Reasonableness, Proportionality and Merits Review

The Hon. Justice Garry Downes AM

Paper delivered to the New South Wales Young Lawyers Public Law CLE Seminar, The Law Society, Sydney
24 September 2008


On 10 November 1947 Lord Greene MR said something which every lawyer and law student knows. The case in which he said it was not, in itself, significant. The decision was hardly reserved. The reasons cover less than eight pages. The other Court of Appeal judges agreed. Lord Greene said this:

"It is true that discretion must be exercised reasonably. Now what does that mean? ... [T]here may be something so absurd that no sensible person could ever dream that it lay within the powers of the authority. Warrington LJ in Short v Poole Corporation (1) gave the example of the red-haired teacher, dismissed because she had red hair. That is unreasonable in one sense. In another sense it is taking into consideration extraneous matters. It is so unreasonable that it might almost be described as being done in bad faith; and, in fact, all these things run into one another." (2)

A little later he said:

"It is true to say that, if a decision on a competent matter is so unreasonable that no reasonable authority could ever have come to it, then the courts can interfere. That, I think, is quite right; but to prove a case of that kind would require something overwhelming, and, in this case, the facts do not come anywhere near anything of that kind." (3)

He repeated the famous phrase once more (4)  before concluding his judgment.

With these words, perhaps partly because of their colourfulness, a ground of review of administrative decision-making was settled for the common law.

In Germany, in the 1870's, a ground of review called Verhaltnismassigkeit was developed. We would call it proportionality, though it was more than that. The principle was adopted in French law in the 1970's. It became a principle applied by the European Court of Human Rights and is supported by the European Convention for the Protection of Human Rights.

The Administrative Appeals Tribunal Act 1975 (Cth) came into force on 1 July 1976. On that day, for the first time in the world, review of administrative decisions on their merits became available in a forum exercising general jurisdiction.

This evening I wish to analyse these three bases for review of administrative action and to make some observations about them which are relevant to Australia.


Lord Greene, and Warrington LJ before him, described a ground of review which would rarely be attracted. The plaintiffs were unsuccessful in both Wednesbury and Poole Corporation. Over the years, however, the strictness of the rules came to be relaxed, both in England and Australia. I remember many times at the bar telling judges that "however narrow the test seems to be, it is surprising how often judges hold that it is satisfied". I was successful, for example, before Bergin J in Ziade v Randwick City Council (5) in which I remember using those very words. Her Honour held that a Council decision to restrict parking to residents, in an area adjacent to a busy cinema, was unreasonable in the Wednesbury sense. The same result might not occur today. Bergin J was able to find:

"I am satisfied that the May resolution was in breach of the Guidelines and therefore invalid and of no force or effect. Alternatively I am satisfied that the May resolution was made in a manner which failed to provide procedural fairness to the plaintiff. I am also satisfied that the May resolution was one which, on the material available to the defendant, no reasonable person would have made." (6)

This was one of the more comprehensive victories I had at the bar!

The liberal tendency has continued in the United Kingdom, but not in Australia.


The Wednesbury principle was conclusively adopted in Australia by the High Court in Parramatta City Council v Pestell. (7)  The judgment of Gibbs J says it most clearly. The relevant passage also illustrates a particular aspect of the principle. Gibbs J said:

"If, in purporting to form its opinion, a council has taken into account matters which the Act, upon its proper construction, indicates are irrelevant to its consideration, or has failed to take into account matters which it ought not to have considered, the opinion will not be regarded as validly formed. Even if the council has not erred in this way an opinion will nevertheless not be valid if it is so unreasonable that no reasonable council could have formed it (see Associated Provincial Picture Houses Ltd v Wednesbury Corporation, (8) and see also Bankstown Municipal Council v Fripp. (9) "  (10)

The passage illustrates the overlap of the reasonableness ground with other grounds of judicial review, particularly the grounds of acting on irrelevant, or not acting on relevant, considerations. It also supports the inference that the ground adds something to, or is wider than, the more precise grounds.

This is a theme which Mason J explained in Minister for Aboriginal Affairs v Peko-Wallsend Ltd (11) where he suggested that in cases in which the legislation did not specify the matters to be taken into account the preferable form of review was by reference to the unreasonableness principle. It may be taken that at this time the principle was still in its ascendancy.

In Re Minister for Immigration and Multicultural Affairs ; Ex p. Applicant S20/2002 (12)  the broader view of the unreasonableness principle suffered a set-back. Gleeson CJ, adopting a contrary position to Mason J, placed emphasis on identifying the precise legal principle said to have been offended against:

"As was pointed out in Minister for Immigration and Multicultural Affairs v Eshetu, (13) to describe reasoning as illogical, or unreasonable, or irrational, may merely be an emphatic way of expressing disagreement with it. If it is suggested that there is a legal consequence, it may be necessary to be more precise as to the nature and quality of the error attributed to the decision-maker, and to identify the legal principle or statutory provision that attracts the suggested consequence." (14)

McHugh and Gummow JJ confined the unreasonableness principle to the exercise of statutory discretions. It did not apply to fact-finding. (15)  Kirby J appears to have accepted this limitation. (16)

The present mood in Australia appears to limit the principle both generally (by confining it to something equating to its original strictures) and specifically (by confining it to the exercise of a statutory discretion). I want to concentrate for a moment on the first of these limitations.

Professors Creyke and McMillan in Control of Government Action: Text, Cases and Commentary (17)  have identified three pieces of supporting evidence. First (contrary to my habitual submission), there are not many decided cases in which the principle has been successfully raised. Secondly, it usually overlaps with a more precise remedy. Thirdly, it needs to be applied against the significant division between merits review and judicial review.

In Attorney-Gereral (NSW) v Quin (18) Brennan J noted that "Wednesbury unreasonableness leaves the merits of a decision or action unaffected unless the decision or action is such as to amount to an abuse of power (19) ". (20) Murray Gleeson AC has distinguished between "review of the merits of administrative decisions … and judicial review based upon principles of legality." (21)

United Kingdom

In the United Kingdom things have gone the other way. Oddly enough, the way was led by a New Zealand judge who, I believe, is the only judge born and educated in a Commonwealth country who ultimately became a full Lord of Appeal in Ordinary. Lord Cooke of Thorndon was President of the New Zealand Court of Appeal from 1986 to 1996. Sitting in the House of Lords, however, after describing Wednesbury as "apparently briefly-considered", he referred to its famous proclamation as a "tautologous formula" representing an undesirable "admonitory circumlocution". (22)  He preferred the simple test "whether the decision in question was one which a reasonable authority could reach." (23)  Lord Cooke also took the view that European concepts of proportionality would produce the same result as the application of English principles of reasonableness. (24)

In R v Ministry of Defence; Ex p. Smith (25)  the Court of Appeal (Bingham MR, Henry and Thorpe LJJ) accepted two propositions. First, the test of unreasonableness was whether "the court is satisfied that the decision is unreasonable in the sense that it is beyond the range of responses open to a reasonable decision-maker." (26)  Secondly, because any human rights context was important, "[t]he more substantial the interference with human rights, the more the court will require by way of justification before it is satisfied that the decision is reasonable…". (27)  The question was whether possessing a homosexual orientation could justify dismissal from the armed forces. The Court decided, reluctantly, I think, and partly because the policy was supported by Parliament, that the practice could not be determined to have the requisite degree of unreasonableness.

Smith was decided shortly before the European Convention on Human Rights (28)  was enacted into law binding the Courts in the United Kingdom. Article 8 created limitations on interference with a citizen's "right to respect for his private and family life, his home and his correspondence." (29)  The Court declined to address this article, however, because at the time of the decision the article did not have the force of law in the United Kingdom.

There was, however, a right to seek review in the European Court of Human Rights where the Convention did apply and to which there was an appeal in Smith. That court passed over the English doctrine of reasonableness, apparently finding it to be unsatisfactory, and applied the European doctrine of proportionality. It upheld the claim and awarded damages (Lustig-Prean and Beckett v The United Kingdom (30) ). Rejecting the approach of the Court of Appeal and applying the principle that an interference with a human right protected by Art 8 will be acceptable only if the interference "answers a pressing social need and, in particular, is proportionate to the legitimate aim pursued", (31) the Court found a breach of Art 8 to be made out.



The adoption of a proportionality test for the validity of administrative action first emerged in Germany in the 1870's. The German word is Verhaltnismassigkeit which literally means relativity. Three tests are applied, only one of which involves proportionality as such. The three tests are:

  1. The measure proposed must be suitable for the purpose;

  2. The measure must be necessary; and

  3. The measure must not be disproportionate.

Proportionality emerged in French Law in the 1970's. It was welcomed first by French academic and member of the Conseil d'Etat Guy Braibant. The French Conseil d'Etat (Council of State) is France's highest court for matters of administrative law.

The principle was applied by the Conseil d'Etat in 1972 to strike down the conversion of a road to a pedestrian precinct. (32) The measure was disproportionate to the need. The principle is now well entrenched in French law where it is associated with a concept of gross error in fact finding and is often explained by reference to a balance or balance sheet (le bilan).

United Kingdom

It must be fair to say that the decision of the European Court of Human Rights in Lustig-Prean softened up the English courts for proportionality. The courts appeared to be ready for it, however, even earlier. Some of the remarks in the Court of Appeal in Smith/Lustig-Prean, and even more clearly, in the Divisional Court in the case, are clear pointers in that direction.

The suggestion that the principle might be adopted in English law goes back to 1985 when Lord Diplock raised the possibility directly (Council of Civil Service Unions v Minister for the Civil Service (CCSU) (33)). That started the debate. The debate was fuelled by the appreciation that proportionality was accepted in Community law. Although European human rights law, where it was found, was only binding in the United Kingdom on review in the European Court of Human Rights, that was still a significant matter. Moreover, it was recognised that the principle would become even more compelling when the European Convention on Human Rights came to bind English courts.

The principle of proportionality now sits alongside unreasonableness in its application in England (R (Daly) v Home Secretary (34)).  Its use is generally confined, however, to human rights or similar issues. Lord Hope has explained the principle in a way which accords very much with the German formulation, namely justification, fairness and proportionality. A measure claimed to erode human rights must satisfy all three.  (35)  As Lord Steyn said in Daly,it is unlikely that the result will now be different in English law whether the principle applied be proportionality or some other ground of review such as reasonableness. (36)


Proportionality has been recognised as such in Australia. The High Court has used it to determine the validity of subordinate legislation (South Australia v Tanner (37) ). Deane J, referring to CCSU, tentatively associated it with grounds for judicial review in Australian Broadcasting Tribunal v Bond. (38)  Spigelman CJ (39)  and others have recognised the relevance of proportionality to reasonableness but Spigelman CJ stated that it was not a separate ground of review. (40)

Sir Anthony Mason (41)  and the late Brad Selway (42) show no enthusiasm for any separate doctrine.

In a way the question is still open in Australia. However, I think that a number of factors point against the likelihood of its adoption as a separate ground. First, it seems to be associated with a broadening of the grounds of review which is an approach that has not been adopted in Australia. Secondly, its adoption in the United Kingdom is closely associated with its geographical and judicial proximity (at least on questions of human rights) with Europe. Thirdly, Australia has a sophisticated system of merits review of administrative decisions which has been in place for more than 20 years and the need to expand judicial review is not a present concern.

Merits Review

What is merits review? The conventional answer is that it is review which is wider than correcting legal error. It extends to a reconsideration of discretionary matters - of the merits of the original decision. The process is often described by saying that the Tribunal must reach "the correct or preferable" decision. (43)  The description, however, glosses over two important components which go to make up merits review. The first component is that the reviewing tribunal considers the merits of the issue; but the second component is that it substitutes its decision for the decision under review. Merits review and judicial review can undoubtedly overlap but the power of courts to substitute a decision is much more limited than is the power of merits review tribunals.

This point was brought home to me at a meeting of an international association of administrative courts and tribunals in Madrid a few years ago. I was explaining the Australian system and suggesting that it was unique. I referred to reconsideration of the merits of decisions. My audience was unimpressed. Then I referred to the substitution of our decision for the decision under review. That made them sit up. "You mean you can make a fresh decision?", they said. "Yes", I said. "Unheard of", they replied.

The essence of merits review is the power to substitute a decision. This is so both as a matter of substance and as a matter of form. It is so, as a matter of substance, because substitution implies the power to address all issues and leaves the reviewing tribunal, as the ultimate decision-maker, in no way bound by what has gone before. It is so, as a matter of form, because that is what s 43 of the Administrative Appeals Tribunal Act says. Section 43, which has been adopted for state administrative review tribunals, says nothing about discretion or merits. It simply empowers the Tribunal to set aside the decision under review and substitute its own decision. That is the sole source of power to consider the merits. Section 16 of the Administrative Decisions (Judicial Review) Act 1977 (Cth), by contrast, does not permit any remaking of the decision under consideration.

Very few administrative law systems permit substitution and never by a body with general jurisdiction, apart from Australia. Specialised tribunals in common law countries such as the United Kingdom, Canada and New Zealand do have the power, but the jurisdiction is necessarily confined. One of the almost accidental consequences of the establishment of a general administrative appeals tribunal in Australia is the proliferation of its jurisdiction so that it now has very wide jurisdiction, under more than 400 acts of the Commonwealth Parliament. The mere existence of a general merits review tribunal has promoted growth in the matters which are subject to its review. The Parliament does not have to look for, or create, a new tribunal, with consequent trouble and expense, when it introduces new legislation affecting rights - one is already available.


Much has been said about the relative positions of reasonableness and proportionality as grounds of judicial review. They have been said to overlap. (44)  Although it is a more complicated test, proportionality may be a more lenient test in operation. For example, although the decision in Smith/Lustig-Prean (45)  was held not to be unreasonable, it was held to be disproportionate. I have not, however, seen any comparative analysis of the content of the two principles other than a recognition that proportionality necessarily operates in a narrower context because a proposed measure must be assessed against a need. (46)

Reasonableness covers a wider field than proportionality. A measure might be proportionate, but its adoption unreasonable, because, for example, of a lack of consultation. It is not clear to me that the tests are the same, or even that they will yield the same result. They are quite different tests. One is a test of rationality; the other is a relationship test. One is an overall and general test; the other is a precise test applied negatively to a previously identified relationship. I expect that a disproportionate solution will almost always be unreasonable; but the reverse is not necessarily true. As I said in McKinnon v Secretary, Department of Treasury, (47)  following the High Court in George v Rockett, (48)  reasonableness, for a decision, requires "the existence of facts which are sufficient to induce that state of mind in a reasonable person". (49)  Proportionality requires a judgment of the relationship between an end, amounting to a need, and a means to satisfy the end. This explains the three stage tests. The judgment is whether the means is proportionate to the need - whether the end justifies the means.

I can see nothing inherent in the two tests which means that one might be more liberal as an administrative law test than the other. If there is a qualitative difference it seems to me that it must be in the way the tests are enunciated. A test of simple reasonableness might be no less liberal than a test of proportionality. However, a test of reasonableness, which is not made out if it satisfies a simple balance, but is required to be a decision which no rational person could ever arrive at, seems to be a less liberal test than one requiring a simple finding that the subject matter of the decision is disproportionate to the need. I have the feeling that a test of proportionality also contains some greater level of subjectiveness. Reasonableness does require some benchmarks even if the selection of the benchmark is itself very subjective. It has been said a number of times that proportionality clearly requires a benchmark (50)  although what the benchmark is has not been authoritatively specified. If the distinctions I have drawn between the two tests are valid they must, if anything, be more valid in Australia, where the strict test of reasonableness still applies, than in the United Kingdom.

What are the reasons which have led to Australia largely rejecting a test of proportionality, which applied alone may be more lenient than the strict test of reasonableness, and affirming the strict test? I think that the primary reason must be the presence in Australia of a sophisticated and comprehensive process of merits review. Australia has rejected a doctrine of deference to the decisions of administrative decision-makers. (51)  However, Australia seems to me to have adopted a subtler approach of limiting interference under the main ground of judicial review which overlaps with merits review, namely reasonableness, for the reason that in most cases merits review will better and more directly deal with complaints.


Reasonableness as a ground of judicial review of administrative action is well established in Australia. It provides a safety valve permitting truly irrational decisions to be set aside, but not replaced. Proportionality can be an element of such a claim, but it will not lead to a decision being set aside unless it is both disproportionate, in particular, and extremely unreasonable, in general. The gap in the law in this area, between the United Kingdom and Europe on the one hand, and Australia on the other, is bound to remain and even widen as the more liberal test of general disproportionality erodes the stricter aspects of reasonableness in the United Kingdom to the point at which the test of reasonableness may reach the state, if it has not already reached it, that the test is one of simple unreasonableness. In Australia, the role performed by proportionality, and more, will be the function of merits review.


  1. [1926] Ch 66 at 90-91.

  2. Associated Provincial Picture Houses Ltd v Wednesbury Corporation [1948] 1 KB 223 at 229 per Lord Greene MR.

  3. Id at 230.

  4. Id at 234.

  5. (2001) 51 NSWLR 342.

  6. Id at 375.

  7. (1972) 128 CLR 305.

  8. Wednesbury, above n2 at 228-229 and 233-234.

  9. (1919) 26 CLR at 403.

  10. Pestell, above n7 at 327 per Gibbs J.

  11. (1986) 162 CLR 24 at 41.

  12. (2003) 198 ALR 59.

  13. (1999) 197 CLR 611 at 626 [40]; 162 ALR 577 at 587 per Gleeson CJ and McHugh J.

  14. S20, above n12 at 61 per Gleeson CJ.

  15. Id at 76.

  16. Id at 87.

  17. Creyke, R., and McMillan, J., Control of Government Action: Text, Cases and Commentary, LexisNexis Butterworths, Sydney, 2005, pp. 724-741.

  18. (1990) 170 CLR 1.

  19. Nottinghamshire County Council v Secretary of State for the Environment [1986] AC 240 at 249.

  20. Quin, above n18 at 36 per Brennan J.

  21. Gleeson, M., ‘Judicial Legitimacy' (2000) 20 Australian Bar Review 4 at 11.

  22. R v Chief Constable of Sussex; Ex Parte International Trader's Ferry Ltd [1999] 2 AC 418 at 452 per Lord Cooke.

  23. Ibid.

  24. Ibid.

  25. [1996] QB 517.

  26. Id at 554.

  27. Ibid.

  28. Convention for Protection of Human Rights and Fundamental Freedoms1950.

  29. Id, Art 8(1).

  30. [1999] ECHR 71; [2000] ECHR 382.

  31. Lustig-Prean [1999] ECHR 71 at [80] per Costa P, Bratza, Loucaides, Kuris, Fuhrmann, Greve and Traja JJ, and Ms S. Dolle.

  32. Ville de Dieppe [1972] CE 8.

  33. [1985] AC 374 at 410 per Lord Diplock.

  34. [2001] 2 AC 532.

  35. R v Shayler [2003] 1 AC 247 at 281 per Lord Hope.

  36. Daly, above n34, at 547 per Lord Steyn.

  37. (1989) 166 CLR 161.

  38. (1990) 170 CLR 321 at 367 per Deane J.

  39. Bruce v Cole (1998) 45 NSWLR 163 at 185 per Spigelman CJ.

  40. Ibid.

  41. Mason, A.,‘The Scope of Judicial Review' (2001) 31 AIAL Forum 21 at 38.

  42. Selway, B., ‘The Principle behind Common Law Judicial Review of Administrative Action - the Search Continues' (2002) 30 Federal Law Review 217 at 224.

  43. Drake v Minister for Immigration and Ethnic Affairs (1979) 24 ALR 577 at 589 per Bowen CJ and Deane J.

  44. See, for example, Daly, above n34 at 547 per Lord Steyn; Tanner, above n37 at 167-168 per Wilson, Dawson, Toohey and Gaudron JJ; Bond, above n38 at 367 per Deane J; and Mason, A., ‘The Scope of Judicial Review', above n41 at 38.

  45. Smith, above n 25; Lustig-Prean, above n 30.

  46. Lustig-Prean, above n31 at [80] per Costa P, Bratza, Loucaides, Kuris, Fuhrmann, Greve and Traja JJ, and S. Dolle; Daly, above n34 at 548 per Lord Steyn; and Shayler, above n35 at 281 per Lord Hope.

  47. (2004) 86 ALD 138.

  48. (1990) 170 CLR 104.

  49. McKinnon, above n47 at 143 per Downes J citing George, Id at 112 per Mason CJ and Brennan, Deane, Dawson, Toohey, Gaudron, and McHugh JJ.  

  50. See, for example, M. Aronson, B. Dyer, and M. Groves, Judicial Review of Administrative Action (3rd ed.), Lawbook Co., Sydney, 2004, p. 345, and P. Johnston, ‘Proportionality in Administrative Law: Wunderkind or Problem Child?' (1996) 26 Western Australian Law Review 138 at 151-152.

  51. City of Enfield v Development Assessment Commission (2000) 199 CLR 135; cf. cases on the deference doctrine as applied in Canada (Canadian Union of Public Employees, Local 963 v New Brunswick Liquor Corporation [1979] 2 S.C.R. 227; U.E.S., Local 298 v Bibeault [1988] 2 S.C.R. 1048; Baker v Canada (Minister of Citizenship and Immigration) [1999] 2 S.C.R. 817) and the United States (Chevron U.S.A., Inc. v Natural Resources Defense Council, Inc. 467 U.S. 837 (1984); United States v. Mead Corp., 533 U.S. 218 (2001)).