Department of Agriculture and Rural Affairs v. Binnie

[1989] VR 836[1989] VR 836

---

(Judgment by: Marks J) Court:

Judges: Young CJ

Marks JTeague J

Judgment date: 9 December 1988

---

Judgment by:
Marks J

This appeal concerns the meaning of "reasonably likely" in s31(1) of the Freedom of Information Act 1982 (the Act).

The respondent, who is concerned with the welfare of animals, obtained a decision on 20 March 1987 from the Administrative Appeals Tribunal (the tribunal) granting him access to some 1600 "Returns of Animal Usage for Experimental Purposes" (the returns) for the period 1984-85 submitted by registered experimenters in accordance with the requirements of the Protection of Animals Act 1966 and its regulations.

The issues before the tribunal arose out of the concern on the one hand of the respondent for the welfare of animals and on the other of medical scientists who experiment on different species of animate life.

The Protection of Animals Act 1966 purports to regulate such experiments by requiring those who perform them to be registered and submit returns: s12(3).

The appellant (the department) had provided the respondent with a list of institutions at which there were registered experimenters and a booklet entitled "Statistics of Final Experimentation, Victoria--June 1985" for the years 1983-84 and instructions on the method of completing animal experimentation returns. The department also told the respondent that he could have access to the returns for the same period provided the name and signatures of the individual experimenters and the names of the institutions were deleted.

The respondent agreed to the deletion of the names and signatures individual experimenters but not to the deletion of the names of the institutions where the experiments were carried out.

In the upshot, the respondent would be informed generally of the institutions where experiments were done and what they were. He might not readily be able to identify what experiments were performed at any particular institution, nor who performed them. If, however, he were told what institution performed particular experiments he might readily be able to identify the group (for example, those registered at that institution) to which an experimenter belonged, if not the particular person.

The issue before the tribunal was only whether the respondent should have access to the returns with or without the names of the institutions and where appropriate, of their departments. Before this court, we were told by the counsel that the issue did and does also involve the deletion of the name the department.

Two institutions, Monash University and The Royal Melbourne Hospital who had filed returns were legally represented before the tribunal and consented to the release of returns relating to them provided the names their relevant departments were deleted.

Thus the tribunal gave its directions for release of all 1600 returns subject to the deletion of the names and signatures of the individual experimenters and, in the case of Monash University and The Royal Melbourne Hospital, the names of departments. It is curious that the latter deletion was not directed in respect of any other institution notwithstanding its non-appearance.

The appeal of the department against the decision of the tribunal to a single judge of this court was dismissed on 23 September 1987. The department now appeals to this court.

In order to have succeeded before the tribunal, the department was required to establish that the returns were exempt documents within the meaning of s31(1)(e) of the Act which reads:

"31.(1)

Subject to this section, a document is an exempt document if its disclosure under this Act would, or would be reasonably likely to

--

(e)

endanger the lives or physical safety of persons engaged in or in connection with law enforcement or persons who have provided confidential information in relation to the enforcement or administration of the law."

There was evidence before the tribunal that only registered experimenters were permitted to perform experiments on animals, that they were required by law to send bi-annual returns to the animal welfare bureau within the department setting out in detail the nature, purpose and methods used in experiments. The freedom of information manager of the department said that he wrote to 48 institutions about the request of the respondent for access to the returns. Only 21 replied, of which 13 expressed opposition to their release through fear of physical harm to the experimenters. This apprehension and fear of identification were said to explain their non-appearance before the tribunal.

A Professor Short of the Department of Physiology at Monash University told the tribunal that he personally and his department had received bomb threats after he had appeared in June 1985 on television in a discussion about animal experimentation. He was given police protection. He spoke also of bomb threats to experimenters in Western Australia, violence in the United Kingdom and the release of mice after a break-in at the Melbourne University. He said that the latter was very dangerous especially if the mice were infected.

The respondent, who appeared in person, asked Professor Short in cross-examination if it were true that Monash University, as part of an experiment, had cemented possums by the head to iron bars and sand-blasted their faces. He answered that it was nonsense.

In his submissions to the tribunal the respondent said he represented 25,000 members of an organisation called "Australians for Animals", that little or no impact would be made by "simply saying that a particular experiment was conducted in one of 60 institutions" and he wanted to be able to "target" particular institutions that carried out experiments that cause suffering to animals.

It was not in dispute before us that the returns were "confidential information in relation to the administration of the law" within the meaning of s31(1)(e).

The essential contention of the department is that the tribunal, and consequently the learned judge who dismissed the appeal, were wrong as to the meaning they attributed to the expression "would be reasonably likely" in s31(1).

It was also submitted that the tribunal did not give reasons as required by s49(3) of the Administrative Appeals Tribunal Act 1984.

For the moment, I concern myself with the main submission regarding s31(1).

In the reasons for decision the tribunal said: "I accept the evidence Professor Short concerning the threat against him and the university and that the threats were in relation to animal experimentation. It is apparent from his evidence that there is a possibility that disclosure of the returns would endanger the lives or physical safety of experimenters. It appears that some types of experimentation, particularly on certain animals, attract more adverse public attention than others. However, the subsection uses the words `would or would be reasonably likely to . . .'. `Likely' is defined in the Shorter Oxford English Dictionary as meaning `seeming as if it would happen, or prove to be as stated; probable'. [See R v Boughey (1986) 20 A Crim R 156 for discussion of the meaning of `likely' and `probable' as distinct from `possible']. The evidence and material before me does not permit me to find that disclosure would or would be reasonably likely to endanger the lives or physical safety of experimenters."

If, as appears from the above to be the case, the tribunal founded it understanding of the expression "reasonably likely" in s31(1) on the dictionary definition mentioned and what was said by the High Court of the word "likely" in Boughey v R (1986) 161 CLR 10 , I think it was mistaken. The definition does not, as will be seen, cover all the relevant meanings. Boughey does not, nor purports to, provide an appropriate guide for the present case. The High Court had before it an application for leave to appeal against a conviction for murder on the ground of misdirection of law to the jury by the trial judge. The applicant was a qualified Tasmanian medical practitioner who had applied manual pressure to the carotid arteries of his mistress, as had been his habit, for sexual arousal. There was medical evidence that this practice caused her death. S157(1) of the Criminal Code (Tasmania) provided that "culpable homicide" was murder if committed, inter alia, "(c) by means of an unlawful act or omission which the offender knew, or ought to have known, to be likely to cause death in the circumstances, although he had no wish to cause death or bodily harm to any person."

Thus the High Court discussed the meaning of "likely" in the context of the proof which the prosecution was obliged to make of an element of the crime of murder.

An examination of the reasons for judgment shows that although the word "likely" was considered to mean "probable" in the context of the Tasmanian legislation, it was not held to mean "a greater than 50 per cent chance" or "more likely than not" (see 161 CLR, at 21) but rather a substantial "real and not remote" chance regardless of whether it is less or more than 50 per cent (see, at 21).

At 14, Gibbs CJ observed: "It is trite to say that the meaning of a word will be influenced by the context in which it appears. In my opinion the word `likely' in s156 and s157 of the Criminal Code Act means `probable' and not `possible'."

At 20 Mason, Wilson and Deane JJ in their joint judgment said: "It is true that the meaning of the words `probable' and `likely' is liable to vary according to the context in which they are used: see Tillmanns Butcheries Pty Ltd v Australasian Meat Industry Employees' Union (1979) 42 FLR 331 , at 346-347; 27 ALR, at 380-382; Australian Telecommunications v Krieg (1976) 14 SASR 303 , at 311; Koufos v C Czarnikow Ltd [1969] 1 AC 350 , at 410-411."

The reasons of the tribunal do not include any articulated consideration of the important qualifications expressed by members of the High Court in Boughey that the relevant word must be considered in its appropriate context and in that case the word "likely" was not the same as "more likely than not" even when considered in the context of a criminal statute concerning murder. It seems the tribunal took "likely" to be synonymous with "probable", and it is not clear whether it understood that the High Court did not necessarily mean "a greater than 50 per cent chance" or "more likely than not".

It is more important that the tribunal appears to have ignored the word "reasonably" which qualified the word "likely" in s31(1) of the Act. In my opinion, the expression, "reasonably likely" has a different connotation from "likely" on its own. It does not require establishment of a more than 50 per cent chance.

Before further discussion of the expression "reasonably likely", I observe that the word "likely" may well vary in meaning according to its context. There can be little doubt that the ordinary use of the word "likely" is synonymous with "probable", that is, there being more than a 50 per cent chance of the event happening. In this regard, useful observations were made by Bray CJ in Australian Telecommunications Commission v Krieg Enterprises Pty Ltd (1975) 14 SASR 303 . However, Deane J, along with other members of the Full Federal Court, in Tillmanns Butcheries Pt, v Ltd v Australasian Meat Industry Employees' Union (1979) 42 FLR 331 ; 27 ALR 367 interpreted "likely" in the context of s45D(1) of the Trade Practices Act 1974 (Cth) not to mean "more than a 50 per cent chance" or "an odds-on chance", the latter expression being that of Lord Hodson in Koufos v C Czarnikow Ltd [1969] 1 AC 350 , at 410. In Sheen v Fields Pty Ltd (1984) 58 ALJR 93 ; 51 ALR 345 the High Court held that "likelihood" in r1, CL21 of the rules made under the Factories and Shops Act 1960 (Qld) meant "a real or not remote" chance or possibility "regardless of whether it was less or more than 50 per cent" and followed the observations of Deane J in Tillmanns Butcheries . Speaking of the word "likely" Deane J said, at (FLR) 346; (ALR) 381:"It can also, in an appropriate context, refer to a real or not remote chance or possibility, regardless of whether it is less or more than 50 per cent. When used with the latter meaning in a phrase which is descriptive of conduct, the word is equivalent to `prone', `with a propensity' or `liable'."

In Waugh v Kippen (1986)160 CLR 156 the High Court applied Sheen in its consideration of "so heavy as to be likely to cause risk of injury" in the Factories and Shops Act 1960 (Qld), General Rule, r1, CL25(1). The court said, at 166-7: "In our view, it is not possible to extract any significant meaning from the word `risk' that is not already comprehended by the word `injury' when used in association with the word `likely'. The employer is required to protect an employee from the risk of injury by preventing a situation from occurring where such a risk is likely. If the obligation was expressed in terms which were designed to forestall a likelihood of injury it would, in our opinion, amount to precisely the same obligation... The test accepted by a majority of the Court in Sheen (1984) 58 ALJR 93 ; 51 ALR 345 therefore supplies an authoritative test in the application of CL25(1)."

These observations are not entirely inapt when considering a provision such as s31(1) which also, albeit in slightly different language, speaks the likelihood of a danger. At this point, however, I need only observe that the meaning of the word "likely" is so sensitive to the context in which it appears that the tribunal, I think, should have in its reasons explained with greater precision the meaning it thought it had.

It is insufficient simply to equate "likely" with "probable" because as appears from my observations both words are sometimes used to mean more likely or more probable than not, and on others in the mathematical sense of degrees of likelihood or probability. In Beecham Group Ltd v Bristol Laboratories Pty Ltd (1968)118 CLR 618 the High Court spoke of "probability" in its mathematical sense, recognising by reference to "a fair probability" that there are degrees of "probabilities". The court there cited with approval the reference by Roper CJ in Eq in Linfield Linen Pty Ltd v Nejain (1951) 51 SR(NSW) 280 to a "fair probability". See also the well known work by Sir Richard Eggleston, Evidence, Proof and Probability.

In Overseas Tankship (UK) Ltd v Miller Steamship Co Pty Ltd Ltd (The Wagon Mound)(No 2) [1967] 1 AC 617 ; [1966] 2 All ER 709 Lord Reid said, at (AC) 634-5; (All ER) 713: "Another word frequently used is `probable'. It is used with various shades of meaning. Sometimes it appears to mean more probable than not, sometimes it appears to include events likely but not very likely to occur, sometimes it has a still wider meaning and refers to events the chance of which is anything more than a bare possibility, and sometimes, when used in conjunction with other adjectives, it appears to serve no purpose beyond rounding off a phrase."

The relevant expression here is "reasonably likely" which has some analogy to "fair probability". It suggests the mathematical approach; the word "reasonably" being a qualification of "likely", alternatively, a measure in colloquial language of the "likelihood".

The expression "reasonably likely" is substantially idiomatic, its meaning not necessarily unlocked by close dissection. In its ordinary use, it speaks of a chance of an event occurring or not occurring which is real--not fanciful or remote. It does not refer to a chance which is more likely than not to occur, that is, one which is "odds on", or where between nil and certainty it should be placed. A chance which in common parlance is described as "reasonable" is one that is "fair", "sufficient" or "worth noting". It is not inapt to attribute such meaning to the expression in s31(1) of the Act.

The word "reasonably" in an analogous context appears frequently in modern legislation. One suspects it is due to its increased use in modern life and its widely understood meaning. It does not necessarily require close attention to the word "reason" but rather to the line it seeks to draw between something worth considering and something not. The Shorter Oxford English Dictionary defines "reasonably" as "sufficiently, fairly".

In the Attorney-General's Department v Cockcroft (1986) 64 ALR 97 the Full Federal Court considered "could reasonably be expected to prejudice" in s43 of the Freedom of Information Act 1982 (Cth).

In their joint judgment Bowen CJ and Beaumont J said, at 106: "It is undesirable to attempt any paraphrase of these words. In particular, it is undesirable to consider the operation of the provision in terms of probabilities or possibilities or the like. To construe s43(1)(c)(ii) as depending in its application upon the occurrence of certain events in terms of any specific degree of likelihood or probability is, in our view, to place an unwarranted gloss upon the relatively plain words of the Act. It is preferable to confine the inquiry to whether the expectation claimed was reasonably based (see Kioa v Minister for Immigration and Ethnic Affairs (1985) 62 ALR 321 per Gibbs CJ and Mason J)." See also Arnold v Queensland (1987) 73 ALR 607 , at 628; News Corporation Ltd v National Companies and Securities Commission (1984) 5 FCR 88 at 95; 57 ALR 550 at 555.

A decision whether the specified disclosure in s31(1)(e) of the Act "would be reasonably likely" to endanger the physical safety of the persons mentioned, requires in a particular case consideration of many factors, not the least of which are the object and purpose of the legislation and of the quality of the risk sought to be guarded against by allowance of the exceptions. It is patent that an object of the Act is to promote public accessibility to records not readily or usually on view while at the same time meeting the need to protect persons who have contributed to those records confidential information which might upset or arouse the anger of significant sections of the population.

In the instant case, the information is about experiments on animals. It cannot be doubted that information of this kind should be generally available. But the evidence shows, and common experience tells, that the activity is one which touches strong emotions, often without justification, in some animal lovers. An example is in the cross-examination of Professor Short to which I have referred. The long existence of widely supported anti-vivisection movements is similar testimony: see Richard French, Anti-Vivisection and Medical Science and Victorian Society: The Anti-Vivisection Movement (1975).

Generally, it is well understood that experiments on species of animate life (and there are many examples where scientists have experimented on themselves) have been of untold benefit to humans. I need only mention that it has been from experiments on animals that scientists have learnt to measure accurately the work done by the heart and every change in the tissues during inflammation, discovered antiseptics (eg the work of the surgeon Lister) and the identity of serums preventative of anthrax, tuberculosis, diphtheria, rabies (hydrophobia), cholera, bubonic plague, typhoid, meningitis, infantile paralysis, myxoedema: see Encyclopaedia Brittanica, 14th ed, vol 1, at 970. The scientist TH Huxley said of the discoveries of Louis Pasteur, who experimented on fowls, sheep and other species of animate life, that the value of them was sufficient to cover the whole cost of the war indemnity paid by France to Germany in 1870: Encyclopaedia Brittanica, 14th ed, under heading "Pasteur".

The tribunal accepted the evidence of Professor Short and that there was a possibility" of the danger mentioned in s31(1)(e) but did not say how strong that possibility was, presumably because it appears to have concluded that a probability was required.

The facts found, and the matters I have mentioned, leave us in as good a position as the tribunal to decide the question. In the circumstances, I am empowered to give effect to my conclusions: Warren v Coombs (1979)142 CLR 531; 23 ALR 405 and Taylor v Johnson (1983)151 CLR 422; 45 ALR 265; see also, s52(5) of the Administrative Appeals Tribunal Act 1984.

In my opinion, disclosure of the names of the institutions and their departments is very likely in practice to disclose, albeit indirectly, the identity of the registered experimenters, if not, the group of which they are members, responsible for individual experiments. If it were assumed that experimenters are not to be visited with danger or risk to their physical safety either through harassment, pressure or some form of violent persuasion there appears little value to anyone in their being identified in the way sought. Certainly, non-disclosure gives minimal, if any, offence to the objects of the Act. Disclosure of the essential information is not in dispute. It is not necessary to show that the risk to which s31(1)(e) refers is from the respondent himself but rather from anyone should the information become generally known.

It must also be acknowledged that exemption applies where it "would be reasonably likely" that there be a danger to physical safety, not that physical harm will occur.

The risk of endangerment might well be thought to be greater than that of actual physical harm. The risk to be guarded against is of an experimenter being placed under threat, that is, in a position where he or she might or might not be physically harmed.

For the reasons I have discussed, and after consideration of all the matters to be taken into account, I consider that the appellant is entitled to an order varying the decision of the tribunal.

It is not necessary to consider the submissions as to adequacy of the tribunal's reasons.

I would allow the appeal, and order that the order of dismissal and for costs of Nathan J be set aside and the decision of the Administrative Appeals Tribunal, 20 March 1987 varied by adding to the deletions referred to in para 1 the following:

"and, save as appears in para 2, the names of all institutions and the names of the departments within them."