reasoning
leading to the decision so that, at the very least, findings on matters of peripheral
importance are clearly excluded.142
Gleeson CJ also warned that the higher the level of particularity employed in
identifying the ‘particular fact’, the harder will it be to demonstrate the decision
was ‘based on’ the existence of that fact.143 This again emphasises the appar-
ent assumption mentioned earlier, that the first ‘no evidence’ limb is primarily
concerned with secondary or ultimate facts, rather than primary facts.
It is the actual process of reasoning employed by the decision maker which
is critical in determining whether a decision is ‘based on’ a particular fact. This
was initially highlighted by the Full Federal Court in Curragh Queensland Min-
ing Ltd v Daniel.144 According to Black CJ (with whom Spender and Gummow
JJ agreed):
A decision may be based upon the existence of many facts; it will be based upon the
existence of each particular fact that is critical to the making of the decision. A small
factual link in a chain of reasoning; if it is truly a link in the chain and there are no
parallel links, may be just as critical to the decision, and just as much a fact upon which
the decision is based, as a fact that is of more obvious immediate importance. A decision
may also be based on a finding of fact that, critically, leads the decision maker to take
one path in the process of reasoning rather than another and so to come to a different

This idea that a decision may be ‘based on’ a fact which is otherwise only
a ‘small factual link’ in the process of reasoning received general endorsement
from the High Court in Rajamanikkam,146 although Gaudron and McHugh JJ

cautioned that the finding of fact must be one ‘without which the decision in
question either could not or would not have been reached’.147 According to their
methodology, whether a decision ‘could’ have been reached without the existence
of a particular factual finding depends on a process of logic or the law to be
applied148 and whether or not it ‘would’ have been made requires analysis of the
decision, the decision-making process itself and the reasons given for reaching
the decision.149 In general, the task centres on the decision maker’s reasons and
whether or not any supposed ‘factual link’ in the chain of reasoning was pivotal
in the sense that it could only have led in the direction of the decision.150
Rajamanikkam151 dealt with a decision by a refugee review tribunal, affirming
a decision of a delegate of the Minister to reject the respondents’ applications for
protection visas under migration law. According to the tribunal, they did not
satisfy the criteria for refugee status inasmuch as there was no well founded
fear of them being persecuted if returned to Sri Lanka.152 The tribunal’s written
reasons listed eight factors in support of its decision, two of which were based
on assumptions by the tribunal that on two occasions during interviews and
questioning, the first respondent had deliberately conveyed a false impression
that it was unsafe for him to return to an area where he had formerly lived. These
assumptions stemmed from the tribunal’s belief that in relation to each occasion,
the first respondent had previously given inconsistent accounts. The tribunal
therefore concluded that he was not a credible witness in relation to the answers
he gave on each of the two occasions.
The tribunal’s belief as to the existence of prior inconsistent statements by the
first respondent was subsequently shown to be wrong. First, it had overlooked
other written material he had provided which removed any cause for doubting his
credibility and secondly, it had simply made a mistake as to what he had actually
said in verbal evidence at the hearing. On that basis, the Full Federal Court
described the tribunal’s assumptions as ‘facts’ of central importance, without
which it would not have reached its decision.153
When the matter reached the High Court, three of the four members of the
majority154 accepted the proposition that, first, the tribunal’s assumptions were
non-existent ‘facts’ and that secondly, the tribunal took account of them in reach-
ing its decision.155 However, the majority were not prepared to find that this, in
itself, meant that the tribunal’s decision was based on these non-existent facts. In
simple terms, that was because the tribunal had listed eight reasons in support
of its final decision, only two of which were based on the non-existent ‘facts’.156
Interestingly, Callinan J thought that the second limb of the ‘no evidence’ rule
should only apply where a decision is shown to be based upon the existence of a
positive fact, rather than a negative finding as to the existence of a fact, as was the
case here. This issue has been the subject of previous judicial consideration,157
although the rationale offered by Callinan J stems from a perceived need to
ensure that the second ‘no evidence’ limb is not available where the basis of
review involves nothing more than an erroneous finding of fact.158
Most administrative decision-making bodies are now obliged to provide rea-
sons for their decisions and for the most part, good practice dictates the provision

of more than one isolated reason. In that respect, it would seem absurd to say
that a decision cannot be ‘based upon’ a non-existent fact where more than one
reason is given. Does Rajamanikkam say that the non-existent facts in that case
were rendered less than ‘critical to the making of the decision’159 simply because
additional reasons in support of the decision were formally recorded? Or was it
because, in a more abstract sense, the additional, unrelated reasons reduced the
non-existent facts to the status of simply being ‘parallel links’160 in the chain of
In another sense, the majority approach in Rajamanikkam separates the pro-
cess of reaching a decision from the ultimate decision itself. On that basis, the
tribunal’s conclusion as to the first respondent’s credibility could be viewed as
simply an interim step which was ‘based upon’ the non-existent facts whilst its
ultimate decision to refuse a protection visa was ‘based on’ on a greater num-
ber of facts, derived from all of the relevant evidence before it.161 As Callinan
J emphasised, the decision under review was not whether the first respondent
had told the truth about two matters arising during the course of the tribunal’s
proceedings but whether or not the tribunal was satisfied that the respondents
were eligible for protection visas.162
The third requirement of the second ‘no evidence’ limb is the need to show
that the particular fact relied upon ‘did not exist’. If the intention pursued by the
framers of the ADJR Act was to ensure that the statutory ‘no evidence’ ground
would be a more restricted version of the common law ‘no evidence’ rule, then it is
certainly achieved in this requirement. This is clear from the fact that an applicant
is required to do more than demonstrate an absence of evidence justifying the
decision. It is necessary to go further and negative the actual existence of the fact
relied on by the decision maker.163
Although onerous enough in itself, this task would be all the more difficult
if applicants were confined to the evidence before the decision maker at the
time of the decision. However, the accepted approach seems to allow recourse to
evidence assembled subsequent to the decision for the purpose of demonstrating
that the fact did not exist at the time the decision was made.164

The relationship between s5(1)(h) and s5(3)
The final aspect of the statutory ‘no evidence’ ground concerns the relationship
between the primary expression of the ground in s5(1)(h) and the manner of
operation of the two limbs in s5(3) of the ADJR Act. The opening words of s5(3)
specify that the ground of review set forth in s5(1)(h) is ‘not to be taken to have
been made out unless’ the provisions of sub-paragraph (a) or (b) of s5(3) apply
and in Television Capricornia Pty Ltd, Wilcox J emphasised that this was consistent
with the overall intention of the legislature to create a restricted version of the
common law ‘no evidence’ rule.165
Assuming for the moment then that s5(3) has a qualifying or limiting function,
does it comprehensively define the content of the primary ‘no evidence’ ground

in s5(1)(h)? In Bond, Mason CJ (with whom Brennan and Deane JJ agreed) said
that the effect of s5(3) is to ‘limit severely the area of operation of the ground in
s5(1)(h)’,166 a view consistent with the idea that s5(3) exhaustively defines the
content of s5(1)(h) in the sense that the primary ground in s5(1)(h) is established
as soon as either of the two limbs in s5(3) is made out. However, that does not
appear to have been the approach adopted by the Federal Court. In Curragh
Queensland Mining Ltd v Daniel167 Black CJ made it clear that it was necessary to
separately establish both the primary ground in s5(1)(h) as well as either limb
in s5(3) and other decisions indicate the same approach.168
The issue was briefly referred to by the High Court in Rajamanikkam where
Gleeson CJ appears to have endorsed the approach taken in Curragh Queens-
land Mining Ltd by Black CJ.169 Two of the other majority judges, Gaudron and
McHugh JJ, subjected the approach taken by Mason CJ in Bond to a certain level
of criticism.170 In their view, to say that s5(3) limited the operation of s5(1)(h)
was to suggest that, somehow, s5(1)(h) bore a more extensive meaning than its
actual terms suggested. In any event, according to their Honours, that view could
only be possible if the Bond ‘harmonisation’ approach to the ‘error of law’ and
‘no evidence’ grounds in ss5(1)(g) and 5(1)(h) was correct, yet, as explained
earlier, no firm opinion was offered on that issue because it was not active in
the case.171 Ultimately, Gaudron and McHugh JJ took the view that s5(1)(h)
is simply a discrete and independent ground of review, the content of which is
identified in s5(3).172 In a similar vein, Kirby J felt that s5(3) did not qualify the
primary ‘no evidence’ ground but was rather a statement of the content of its

The ADJR Act: Where to now for the relationship between
s5(1)(h) and s5(3)?
At the very least, the inclusion of a ‘no evidence’ ground in the ‘ADJR Act system’
of judicial review acknowledges at the very least that under common law, a
complete absence of evidence on which to support a decision constitutes an error
of law. At the same time, the inclusion of a general ‘error of law’ ground has
left room for some uncertainty as to the precise boundaries of each of these
grounds. As the preceding examination also shows, the specific ‘no evidence’
ground comprises a number of somewhat technical requirements and the rather
confined and restrictive judicial approach taken in interpreting its terms means
that it is difficult to predict how it will evolve, in comparison with judicial review
for ‘no evidence’ under the general law.

In an overall sense, any movement towards a broader approach in the gen-
eral law in Australia concerning judicial review for ‘no evidence’ appears small,
compared with developments under English law. Whilst there are certainly
some suggestions that broader means of judicial supervision of fact finding and

evidence may evolve,174 the basic approach seems to remain for the moment
firmly within the conventional boundaries of judicial review, defined by the
enduring influence of the ‘legality merits’ distinction.
The same is largely true of the reformed statutory ‘ADJR Act system’ of judicial
review, which now exists in a number of Australian jurisdictions. Although in
some respects it appears to liberalise pre-existing restrictions on review for ‘no
evidence’, it appears to remain, by and large, reflective of the basic common
law position in Australia. Indeed, as the previous examination of recent case law
show, it appears to have unduly complicated the basic position by creating an
overlay of technical and restrictive requirements. Perhaps in that respect, greater
possibility for future development of this area of law in Australia may lie within
the general law of judicial review.
Failure to exercise discretion
or perform duties
Maria O’Sullivan

Administrative decision makers may be given a discretionary power under a
statute or required to perform a duty. Of these two, the law relating to the exercise
of a discretionary power has been the greater subject of substantial jurisprudence
and academic commentary. The central debate in this context is the extent to
which administrative law should constrain the exercise of discretion – that is, how
to achieve a balance between ensuring that decision makers consider the merits
of individual cases, whilst also recognising the bureaucratic imperatives of con-
sistency and efficiency. In this respect, this topic presents a challenge for admin-
istrative law which must (to some extent at least) reflect the political reality of
administrative decision-making, whilst also ensuring that decisions are made
The common thread in the jurisprudence in this area is that a decision maker
granted power by Parliament under a statute must be the person to make the
decision. Thus he or she cannot fetter that discretion, improperly delegate to
another or, if a duty is applicable, fail to perform that duty. Before discussing the
grounds for judicial review which arise in this respect, it is necessary to discuss
the meaning of the term ‘discretion’ in administrative law and to set the grounds
in their political context.

The meaning of discretion

The term ‘discretion’ indicates the existence of a level of choice in making a
decision and is usually expressed by use of the word ‘may’ in a statutory provision.1
This can be contrasted with those statutory provisions which require a decision
maker to make a decision if certain criteria are satisfied.2 The important aspect of
a discretion in terms of administrative law is that it is designed to allow a decision
maker to examine the merits of a particular case, rather than applying criteria in
an automatic fashion.


The political context

Although many statutory provisions place discretionary powers in the hands of
ministers, in practice it is normally administrators in government departments
who make, or at least play a central role in making, those decisions. This is largely
due to the fact that ministers are not able to attend personally to every decision
in their portfolio. Thus, in the vast majority of cases, a minister will delegate such
powers to public servants in his or her department. Alternatively, he or she may
ask departmental officers to write a report on the issue, with recommendations
as to the best course of action to take. The question this raises for administrative
law is how to accommodate these practices, whilst also ensuring that powers are
exercised within the bounds of the statute.
Additionally, the principle of responsible government provides that a minister
is accountable to Parliament for the operations of his or her portfolio (also known
as ‘ministerial responsibility’). The extent to which administrative law principles
should take account of this political accountability structure underpins some of
the debates about the control of discretion. One controversial question is the
extent to which the political relationships and practices which arise from respon-
sible government ought to influence the application by the courts of principles of
judicial review. This issue is complicated by the uncertain nature of responsible
government in Australian law. The High Court has accepted that the principle
of responsible government underpins Australia’s Constitution and governmen-
tal structures.3 It has also accepted that responsible government is a dynamic
concept.4 However, it has hesitated to explain clearly what the doctrine entails
or requires. The extent to which principles of responsible government interact
with judicial review principles therefore remains unclear.

Applicable grounds of judicial review

There are essentially three ways in which an administrative decision maker may
fail to exercise a discretion and therefore act ultra vires. First, a decision maker
cannot make a decision under the ‘dictation’ of another. Secondly, the decision
maker must not fetter the exercise of the discretion by inflexibly applying a rule
or policy or by adhering to an undertaking. Thirdly, the decision maker cannot
delegate the power to another decision maker unless this is permitted by the
statute. Additionally, there are also specific remedies available for a failure to
perform a duty.

Acting under dictation
This ground arises where a decision maker acts under the instructions or
influence of another person, or of a policy, in such a way that he or she fails

to consider the merits of a particular case. There is some overlap between the
ground of acting under dictation and that of inflexible application of policy.5
The applicable ground of review for acting under dictation is s5(1)(e) with
5(2)(e) of the Administrative Decisions (Judicial Review) Act 1977 (Cth) (ADJR
Act), which provides that an applicant can apply for an order of review in relation
to ‘an exercise of a personal discretionary power at the direction or behest of
another person’.6
In H Lavender and Son Ltd v Minister of Housing and Local Government,7 the
issue was whether the Minister for Housing had acted under the dictation of the
Minister for Agriculture in refusing planning permission to Lavender. The Court
held that, although the minister could consider the views of others, the minister
must be ‘open to persuasion’ in relation to the decision and the application of the
policy.8 Importantly, the Court stated that whilst the decision of the Minister for
Agriculture might be a decisive factor for the Minister of Housing when taking
into account all relevant considerations, a policy could not be applied in such a
way that it was the only material consideration.9
Lavender involved the dictation of a minister by another minister. However,
what principles should apply where Parliament specifically grants a discretion to
a senior public servant, rather than a minister? Should that official act pursuant
to ministerial directions? The Australian line of authority on this point is obscure
due to divergent judicial opinion in the two leading High Court cases on this
issue: Ipec Air and Ansett.
In R v Anderson; Ex parte Ipec Air Pty Ltd,10 Ipec Air applied to the Director-
General of Civil Aviation (the head of the Department of Civil Aviation) for per-
mission to import aircraft under the Customs (Prohibited Imports) Regulations.
The Director-General ultimately refused to grant permission for the importation
after the Minister for Civil Aviation informed him that such permission would
be against the government’s ‘two airlines’ policy.11 The High Court, by a three to
two majority, held that the ground of acting under dictation was not made out
on the facts.12 The judgments of their Honours reflect essentially three differ-
ent positions on the issue of acting under dictation. Kitto and Menzies JJ took a
stricter view of the ability of ministers to direct public servants.13 They conceded
that a public servant could take account of some matter of general government
policy, as long as he or she arrived at a decision of his or her own.14 However,
they noted that the Regulations committed the decision-making power to the
Director-General which indicated that the decision was intended to be made
at a departmental rather than ministerial (that is, the political) level.15 In this
case, the minister had effectively made the decision, thereby contradicting the
Regulations.16 Importantly, Menzies J made a distinction between the function
of ministers and heads of departments, noting that: ‘the sound theory behind
conferring a discretion upon a department head rather than his minister is that
government policy should not outweigh every other consideration’.17
Taylor and Owen JJ adopted an ‘intermediate’ position, finding that govern-
ment policy was a proper matter for consideration by the Director-General and,

on the evidence, that the decision had in fact been made by him.18 At the other end
of the spectrum, Windeyer J found that the Director-General must have regard
to government policy and ‘exercise his functions accordingly’.19 He added that
‘the only consideration by which the Director-General could properly have been
guided was the policy of the Government’,20 and that the Director was under a
duty to ‘obey all lawful directions of the Minister’.21 In coming to this conclusion,
his Honour referred to public service legislation and the doctrine of ministerial
The later High Court case of Ansett Transport Industries (Operations) Pty Ltd v
Commonwealth23 concerned the Secretary of the Department of Transport who
was given power under the Customs (Prohibited Imports) Regulations to permit
the importation of aircraft (thus, an area raising economic policy issues simi-
lar to those in Ipec). The High Court expressed differing views of the validity
of ministerial directions. The majority stated that a decision taken by a public
servant at the direction of his or her minister does not constitute acting under
dictation. Regarding the proper influence of government policy, Barwick CJ and
Murphy J, the only members of the court who had been ministers prior to their
judicial appointment, said that in an area such as aviation policy (raising impor-
tant political and economic issues of national concern) the head of a department
would be bound to carry out government policy and the lawful directions of his
or her minister.24 In taking this view Murphy J seemed to be influenced greatly
by the notion of ministerial responsibility.25 Likewise, Aickin J stated that there
is ‘nothing improper’ in a minister directing the secretary of a department to act
in a particular manner and, in fact, in some cases the secretary might be under a
duty to act in accordance with the policy of the government of the day.26 Gibbs
J stated that it would not be wrong for the secretary in exercising his discretion
‘to give weight, and indeed conclusive weight, to the policy of the government’.27
The majority Justices adopted an approach similar to that of Windeyer
in Ipec.
Mason J, on the other hand, was more cautious. He took a similar approach to
that of Kitto and Menzies JJ in Ipec by pointing out that the Regulations placed
the discretion in the hands of the Director, rather than the minister. Thus, if the
secretary was obliged to act according to ministerial direction, then the decision is
no longer the secretary’s.28 Mason J did recognise the importance of government
policy in the area of aircraft importation and thus noted that the decision maker
may ‘have regard to any relevant government policy’, but emphasised that he
or she nevertheless has to decide for themselves ‘whether the existence of the
policy is decisive of the application’.29 In this regard, the secretary is not ‘entitled
to abdicate his responsibility for making a decision by merely acting on a direction
given to him by the minister’.30 Mason J explicitly disagreed with the approach
taken by Windeyer J in Ipec.31
All the judges in Ipec and Ansett accepted that government policy may consti-
tute a relevant consideration to be taken into account by a decision maker. The
point where they diverge is the extent to which a decision maker can give such
policy ‘decisive’ or ‘conclusive’ weight. However, the line of authority (such as

it is) appears to be that in certain circumstances, a departmental officer may be
obliged to follow government policy or comply with ministerial directions, even
though there is no explicit statutory provision imposing such an obligation.
Ipec and Ansett concerned the heads of government departments. Should such
an approach also be applicable to statutory authorities and tribunals estab-
lished as independent bodies? This question arose in the 1981 case of Bread
Manufacturers,32 where the High Court held that the Prices Commission (an inde-
pendent body) was not precluded from ascertaining the minister’s view before
making a particular order.33 This conclusion seemed to depend a great deal on
the statute in question which gave the minister the power to veto a decision by
the Commission in relation to the setting of prices.34 The Court refrained from
formulating a clear legal rule which could be applied to acting under dictation
cases (in light of the diverging approaches of the Court in Ipec and Ansett). Mason
and Wilson JJ merely stated that it was not possible to formulate a general rule
as the position depended on a variety of considerations, including the ‘nature of
the question to be decided, the character of the tribunal, and the general drift
of the statutory provisions . . . as well as the views expressed on behalf of the
Government’.35 These comments illustrate the difficulties posed in adopting set
principles in administrative law given that much will depend on the individual
statute in question. Since the decisions in the three cases mentioned above, the
law relating to the acting under dictation ground remains unclear in a number
of respects. One aspect pertains to the extent to which the courts should give
effect to statutory provisions permitting ministers to give directions to decision
makers as to the exercise of their statutory powers.36 Such provisions, which
are increasingly enacted, tend to allow only for general directions (rather than
directions as to the outcome of a particular decision).37 Can a minister direct an
administrator as to the decision to be reached, or are the directions to be lim-
ited to only general guidance as to the exercise of the discretion? The position
very much depends on the statute in question and the law is still unclear on this
issue.38 Where there is an explicit statutory provision allowing for general minis-
terial directions, case law suggests that the minister will generally not be able to
direct a decision maker as to the outcome of a decision.39 However, it should be
noted that there is some conflicting jurisprudence in this regard.40 Sir Anthony
Mason, writing extra-judicially in 1989, observed:

One of the unresolved problems of administrative justice is that we have failed to
evolve principles spelling out the circumstances in which a decision maker must act
independently of political direction or influence, as compared to those in which he is
subject to such direction or influence. The questions which were not finally answered
in R v Anderson: Ex parte Ipec-Air Pty Ltd and Ansett Transport Industries (Operations)
Pty Ltd v The Commonwealth still remain unanswered.41

Another somewhat controversial line of cases have been those dealing with
cabinet influence on ministerial decisions. Australian courts have held that, at
least in some circumstances, a minister may consult cabinet prior to making an
administrative decision without infringing the rule against acting under

dictation.42 On a more positive note, the courts have made progress in defin-
ing what is meant by ‘behest’ as that term is used in the ADJR Act. For instance,
in Telstra Corp Ltd v Kendall, the Full Federal Court held that ‘the word “behest”
cannot simply be a substitution for request’.43 This indicates that it may be unnec-
essary to establish that a decision maker acted under duress in order to argue
this ground of judicial review.
Examining the approaches taken in the above cases more broadly, a number
of questions are raised. Do the approaches taken in the cases remain relevant
in today’s political environment? For instance, a number of the judges in Ipec
and Ansett placed an emphasis on ministerial responsibility. However, does that
emphasis remain well-founded in practice? Furthermore, can it be said that the
courts have given sufficient effect to parliamentary intention in the way they
have stated the principles in this area? Some of these issues will be canvassed in
the concluding section of this chapter.

Inflexible application of policy
A decision maker may take into account government policy in certain circum-
stances. The term ‘policy’ in this context can encompass high-level ministerial
policy, departmental policy or guidelines, or a personal rule or policy of a decision
maker. It is thought that policy provides many benefits to administrative decision-
making, such as encouraging consistency, certainty, and efficiency – particularly
where numerous decision makers are making the same kind of decision across
a department.44 But policy also has significant disadvantages. For instance, an
administrative decision maker may become overly reliant on a policy, thereby
failing to consider the individual merits of a case. Thus, administrative law limits
the use of policy in a number of ways.
First, a policy must be consistent with the relevant statute. If it is not, an
applicant may argue that the policy is outside the scope of the Act pursuant to
s5(1)(d) of the ADJR Act,45 or that it is an irrelevant consideration.46 Secondly,
a lawful policy cannot be inflexibly applied by a decision maker. This ground
is codified in ss5(1)(e) with 5(2)(f ) of the ADJR Act which provides that an
applicant may lodge an order of review in relation to ‘an exercise of a discretionary
power in accordance with a rule or policy without regard to the merits of the
particular case’.47
In British Oxygen Co Ltd v Minister of Technology,48 the House of Lords held
that the general rule is that a decision maker exercising a statutory discretion
must not ‘shut his ears to an application’.49 Delivering judgment for the court,
Lord Reid noted that there could be no objection to a ministry or large authority
which has had to deal with a multitude of similar applications adopting a rule
or policy to deal with those applications, provided the authority keeps an open
mind to any applicant who wants to argue that their case is exceptional – that is,
a decision maker must always be willing to ‘listen to anyone with something new
to say’.50

The notion of an ‘open mind’ in relation to the application of policy was consid-
ered by the High Court in Green v Daniels.51 The applicant, who had recently left
school, was denied unemployment benefits on the basis that department policy
provided that school leavers could not receive such benefits until the beginning
of the next school year. Section 107 of the Social Security Act 1947 (Cth) set out
specific criteria for the grant of unemployment benefits, among which was a
requirement that the Director-General (the head of the department) had to be
satisfied that the applicant was unemployed. There was nothing in the legislation
which stated that school leavers were ineligible for these benefits.
The High Court found that the Director-General had not exercised his discre-
tion lawfully as required under the statute. The Court emphasised that the Act
did not confer ‘general discretion’ on the Director-General but set out ‘specific cri-
teria’ for the exercise of his discretion.52 On this view, the Director-General could
provide guidelines to decision makers in relation to the statute but they could
not be inconsistent with ‘a proper observance of the statutory criteria’.53 The
existence of specific criteria in s107 was a central aspect of the decision in Green v
Daniels. In this respect, some later cases have distinguished Green on this basis.54
This indicates that it may be difficult to establish that a policy is inconsistent with
a statute where the particular provision provides only a general discretion.
In later cases, the Australian Federal Court has emphasised the need for
decision makers to give ‘genuine and realistic consideration’ to the merits of
a case and be ready, where necessary, to depart from any applicable policy.55
Thus, a decision maker will be required to reflect this in his or her statement of
The decision maker in Green v Daniels was the head of a government depart-
ment. But to what extent are independent statutory authorities and tribunals
expected to take account of government policy? As with the principles enunci-
ated for the acting under dictation ground above, the general principle is that
much depends on context, that is:

. . . the extent to which an independent body may reflect established government policy
depends upon the character of the body, the nature of its functions and the relevance
to that charter and functions of the policy in question. There is no absolute rule that
the body must ignore known government policy. On the other hand, it must not be
so influenced by the policy that it fails to perform its own functions, as the statute

Finally, it should be noted that particular principles relating to government policy
apply to the Commonwealth Administrative Appeals Tribunal (AAT) and the
equivalent tribunal of general jurisdiction that exists in many states.57

Fettering of discretion by undertakings
The current position in Australian administrative law is that administrators are
not able to fetter the exercise of their discretion under a statute by treating

themselves bound by either a representation or a contractual undertaking. In
this regard, Australian courts have generally not supported the use of estoppel in
administrative law. For instance, in Minister for Immigration v Kurtovic,58 the Full
Court of the Federal Court held that an estoppel cannot be raised so as to prevent
the performance of a statutory duty or to hinder the exercise of a statutory dis-
cretion, as this would amount to an impermissible fetter upon the future exercise
of discretion conferred by the relevant statutory provision. A similar approach
was taken by the High Court in Quin in relation to change of policy.59 Putting
this aside, applicants may have some procedural rights in such a situation, that
is, they may be able to argue that they are entitled to natural justice (the right to
be heard) arising from a representation or the existence of a policy.60

Improper delegation
If a parliament vests a discretion in a particular decision maker, it is normally that
decision maker who must exercise the power. In many cases a statutory discretion
granted to a nominated person (usually a minister or department head) may be
exercised by an administrator in a government department in two ways: via
delegation or by the agency principle.
The common law prescribes a presumption against the delegation (or sub-
delegation) of power.61 This presumption springs from the principle that the
power should be exercised by the decision maker named in the statute. Delegation
is therefore only permitted if this common law presumption can be rebutted by
either an express or an implied authority to delegate in the relevant legislation.
Use of the agency principle is slightly different to this in that it avoids the common
law presumption against delegation by utilising a principal-agency relationship.
The primary difference in practice is that a delegate makes a decision in his or
her own name whereas an agent makes the decision in the name of the designated
decision maker.
There is no ground in the ADJR Act which specifically provides for ‘improper
delegation’. Instead, an applicant may utilise s5(1)(c), relating to lack of juris-
diction, or s5(1)(d), relating to unauthorised decisions.62 This ground may arise
not simply when it appears from the face of a decision that an unauthorised per-
son has made the decision (for example, because it is signed by a public servant,
without proper delegation from the minister) but also where a decision maker
simply ‘rubber-stamps’ a recommendation of his or her staff.


A statute will often set out an express power for the nominated decision maker
to delegate all of his or her powers under the statute. Regard must then be had
to any restrictions or conditions governing the power (such as a requirement to
publish the instrument of delegation).

In some circumstances, a power to delegate may be implied from the words,
structure and subject matter of a statute. In Ex parte Forster; re University of
Sydney,63 the Full Court of the Supreme Court of New South Wales considered
the validity of a decision by the University of Sydney to exclude a student for unsat-
isfactory performance. The central issue was whether the University Senate, in
delegating power to faculties and faculty committees to decide on student admis-
sion and exclusion, breached the prohibition against sub-delegation. The Court
held that, on the facts, the common law presumption against sub-delegation had
not been infringed, as it had been rebutted by construction of the statute and the
context in which it operated. The Court noted that the implication of delegation
from a statute must be considered ‘with due regard to the purpose and objects
of the statute, the character of the power which is conferred, the exigencies of
the occasions which may arise with respect to its exercise and other relevant
In applying this test to the facts, the Court noted that the objects of the relevant
legislation was the entire management of the affairs of a university and thus, the
University could not function without an ‘ample facility for delegation’.65 The
Court also noted that the importance of the subject matter may have a bearing
upon the permissibility of delegation or the appropriateness of the body to whom
the delegation is made.66

Agency/alter ego principle

The common law also recognises that in certain cases, the agency principle may
apply to some administrative decisions (also known as the ‘alter ego’ principle).
There is a slight difference in the way in which a power to delegate is implied from
a statute compared with implication of a power to act via agents. The implication
of delegation is made by direct and close reference to the terms of the statute,
with particular focus on evincing parliamentary intention. In contrast, use of the
agency principle looks not just to the statute, but also more broadly at the entire
decision-making process to establish whether a power to act via agents can be
implied due to administrative necessity.
In O’Reilly v Commissioners of the State Bank of Victoria,67 the Commissioner of
Taxation had validly delegated to the Deputy Commissioner of Taxation his power
to issue a notice under section 264 of the Income Tax Assessment Act 1936 (Cth) to
call witnesses to the taxation office to give evidence. This was done pursuant to an
express power of delegation in the Taxation Administration Act 1953 (Cth). That
power did not permit the Deputy Commissioner to sub-delegate. Subsequently,
the Deputy Commissioner gave written authorisations to Chief Investigative Offi-
cers in the department to issue notices in the Deputy Commissioner’s name to
require persons to attend the taxation office. In this instance, a notice calling
two witnesses to appear was sent by a departmental official who stamped the
notices with the signature of the Deputy Commissioner, pursuant to a general

authorisation by the Deputy Commissioner. In argument before the High Court,
the two persons affected argued that the notices were invalid on the basis that
they breached the common law rule against sub-delegation. In response, the
Deputy Commissioner argued that the agency principle applied.
The High Court held that the notices were valid as the departmental offi-
cial had acted as an authorised agent of the Deputy Commissioner.68 The court
relied on the principle of the leading English case of Carltona v Commissioners
of Works69 where the English Court of Appeal held that a minister given ‘mul-
tifarious functions’ was entitled to exercise his or her power via agents.70 That
approach was heavily influenced by the principle of ministerial responsibility,
with the Court noting that ministers would have to answer to Parliament for any
mistake or incompetency by a departmental official exercising such powers.71
The Carltona principle was not directly on point with the facts in O’Reilly given
that the latter involved the powers of a senior public servant, rather than those of
a minister. It was therefore questionable whether the English Court of Appeal’s
reliance on ministerial responsibility could be applied to the O’Reilly scenario
(given that heads of departments are not directly responsible to parliament in
the same way as ministers). The High Court answered this by recognising that the
decision in Carltona ‘depended in part’ on the special constitutional position of
ministers,72 but found that Carltona, and the line of decisions following it, also rest
on the recognition that ministers have multifarious functions and that ‘[m]inisters
are not alone in that position’.73 Thus, the Court held that the Carltona principle
was ‘equally persuasive to the head of any large government department’.74 The
Court added:

No permanent head of a department in the Public Service is expected to discharge per-
sonally all the duties which are performed in his name and for which he is accountable
to the responsible Minister.75

Consequently, the Court found that ‘there exists, as the Parliament must have
known, a practical necessity that the powers conferred on the Commissioner by
the Act should be exercised by the officers of the Department who were acting as
his authorised agents’.76
In dealing with the existence of the express power of delegation in the statute,
the Court held that practical administrative necessity may require a decision
maker to act via an agent despite the existence of an express power to delegate.77
The reasoning in O’Reilly can be questioned on a number of bases. One ques-
tion is whether the court accurately and properly gave effect to the intention
of Parliament in applying the agency principle in a case where Parliament had
specifically stated that the powers in question were not to be sub-delegated. If the
rule in Carltona is designed to be one of statutory interpretation, one can argue
that it did not function well in O’Reilly because the approach of the High Court
seemed at odds with the apparent intention of parliament. A related criticism
of the reasoning used in O’Reilly is that it tends to transform the interpretive
approach favoured in Carltona into one based on administrative convenience.

Failure to perform a duty
In circumstances where a decision maker has a ‘duty’ to make a decision (as that
term is defined in the ADJR Act), an applicant may also lodge an application
under s7 of the ADJR Act on the ground that the decision maker has failed to
make that decision.78 If there is no section in the relevant statute setting out a
time period for the making of the decision, the ADJR Act allows an applicant to
apply for an order on the grounds that there has been ‘unreasonable delay’ in
making the decision.79
Alternatively, an applicant may apply under the common law for the writ
of mandamus to compel an administrator to perform a statutory duty accord-
ing to law. Further, if there is a serious, extended delay in the handing down
of a decision, an applicant may also be able to argue that the delay has pre-
vented them from receiving a fair hearing pursuant to the principles of natural

An appraisal
Some justices in Ipec and Ansett justified their conclusions by invoking the doc-
trine of ministerial responsibility.81 Considering the contemporary political envi-
ronment, many commentators have stated that ministerial responsibility is no
longer a strong mechanism of accountability.82 To the extent that the reasoning
in those judgments relies on the effectiveness of ministerial responsibility as a
means of political accountability, it may be flawed.
A debatable issue is whether the courts have given effect to parliamentary
intention in the application of the abovementioned grounds of review. In O’Reilly,
the High Court justified the application of the agency principle on the basis that
Parliament could not have intended for the Commissioner to function without
a facility to operate via agency. This was despite the fact that Parliament had
enacted a clear statutory provision limiting delegation. The High Court reached
this conclusion by an interpretation of parliamentary intention that rested upon
the doubtful reasoning that ‘there exists, as the Parliament must have known’ a
practical necessity for application of the agency principle.83 This, however, is
highly questionable given that Parliament specifically limited the Deputy Com-
missioner’s ability to sub-delegate in the statute. If the role of the courts is
to ensure that decision makers act in accordance with the will of parliament
(expressed via legislation), is such an avoidance of an express prohibition on
sub-delegation justifiable?
Likewise, the approach of the majority of the High Court in Ipec and Ansett is
questionable as they appeared to take the view that, despite parliament specifi-
cally nominating a senior public servant (rather than the minister) as the reposi-
tory of a discretionary power, that power should be exercised in accordance with
ministerial policy. Arguably, the dissent of Menzies J in Ipec gave better effect to
Parliament’s intention in this regard. As has been stated in case law on this issue:

The principle underlying the rule against replacement of statutory discretion by a statu-
tory prohibition is simply respect for parliamentary sovereignty. Where parliament says
that in certain circumstances there is a discretion to grant permission, then no official
may replace that law by one to the opposite effect . . . 84

If the role of courts is to ensure that the boundaries of statutory powers which
are set down by parliament are respected, one may question whether this function
is being undertaken in some of the grounds discussed in this chapter. Moreover,
these concerns highlight some of the pertinent issues facing the future develop-
ment of administrative law as a whole. Namely, how should the law accommodate
the reality of bureaucratic decision-making processes, whilst also protecting the
rights of individuals and furthering the values of justice?
Procedural fairness: The hearing rule
Linda Pearson

The essential requirements of procedural fairness are fairness and detachment.1
Those requirements find expression in the two rules of procedural fairness. The
first is the hearing rule (audi alteram partem), which is the requirement to give
notice to the person affected by a decision that a decision is to be made, to disclose
information or material on which the decision maker proposes to rely, and to allow
an opportunity to put a case. The second is the rule against bias (nemo debet esse
judex in propria sua causa), which is the requirement that the decision maker be
free of actual bias or prejudgment, or the perception of prejudgment.2
The hearing rule is a broad topic that could potentially cover a vast range of
procedural requirements that arise from the principles of procedural fairness.
This chapter addresses four aspects of those requirements: (a) the application
of the hearing rule; (b) the exclusion or limitation of the hearing rule; (c) the
content of the hearing rule; and (d) the consequences of breach of the hearing
It is important to note at the outset that the term ‘hearing’ does not refer
just to that part of a decision-making process that may take the form of an oral
presentation of evidence and argument. Unlike judicial decision-making in an
adversarial system, which has its primary focus on the formal oral hearing of
evidence and submissions, administrative decision-making adopts many forms,
and need not necessarily involve an oral component. Analysis of the fairness of
the ‘hearing’ in this context necessarily requires consideration of the entirety
of the decision-making process. In some contexts, for example review by the
tribunals established to review migration decisions (the Migration Review Tri-
bunal [MRT] and the Refugee Review Tribunal [RRT]), the legislature will set
out detailed procedural requirements, which may include an entitlement to an
oral hearing. Where this applies, separate (and often difficult) issues may arise,


in particular, the question whether those requirements are exhaustive, or what
is the consequence of any breach.

Application of the hearing rule

The current Australian test for determining whether procedural fairness is
required was formulated by Mason J in Kioa v West.3 Mason J explained:

The law has now developed to a point where it may be accepted that there is a common
law duty to act fairly, in the sense of according procedural fairness, in the making
of administrative decisions which affect rights, interests and legitimate expectations,
subject to the clear manifestation of a contrary statutory intention.

In formulating the threshold test in this way, Mason J reflected, and sum-
marised, the twentieth century expansion of the ambit of procedural fairness,
from decisions affecting rights such as property rights,4 to interests, including
livelihood and reputation,5 and finally to expectations arising from undertakings,
conduct or, in some circumstances, the nature of the application.6 The histori-
cal development of the application of natural justice, or procedural fairness, to
administrative decision-making is traced by Holloway in his study of the High
Court,7 and by other text writers,8 and will not be repeated here. The first signif-
icant step in that development was the move away from the requirement that the
decision maker be under a ‘duty to act judicially’ in addition to having the legal
authority to make decisions affecting rights, which effectively limited procedural
fairness to those bodies engaged in adjudication.9 The decisions of the House of
Lords in Ridge v Baldwin,10 the Privy Council in Durayappah v Fernando,11 and the
Australian High Court in Banks v Transport Regulation Board (Vic)12 opened the
way for recognition, and protection, of interests falling short of rights in the strict
sense. The second significant step was the acceptance of the concept of ‘legitimate
expectation’, which Mason J in Kioa described as arising from some statement or
undertaking by the decision maker, from the nature of the application, or from
the existence of a regular practice.13 As noted below, the concept of legitimate
expectation (at least as an indicator of the application of procedural fairness) is
in retreat. The current focus for the application of procedural fairness is more
generally on the nature of the ‘interest’ affected by an administrative decision.
The range of interests to which procedural fairness will apply is extremely broad,
and includes interests such as status, business and personal reputation, liberty,
confidentiality, and livelihood and other financial interests.14
In the course of the twentieth century expansion of the range of circumstances
in which procedural fairness is required, there was a shift from describing the
duty as a duty ‘to act judicially’, to a duty to accord natural justice, and ultimately
to a duty to comply with the requirements of ‘procedural fairness’. For a period
there was an attempt at characterisation of a ‘duty to act fairly’ as being dis-
tinct from a duty to apply natural justice, and as requiring some (more limited)

form of procedural safeguards where natural justice did not apply.15 In Kioa,
Mason J suggested that the distinction was one of label rather than substance,
and expressed a preference for the term ‘procedural fairness’ for administrative
decision-making, commenting that the term natural justice ‘has been associated,
perhaps too closely associated, with procedures followed by courts of law’.16
However, the label does have implications for content, as Mason J stated that
‘the expression “procedural fairness” more aptly conveys the notion of a flexible
obligation to adopt fair procedures which are appropriate and adapted to the
circumstances of the particular case’.17
It is difficult to entirely separate the ‘implication question’,18 or the ‘thresh-
old test’,19 from the question of the content of procedural fairness. In part, this
is because the broadening of the range of circumstances to which procedural
fairness applies is potentially unlimited. Deane J acknowledged this point in
Haoucher v Minister for Immigration and Ethnic Affairs20 when he explained:

Indeed, the law seems to me to be moving towards a conceptually more satisfying posi-
tion where common law requirements of procedural fairness will, in the absence of a
clear contrary legislative intent, be recognised as applying generally to governmental
executive decision-making . . . and where the question whether the particular deci-
sion affects the rights, interests, status or legitimate expectations of a person in his
or her individual capacity is relevant to the ascertainment of the practical content, if
any, of those requirements in the circumstances of a particular case and of the stand-
ing of a particular individual to attack the validity of the particular decision in those

More recently, in Re Minister for Immigration and Multicultural Affairs; ex parte
Lam,21 McHugh and Gummow JJ approved the following explanation of the
requirements of procedural fairness that McHugh had provided in Teoh’s case:

The rational development of this branch of the law requires acceptance of the view
that the rules of procedural fairness are presumptively applicable to all administrative
and similar decisions made by public tribunals and officials. In the absence of a clear
contrary legislative intention, those rules require a decision maker ‘to bring to a person’s
attention the critical issue or factor on which the decision is likely to turn so that he
may have an opportunity of dealing with it’ (Kioa at 587). If that approach is adopted,
there is no need for any doctrine of legitimate expectations. The question becomes,
what does fairness require in all the circumstances of the case?22

These recent statements by the High Court reflect a more widely growing
dissatisfaction with the use of legitimate expectation as an indicator of the appli-
cation of the requirements of procedural fairness, which stems from Brennan J’s
longstanding objection to the concept of legitimate expectation.23 But expecta-
tions continue to be relevant to the content of the hearing rule. In Lam24 Gleeson
CJ commented that the content of procedural fairness may be affected by what
is said or done during the decision-making process – the relevant factor being
unfairness, and not the disappointment of an expectation.

Leaving aside the debate as to the utility of the concept of legitimate expecta-
tion, it may be that these comments foreshadow a shift away from the Kioa test,
and its focus on the nature and extent of the interests affected, and towards a
presumptive application of the requirements of procedural fairness for all admin-
istrative decision-making. Such a presumption has echoes in the remarks of Lord
Loreburn LC in Board of Education v Rice,25 noting that the duty ‘to act in good
faith and fairly listen to both sides’ was ‘a duty lying upon everyone who decides
anything’. As Creyke and McMillan note, however, while such an approach has the
advantage of ‘apparent simplicity and popularity’, there are difficulties, in par-
ticular, the need to acknowledge some exceptions, for example decision-making
by Cabinet, exercises of prerogative power, decisions of a subordinate legisla-
tive character, and some ‘policy’ decisions.26 Such exceptions could, perhaps, be
excluded from the ambit of judicial review on the broader basis of justiciability.27
Even if not so excluded, the consequence of such an approach might be, as fore-
shadowed by Brennan J in Kioa,28 that if it is assumed that procedural fairness
applies generally in administrative decision-making, in some circumstances the
content of the principles may be diminished, ‘even to nothingness’, to avoid frus-
trating the purpose for which the statutory power in issue was conferred.

Exclusion or limitation of a duty to accord
procedural fairness

The High Court decision in Kioa crystallised debate concerning the source of the
obligation to comply with the requirements of procedural fairness. For Mason J,
the obligation to accord procedural fairness where rights, interests or legitimate
expectations were affected was a common law obligation. Brennan J differed
from Mason J in his analysis of the source of the obligation.29 For Brennan J
the qualification of a statutory power by the requirement to observe procedural
fairness derived from an implied legislative intent, applying ‘to any statutory
power the exercise of which is apt to affect the interests of an individual alone
or apt to affect his interests in a manner which is substantially different from the
manner in which its exercise is apt to affect the interests of the public’.30
The difference in approach to the implication of procedural fairness is more
significant in appearance than in reality. While Brennan J maintained his view
that there was no freestanding common law right to be accorded procedural
fairness, independent of statute, in some of the key High Court procedural fairness
decisions following Kioa, Brennan J and Mason J were in agreement as to the
outcome.31 The High Court decision in Re Refugee Review Tribunal; ex parte Aala32
added a further dimension to the debate, in treating procedural fairness as an
implied condition or limitation on the exercise of statutory power that must be
complied with for a valid exercise of that power.
Regardless of the source of the obligation, the approaches of both Mason J
and Brennan J anticipate the possibility that procedural fairness can be excluded
or limited. In Kioa,33 Mason J accepted that ‘the clear manifestation of a contrary

statutory intention’ could override the common law implication of the require-
ment to accord procedural fairness. Starting from his position that the implication
of procedural fairness was a matter of discerning an implied legislative intent,
Brennan J noted that the presumption of procedural fairness ‘may be displaced by
the text of the statute, the nature of the power and the administrative framework
created by the statute within which the power is to be exercised’.34
The suggestion that for most purposes there is a presumptive application of
the requirements of procedural fairness in administrative decision-making raises
squarely the issue of the possible exclusion or limitation of the requirements of
procedural fairness. Exclusion of a duty to accord procedural fairness can be
express, or arise by implication. Interpretation of the relevant statute is required,
and this interpretation exercise requires consideration of whether the legislature
intends to exclude all the requirements of procedural fairness (in which case
any legislated procedure would operate as code), or only some (and if so, which
The difficulty in finding a clear expression of legislative intent is illustrated by
the decision of the High Court in Re Minister for Immigration and Multicultural
Affairs; ex parte Miah.35 Miah challenged the decision of a delegate of the minister
to refuse his application for a protection visa. The delegate decided that Miah’s
fear of persecution in his country of nationality was not well-founded because of
a change of government which had occurred after the application was lodged.
The delegate did not inform Miah of the intention to rely on this information,
or give him an opportunity to comment on it. Subdivision AB of Division 3 of
Part 2 of the Migration Act 1958 was headed ‘Code of procedure for dealing fairly,
efficiently and quickly with visa applications’. Subdivision AB set out some pro-
cedural requirements, including s54, which provided that a decision could be
made without giving the applicant an opportunity to make oral or written sub-
missions, and s57, which required the decision maker to invite the visa applicant
to comment on information relevant to the decision ‘that is specifically about the
applicant’. There was a right of appeal to the RRT, however the applicant had not
applied for review within the prescribed time limit, and sought constitutional
writs in the High Court under s75(v) of the Constitution.36 Gaudron J noted the
different approaches of Mason J and Brennan J to the implication question, and

. . . if natural justice is a common law duty, the question is whether the provisions of
that subdivision manifest a clear intention that that duty be excluded. On the other
hand, if the rules of natural justice are seen as implied by the common law, the question
is whether the provisions of subdivision AB manifest an intention that that implication
not be made. Whatever approach is adopted, in the end the question is whether the leg-
islation, ‘on its proper construction, relevantly (and validly) limit[s] or extinguishe[s]
[the] obligation to accord procedural fairness’.37

Gleeson CJ and Hayne J noted that there is a difference between a code of
procedure for dealing with visa applications and a comprehensive statement of
the requirements of natural justice. For example, those requirements include the

requirement of absence from bias. However, on their reading of the procedural
requirements of subdivision AB, together with the requirement on the decision
maker to give reasons, and the entitlement to full review on the merits, there
was an intention on the part of the legislature to prescribe the circumstances
in which the applicant was entitled to make submissions or provide additional
information.38 The majority (Gaudron, McHugh and Kirby JJ) held that the Act
did not exclude the application of the common law principles of procedural
fairness. McHugh J commented that ‘the use of the term “code” is too weak a
reason to conclude that Parliament intended to limit the requirements of natural
justice to what is provided in subdivision AB’.39 The majority concluded that the
delegate was required to provide an opportunity to the applicant to put a case by
reference to the change of government.
The response of the government was to move Parliament to amend the Migra-
tion Act, by inserting s51A in subdivision AB:

(1) 51A (1) This Subdivision is taken to be an exhaustive statement of the require-
ments of the natural justice hearing rule in relation to the matters it deals with.
(2) Sections 494A to 494D, in so far as they relate to this Subdivision, are taken to be
an exhaustive statement of the requirements of the natural justice hearing rule
in relation to the matters they deal with.

Similar provisions were inserted in Part 5 Division 5, applying to merits review
by the MRT (s357A), and in Part 7 Division 4, applying to merits review by the
RRT (s422B).
A divergence of views concerning the scope of the phrase ‘in relation to the
matters it deals with’ in s51A and its equivalents soon emerged. Some decisions of
the Federal Court adopted a confined view, treating this phrase as applying only
to the exact text of the procedural requirements set out in the relevant Division
or Subdivision.40 In Moradian,41 for example, Gray J held that the ‘indirect refer-
ences, uncertain inferences or equivocal considerations’ in s51A did not disclose
an intention on the part of the legislature to exclude the principles of procedural
fairness with sufficient certainty. Other judicial decisions adopted a wider view,
identifying a legislative intention that the relevant provisions cover all procedural
aspects of the consideration of the application or the conduct of the review.
A Full Court of the Federal Court has attempted to resolve this divergence in
favour of the wider view, holding in Minister for Immigration and Multicultural
Affairs v Lay Lat42 that what was intended was that subdivision AB ‘provide com-
prehensive procedural codes which contain detailed provisions for procedural
fairness but which exclude the common law natural justice hearing rule’.43 The
same Full Court has applied this reasoning to s422B (and thus, by implication to
s357A).44 This conclusion does not necessarily affect the availability of judicial
review, as procedural irregularities may still be reviewable on the basis that there
has been a failure to properly ‘review’ the decision.45 However, the response of
the courts to these provisions is illustrative of the profound tension between the
courts and executive as to the proper scope and operation of judicial review,

particularly of migration decisions, and the difficulty in discerning ‘the express
terms or necessary implication’46 required to exclude the common law principles.

Exclusion by implication
In some instances, the characterisation of a decision as being ‘preliminary’ or
merely a step (for example, a recommendation) on the path of the decision-
making process will be seen as an indication of an implied intention to exclude
a duty to accord procedural fairness. More commonly, the issue will arise where
there is a right of appeal against the decision. The existence of a right of appeal will
not of itself exclude procedural fairness. In Miah47 McHugh J discussed the factors
to which the courts will have regard. Those factors relate both to the nature of
the original decision – whether it is preliminary or final, private or public, urgent,
and whether there are any specified formal procedures – and to the nature of the
appeal – whether it is to an internal or external (particularly judicial) body, and
whether it is limited or a full de novo rehearing48 – and, ultimately, the nature of
the interest of and consequences for the person affected, and the subject matter
of the legislation. Applying those factors, McHugh J concluded that even the de
novo right of review by an independent tribunal was not sufficient to indicate
an intention to exclude the application of procedural fairness to the delegate’s
An alternative approach is to view the provision of a right to appeal not as an
indication of legislative intent to exclude the requirements of procedural fairness,
but rather as an indication that to the extent that procedural fairness is still
required, judicial review is not the appropriate recourse, and the right of appeal
must be exercised. This explains the difference in reasoning between Barwick CJ
and Jacobs J in Twist v Randwick Municipal Council.49 It made no difference to
the outcome in Twist; however it would be open to the person affected to seek
the intervention of the court (such as by mandamus to direct that a hearing be
given) before the initial decision is made.

Requirements of the hearing rule

The essence of the hearing rule can be simply stated, and is that the person whose
rights, interests or legitimate expectations are likely to be affected by an admin-
istrative decision is given the chance to be heard before the decision is made. The
difficulty comes in identifying what is actually required in the circumstances of a
particular case. Writing extra-judicially, in 1986 Sir Gerard Brennan commented
that ‘[T]he imprecision in the content of “natural justice” and the ex post facto
declaration of that content is one of the unsolved problems of administrative law
and practice’.50 The imprecision to which Sir Gerard Brennan refers is a reflec-
tion of the diversity of administrative contexts in which the hearing rule must
be applied. There is no universal standard, and the requirements of procedural

fairness ‘depend on the circumstances of the case, the nature of the inquiry, the
rules under which the tribunal is acting, the subject-matter that is being dealt
with, and so forth . . . ’51
Writing in 1974, when the English and Australian courts were reinvigorating
the principles and application of natural justice,52 Ganz commented that: ‘The
greatest disservice that administrative lawyers can render administrative law is
to mould the administrative process in their own image.’53 Ganz went on to char-
acterise the rules of natural justice as being ‘modelled on the gladiatorial combat
between two parties before an impartial judge’.54 While the notion of ‘gladiato-
rial combat’ was possibly once an accurate reflection of the judicial process, it is
increasingly less so now. It has never been an appropriate characterisation of the
administrative decision-making process. Ganz’s point is, however, an important
one. While an uncritical adoption of the adversarial paradigm to administrative
decision-making processes is inappropriate, it is not surprising that it is a natural
reference point for the judges whose decisions have framed the content of the
hearing rule. And unless administrative decision-making is regarded as entirely
distinct from judicial decision-making, it is difficult to see why judicial proce-
dures should be completely disregarded in framing the requirements of a fair
An early acknowledgement that the rules formulated for an adversarial judi-
cial process are not appropriate for an administrative process came in Board of
Education v Rice,55 where Lord Loreburn LC characterised the Board of Education
as ‘in the nature of an arbitral tribunal’, and noted that the Board could ‘obtain
information in any way they think best, always giving a fair opportunity to those
who are parties in the controversy for correcting or contradicting any relevant
statement prejudicial to their view’.
Holloway has criticised this part of Lord Loreburn’s reasoning as ‘injecting a
note of “thinness” into the doctrine [of natural justice]’, and more significantly as
providing ‘the lever with which the procedural impositions of natural justice could
be pared back’.56 Holloway includes the other key decision of the early twentieth
century, Local Government Board v Arlidge,57 in this charge, noting an ‘explicit
doctrinal relaxation’ of the obligations associated with procedural fairness.58
Whatever their impact on judicial reasoning during the early twentieth
century,59 viewed some ninety years on, these decisions make sense. It is unlikely
that these Boards have any direct modern equivalent, in terms of either their func-
tions or their constitution.60 However, the acknowledgement that there is both
an irreducible minimum requirement for procedural fairness, and flexibility in
its application, depending on the context, is crucial.
The irreducible minimum requirement of the hearing rule has been expressed
in a number of ways. Lord Loreburn referred to a duty to ‘act in good faith and
fairly listen to both sides’.61 While this formulation captures both the bias and the
hearing requirements of procedural fairness, it reflects more of an adversarial
context than is generally relevant in administrative decision-making. Tucker LJ
referred to one essential, ‘that the person concerned should have a reasonable

opportunity of presenting his case’.62 Creyke and McMillan identify the minimum
requirements as:63
• prior notice that a decision will be made;
• disclosure of an outline or the substance of the information on which the decision
is proposed to be based (that is, a summary of the case that has to be met); and
• an opportunity to comment on that information, and to present the individual’s own
The central issues are how to apply these minimum requirements for the broad
range of administrative decisions which are subject to procedural fairness and
how to mould them according to the different circumstances of each case.
Much of the recent debate concerning the content of procedural fairness has
occurred in the context of review of migration decisions, in particular those of
the merits review tribunals, the RRT and MRT.64 It is significant that despite the
efforts of the Commonwealth Government to restrict judicial review in this area,
it continues to constitute a major part of the federal courts’ administrative law
workload.65 In particular, legislative attempts to preclude judicial review on the
ground of breach of the hearing rules of procedural fairness,66 and more recently
to define the content of the hearing rule,67 have neither limited the number of
applications for judicial review, nor assisted in illuminating the requirements of
the hearing rule.
Judicial review of the decisions of the migration tribunals highlights the
tension between the judicial model of natural justice and broader administra-
tive considerations. The migration tribunals adopt a non-adversarial model of
decision-making, with certain legislated procedural obligations. Possibly the high
point of criticism of the non-adversarial model adopted by the RRT came in
Selliah v Minister for Immigration and Multicultural Affairs,68 when the court
stated that:
. . . hearings before the Tribunal are virtually unique in Australian legal procedures
and in the common law system generally. They are not interrogatory or adversarial
proceedings: Abebe v The Commonwealth of Australia [1999] HCA 14, lawyers are gen-
erally absent, and the appropriateness of providing an interpreter is apparently within
the sole discretion of the Tribunal. The Tribunal will normally not be proficient in the
first language of the asylum seeker and will generally know nothing of the culture or
practices of the person’s country of origin or the idiom of its language. The Tribunal is
both judge and interrogator, is at liberty to conduct the interview any way it wishes,
without order, predictability, or consistency of subject matter, and may use any out-
side material it wishes without giving the person being interrogated the opportunity of
reading and understanding the material before being questioned about it. Moreover,
the Tribunal has been known to rely on supposed inconsistencies in the factual account
being given without stating the alleged contradictions to the interviewee and giving
the person an opportunity to explain them.
These methods contravene every basic safeguard established by our inherited system
of law for 400 years . . .

While the court was disturbed about these features of the legislation, the power
of the legislature to enact such restrictions was not disputed. More recently, in

Applicant VEAL of 2002 v Minister for Immigration and Multicultural and Indige-
nous Affairs69 Gleeson CJ, Gummow, Kirby, Hayne and Heydon JJ described the
functions of the RRT in the following terms:

The Tribunal was not an independent arbiter charged with deciding an issue joined
between adversaries. The Tribunal was required to review a decision of the Executive
made under the Act and for that purpose the Tribunal was bound to make its own
inquiries and form its own views upon the claim which the appellant made. And the
Tribunal had to decide whether the appellant was entitled to the visa he claimed.

Not only is the tone less strident, but there appears to be an acceptance that
the process is not necessarily defective because it fails to mirror an idealised
judicial model: it is simply different. In part this may reflect the concern of the
High Court as to the constitutional dimensions relevant to the federal tribunals.
In NAIS v Minister for Immigration and Multicultural and Indigenous Affairs,70
Gummow J noted that the adoption of the paradigm of judicial processes of
decision-making by the federal tribunals was ‘rarely helpful because it is apt to
blur the constitutionally entrenched distinctions between judicial and executive
power’. It is important not to make too much of this distinction, which is, in
any event, relevant primarily to Commonwealth tribunals. Even though the
migration tribunals are ‘carrying out an administrative function on behalf of
the Executive’, there is an irreducible requirement that they adopt procedures
that are fair, ‘so that “the practical requirements of fairness” appropriate for the
application of the rule of law are observed’.71
Of course, the non-adversarial tribunals are only one form of administrative
decision maker. The High Court has made it clear, however, that tribunals sit in
the non-judicial camp for the purposes of jurisdictional defects and excesses.72
It is interesting to note that while the courts have acknowledged that the bias
rule may apply less stringently to ministers than to courts or tribunals,73 there
has not been a similar consideration given to the application of the hearing rule,
other that in acknowledging that application and content may vary according to
the nature of the decision.74
Aronson, Dyer and Groves argue that, while the courts have acknowledged
that it is inappropriate to expect administrative decision makers to follow curial
procedures, the procedures imposed by the common law are shaped by the values
and assumptions of the adversarial tradition.75 A recent study of inquisitorial
processes in tribunals has identified as an issue the location of the tribunal system
within a system where statutory appeal or judicial review are conducted by courts
which operate in an adversarial manner.76 That study flagged two key concerns
for tribunals which are equipped with investigative powers and which adopt an
inquisitorial role, which do not arise for those tribunals which operate in an
adversarial model: avoiding a perception of lack of impartiality in the course of
eliciting information and testing evidence, and identifying the extent to which the
tribunal is required to inquire further than the material or information provided
by an applicant.77

The central requirements of the hearing rule were outlined by the Federal
Court in Commissioner for Australian Capital Territory Revenue v Alphaone Pty
Ltd.78 The court explained:

Where the exercise of a statutory power attracts the requirement for procedural fair-
ness, a person likely to be affected by the decision is entitled to put information and
submissions to the decision maker in support of an outcome that supports his or her
interests. That entitlement extends to the right to rebut or qualify by further informa-
tion, and comment by way of submission, upon adverse material from other sources
which is put before the decision maker. It also extends to require the decision maker to
identify to the person affected any issue critical to the decision which is not apparent
from its nature or the terms of the statute under which it is made. The decision maker
is required to advise of any adverse conclusion which has been arrived at which would
not obviously be open on the known material. Subject to these qualifications however,
a decision maker is not obliged to expose his or her mental processes or provisional
views to comment before making the decision in question.

The distinction between identifying critical issues and exposing the decision
maker’s ‘mental processes or provisional views’ is often difficult to draw. The
courts do not require decision makers to give the person affected ‘a running
commentary’ on the prospects of success,79 or to put to the person concerns
which may be inclining the decision maker towards an adverse finding.80 There
may, however, be circumstances in which a decision maker is required to alert
the person affected of the possibility of an adverse conclusion,81 and the decision
maker must be careful not to mislead the person affected, for example, by allowing
an implication to arise that a document has been accepted as genuine.82 While
the decision maker is not required to make an applicant’s case, the decision maker
may in some instances be obliged to identify for the affected person the issues
that need to be addressed. As McHugh J put it in Minister for Immigration and
Ethnic Affairs v Teoh83 the obligation is to bring to a person’s attention ‘the critical
issue or factor on which the administrative decision is likely to turn so that he
may have an opportunity of dealing with it’.
This can be problematic for those non-adversarial tribunals, such as the tri-
bunals created to determine migration issues, and the Social Security Appeals
Tribunal, where the respondent decision maker plays no active role in relation
to the presentation of evidence. In addition to the danger of creating a percep-
tion of possible prejudgment, there are significant resource and efficiency issues
in imposing obligations to assist an applicant to understand what the critical
issues are. In this context, it is worth noting that the Administrative Decisions
Tribunal of New South Wales has a statutory obligation to take such measures as


ńňđ. 11
(âńĺăî 19)