An, R (on the application of) & Anor v Secretary of State for the Home Department & Ors

Mr Bowen submits that Munby J was wrong to hold that the applicable standard of proof is “the ordinary civil standard of proof ‘on a balance of probability’” (para 71). Equally, the obiter dicta in Reid v. Secretary of State for Scotland [1999] 2 AC 512, 539 and R (H) v. London North and East Region Mental Health Review Tribunal [2002] QB 1 at para 32, to the effect that proof is on a balance of probabilities, are mistaken. The issue of standard of proof in this context has not previously been the subject of a considered decision.

Drawing on an article by Dr Mike Redmayne, Standards of Proof in Civil Litigation (1999) MLR 167, and observations of Burger CJ in Addington v. Texas (cited above), Mr Bowen submits that the ordinary civil standard of proof on the balance of probabilities represents an equal sharing (50:50) of the litigation risk between the parties, on the basis that the same utility may be ascribed to one party wrongly (i.e. unjustly) winning a case as to another wrongly losing it. The criminal standard, on the other hand, places almost the entire litigation risk on the state, on the basis that there is an interest of transcending value at stake (the liberty of the accused) and society places a much higher value upon ensuring that the innocent are not convicted than it does upon ensuring that the guilty are convicted.

Mr Bowen submits that there are, however, many circumstances in which an interest of transcending value falls to be determined in a civil context where an equal sharing of the litigation risk is not appropriate. The common law has long recognised that in certain contexts the ordinary civil standard of proof may accordingly require modification, as where serious allegations of a criminal nature are made in civil cases or the consequences of a finding one way or the other will be particularly grave for the individual concerned, for third parties or for the general public. In some cases more than one such feature may exist and the competing utilities of the individual’s rights and those of third parties or the wider public interest may have to be balanced.

Mr Bowen suggests that four different solutions have been adopted as regards the standard of proof in such cases:

Reference is also made to the “convincing evidence” standard (“convincingly established”) that has been applied by the ECtHR when determining whether an interference with various qualified rights under the ECHR is necessary and proportionate. The same approach has been adopted by the ECtHR, and in this jurisdiction by the Court of Appeal in R (N) v. M [2003] 1 WLR 562, in determining whether there is a medical necessity for compulsory medical treatment that would otherwise amount to a violation of article 3. It is submitted that, although the question has not been directly answered in the Strasbourg case-law, the same standard should apply under article 5 to the justification of a detention in hospital on grounds of unsoundness of mind.

By analogy with those situations where a higher standard of proof has been held to be appropriate (quasi-criminal cases, disciplinary proceedings, contempt proceedings) and by reference to the approach adopted in other common law jurisdictions, it is submitted that the transcending values of liberty and autonomy are such that the social cost of erroneous detentions must be seen as greater than that of erroneous decisions to release. While the potential cost of erroneous decisions to release is that patients so released will harm themselves or others, in the generality of cases that is an extremely small risk.

In these circumstances it is submitted that a higher than ordinary civil standard of proof should be applied to decisions under sections 72 and 73. In the course of his submissions Mr Bowen put forward various formulations of an appropriate self-direction by the tribunal. By the end of his reply he had refined them into four possibilities, all of which were said to meet the requirements of fairness and of article 5, though the first was preferred: (i) the tribunal must be satisfied to a high degree of probability that the criteria for detention are made out; (ii) the tribunal must be so satisfied by convincing evidence; (iii) the tribunal must be so satisfied by clear and convincing evidence; and (iv) the tribunal must be so satisfied to the civil standard of proof with the strictness appropriate to the seriousness of the matters to be proved and the implications of proving them.

We think it helpful to start by looking at the main English authorities chronologically rather than by reference to the classifications adopted by Mr Bowen, though we leave to one side, for the time being, the cases dealing specifically with the ECHR.

Bater v. Bater [1951] P 35 concerned a wife’s petition for divorce on the ground of cruelty. The Court of Appeal held that it was not a misdirection for the trial judge to have stated that the petitioner must prove her case beyond reasonable doubt. Bucknill LJ, with whom Somervell LJ agreed, considered that “a high standard of proof” was required because of the importance of such a case to the parties and the community. Denning LJ stated (at pages 36-37):

“The difference of opinion which has been evoked about the standard of proof in recent cases may well turn out to be more a matter of words than anything else. It is of course true that by our law a higher standard of proof is required in criminal cases than in civil cases. But this is subject to the qualification that there is no absolute standard in either case. In criminal cases the charge must be proved beyond reasonable doubt, but there may be degrees of proof within that standard.

As Best CJ and many other great judges have said, ‘in proportion as the crime is enormous, so ought the proof to be clear’. So also in civil cases, the case may be proved by a preponderance of probability, but there may be degrees of probability within that standard. The degree depends on the subject-matter. A civil court, when considering a charge of fraud, will naturally require for itself a higher degree of probability than that which it would require when asking if negligence is established. It does not adopt so high a degree as a criminal court, even when it is considering a charge of a criminal nature; but still it does require a degree of probability which is commensurate with the occasion. Likewise, a divorce court should require a degree of probability which is proportionate to the subject-matter.”

The issue in Hornal v. Neuberger Products Ltd. [1957] 1 QB 247 was the standard of proof in a civil claim for fraudulent misrepresentation. The Court of Appeal held that the trial judge had directed himself correctly by reference to the balance of probability. Denning LJ referred back to the views he had expressed in Bater v. Bater. Hodson LJ expressed his complete agreement with those views, adding (at page 264):

“Just as in civil cases the balance of probability may be more readily tilted in one case than in another, so in criminal cases proof beyond reasonable doubt may more readily be attained in some cases than in others.”

Morris LJ’s observations (at page 266) are particularly illuminating:

“It is, I think, clear from the authorities that a difference of approach in civil cases has been recognized. Many judicial utterances show this. The phrase ‘balance of probabilities’ is often employed as a convenient phrase to express the basis upon which civil issues are decided. It may well be that no clear-cut logical reconciliation can be formulated in regard to the authorities on these topics. But perhaps they illustrate that ‘the life of the law is not logic but experience.’ In some criminal cases liberty may be involved; in some it may not. In some civil cases the issues may involve questions of reputation which can transcend in importance even questions of personal liberty. Good name in man or woman is ‘the immediate jewel of their souls.’

But in truth no real mischief results from an acceptance of the fact that there is some difference of approach in civil actions. Particularly is this so if the words which are used to define that approach are the servants but not the masters of meaning. Though no court and no jury would give less careful attention to issues lacking gravity than to those marked by it, the very elements of gravity become a part of the whole range of circumstances which have to be weighed in the scale when deciding as to the balance of probabilities. This view was denoted by Denning LJ when in his judgment in Bater v. Bater he spoke of a ‘degree of probability which is commensurate with the occasion’ and of ‘a degree of probability which is proportionate to the subject-matter.’

In English law the citizen is regarded as being a free man of good repute. Issues may be raised in a civil action which affect character and reputation, and these will not be forgotten by judges and juries when considering the probabilities in regard to whatever misconduct is alleged. …"

The factual background to In re Dellow’s Will Trusts [1964] 1 WLR 451 was that a husband and wife, having made mutual wills each leaving their estate to the other, had been found dead in their home from coal gas poisoning. The husband was presumed to have died first. The question then arose as to whether the wife had feloniously killed him. As regards the standard of proof, Ungoed-Thomas J referred to the observations of Morris LJ in Hornal v. Neuberger Products Ltd. and continued (at pages 454-455):

“It seems to me that in civil cases it is not so much that a different standard of proof is required in different circumstances varying according to the gravity of the issue, but, as Morris LJ says, the gravity of the issue becomes part of the circumstances which the court has to take into consideration in deciding whether or not the burden of proof has been discharged. The more serious the allegation the more cogent is the evidence required to overcome the unlikelihood of what is alleged and thus to prove it. This is perhaps a somewhat academic distinction and the practical result is stated by Denning LJ: ‘The more serious the allegation the higher the degree of probability that is required: but it need not, in a civil case, reach the very high standard required by the criminal law.’ In this case the issue is whether or not the wife feloniously killed the husband. There can hardly be a graver issue than that, and its gravity weighs very heavily against establishing that such a killing took place, even for the purpose of deciding a civil issue.”

The issue in Blyth v. Blyth [1966] AC 643 was the standard of proof applicable to the question whether adultery had been condoned. The House of Lords held by a majority that it was the balance of probability. Lord Denning, referring back to Bater v. Bater and to Hornal v. Neuberger Products Ltd, said this (at page 669):

“In short it comes to this: so far as the grounds for divorce are concerned, the case, like any civil case, may be proved by a preponderance of probability, but the degree of probability depends on the subject-matter. In proportion as the offence is grave, so ought the proof to be clear. So far as the bars to divorce are concerned, like connivance or condonation, the petitioner need only show that on balance of probability he did not connive or condone as the case may be” (original emphasis).

In R v. Secretary of State for the Home Department, ex p. Khawaja [1984] AC 74 it was held by the House of Lords that, on an application for judicial review of an order detaining a person as an illegal entrant, it was for the executive to prove to the satisfaction of the court, on the balance of probabilities, the facts relied on by the immigration officer as justifying his conclusion that the applicant was an illegal entrant. Lord Scarman dealt at length with the standard of proof. Having cited Bater v. Bater, Hornal v. Neuberger Products Ltd., In re Dellow’s Will Trusts and Blyth v. Blyth, he continued (at pages 113F-114C):

“My Lords, I would adopt as appropriate to cases of restraint put by the executive upon the liberty of the individual the civil standard flexibly applied in the way set forth in the cases cited: and I would direct particular attention to the words of Morris LJ already quoted. It is not necessary to import into the civil proceedings of judicial review the formula devised by judges for the guidance of juries in criminal cases. Liberty is at stake: that is, as the court recognised in Bater v. Bater [1951] P. 35 and in Hornal v. Neuberger Products Ltd.[1957] 1 Q.B. 247, a grave matter. The reviewing court will therefore require to be satisfied that the facts which are required for the justification of the restraint put upon liberty do exist. The flexibility of the civil standard of proof suffices to ensure that the court will require the high degree of probability which is appropriate to what is at stake. "... the nature and gravity of an issue necessarily determines the manner of attaining reasonable satisfaction of the truth of the issue": Dixon J. in Wright v. Wright (1948) 77 C.L.R. 191, 210. I would, therefore, adopt the civil standard flexibly applied in the way described in the case law to which I have referred. And I completely agree with the observation made by my noble and learned friend, Lord Bridge of Harwich, that the difficulties of proof in many immigration cases afford no valid ground for lowering the standard of proof required.

Accordingly, it is enough to say that, where the burden lies on the executive to justify the exercise of a power of detention, the facts relied on as justification must be proved to the satisfaction of the court. A preponderance of probability suffices: but the degree of probability must be such that the court is satisfied. The strictness of the criminal formula is unnecessary to enable justice to be done: and its lack of flexibility in a jurisdiction where the technicalities of the law of evidence must not be allowed to become the master of the court could be a positive disadvantage inhibiting the efficacy of the developing safeguard of judicial review in the field of public law.”

Lord Fraser expressed agreement with Lord Scarman on that issue, stating (at page 97G):

“the appropriate standard is that which applies generally in civil proceedings, namely proof on a balance of probabilities, the degree of probability being proportionate to the nature and gravity of the issue. As cases such as those in the present appeals involve grave issues of personal liberty, the degree of probability required will be high.”

Similarly, Lord Bridge, with whom Lord Templeman expressed agreement, concluded (at page 124E):

“the civil standard of proof by a preponderance of probability will suffice, always provided that, in view of the gravity of the charge of fraud which has to be made out and of the consequences which will follow if it is, the court should not be satisfied with anything less than probability of a high degree.”

The next authority in the chronological sequence is one of central importance. In re H (Minors) (Sexual Abuse: Standard of Proof) [1996] AC 563 concerned section 31 of the Children Act 1989 which, in summary, empowers the court to make an order placing a child in the care of the local authority if satisfied that the child (i) is suffering significant harm or (ii) is likely to do so. The mother in that case had four children, all girls. The local authority applied for a care order in respect of the three youngest girls, basing its application solely on allegations of sexual abuse of the eldest girl. The House of Lords held by a majority that, just as there must be facts, properly proved to the court’s satisfaction if disputed, on which the court can properly conclude that a child “is suffering” harm, so too there must be facts from which the court can properly conclude that a child “is likely to suffer” harm (i.e., as was also held, that there is a real possibility that the child will suffer harm in the future); and here too, if the facts are disputed, the court must resolve the dispute so far as necessary to reach a proper conclusion on the issue.

That is the context within which their Lordships considered a further issue, as to the standard of proof required to prove relevant facts, such as the allegations of sexual abuse on which the application was founded. On that issue Lord Nicholls stated, in a passage at pages 586C-587E that we think it necessary to quote almost in full:

“Where the matters in issue are facts the standard of proof required in non-criminal proceedings is the preponderance of probability, usually referred to as the balance of probability. This is the established general principle. There are exceptions such as contempt of court applications, but I can see no reason for thinking that family proceedings are, or should be, an exception. … Family proceedings often raise very serious issues, but so do other forms of civil proceedings.

The balance of probability standard means that a court is satisfied an event occurred if the court considers that, on the evidence, the occurrence of the event was more likely than not. When assessing the probabilities the court will have in mind as a factor, to whatever extent is appropriate in the particular case, that the more serious the allegation the less likely it is that the event occurred and, hence, the stronger should be the evidence before the court concludes that the allegation is established on the balance of probability. Fraud is usually less likely than negligence. Deliberate physical injury is usually less likely than accidental physical injury. A step-father is usually less likely to have repeatedly raped and had non-consensual oral sex with his under age stepdaughter than on some occasion to have lost his temper and slapped her. Built into the preponderance of probability standard is a generous degree of flexibility in respect of the seriousness of the allegation.

Although the result is much the same, this does not mean that where a serious allegation is in issue the standard of proof required is higher. It means only that the inherent probability or improbability of an event is itself a matter to be taken into account when weighing the probabilities and deciding whether, on balance, the event occurred. The more improbable the event, the stronger must be the evidence that it did occur before, on the balance of probability, its occurrence will be established. Ungoed-Thomas J expressed this neatly in In re Dellow's Will Trusts [1964] 1 WLR 451, 455: ‘The more serious the allegation the more cogent is the evidence required to overcome the unlikelihood of what is alleged and thus to prove it.’

This substantially accords with the approach adopted in authorities such as the well known judgment of Morris L.J. in Hornal v. Neuberger Products Ltd. [1957] 1 QB 247, 266. This approach also provides a means by which the balance of probability standard can accommodate one’s instinctive feeling that even in civil proceedings a court should be more sure before finding serious allegations proved than when deciding less serious or trivial matters.

No doubt it is this feeling which prompts judicial comment from time to time that grave issues call for proof to a standard higher than the preponderance of probability. … The law looks for probability, not certainty. Certainty is seldom attainable. But probability is an unsatisfactorily vague criterion because there are degrees of probability. In establishing principles regarding the standard of proof, therefore, the law seeks to define the degree of probability appropriate for different types of proceedings. Proof beyond reasonable doubt, in whatever form of words expressed, is one standard. Proof on a preponderance of probability is another, lower standard having the in-built flexibility already mentioned. If the balance of probability standard were departed from, and a third standard were substituted in civil cases, it would be necessary to identify what the standard is and when it applies. Herein lies a difficulty. If the standard were to be higher than the balance of probability but lower than the criminal standard of proof beyond reasonable doubt, what would it be? The only alternative which suggests itself is that the standard should be commensurate with the gravity of the allegation and the seriousness of the consequences. A formula to this effect has its attraction. But I doubt whether in practice it would add much to the present test in civil cases, and it would risk causing confusion and uncertainty. As at present advised I think it is better to stick to the existing, established law on this subject. I can see no compelling need for a change.”

Lords Goff and Mustill agreed with Lord Nicholls. Lord Browne-Wilkinson, although dissenting in part on other issues, agreed with Lord Nicholls that “the standard of proof is the ordinary civil standard, i.e. balance of probabilities” (page 572E).

In re H was considered in Secretary of State for the Home Department v. Rehman [2003] 1 AC 153, which concerned the making of a deportation order under section 3(5)(b) of the Immigration Act 1971 on the ground that deportation would be conducive to the public good in the interests of national security. The Special Immigration Appeals Commission held that the Secretary of State had not established to a high degree of probability that the applicant had been or was likely to be a threat to national security. The Court of Appeal allowed the Secretary of State’s appeal. The House of Lords, in upholding the Court of Appeal’s decision, held that the concept of standard of proof was not applicable to the evaluation of whether the risk to national security was sufficient to justify deportation, but that where past acts were relied on they should be proved to the civil standard of proof. As to that, Lord Hoffmann stated:

“55.

I turn next to the Commission's views on the standard of proof. By way of preliminary I feel bound to say that I think that a 'high civil balance of probabilities' is an unfortunate mixed metaphor. The civil standard of proof always means more likely than not. The only higher degree of probability required by the law is the criminal standard. But, as Lord Nicholls of Birkenhead explained in In re H (Minors) (Sexual Abuse: Standard of Proof) [1996] AC 563, 586, some things are inherently more likely than others. It would need more cogent evidence to satisfy one that the creature seen walking in Regent's Park was more likely than not to have been a lioness than to be satisfied to the same standard of probability that it was an Alsatian. On this basis, cogent evidence is generally required to satisfy a civil tribunal that a person has been fraudulent or behaved in some other reprehensible manner. But the question is always whether the tribunal thinks it more probable than not.”

Lords Clyde and Hutton agreed with Lord Hoffmann. Lord Slynn, with whom Lords Steyn and Hutton agreed, referred simply to the need to prove facts “to the civil standard of proof” (para 22).

In view of the submissions made by Mr Bowen, we should note that Ex p. Khawaja was not cited to the House of Lords in In re H (though all the other cases that we have mentioned so far in this chronological survey were), but that it was cited in Rehman.

The next case to consider is B v. Chief Constable of Avon and Somerset Constabulary [2001] 1 WLR 340, which arose out of an application to the magistrates’ court for a sex offender order under section 2 of the Crime and Disorder Act 1998. On such an application it must be proved that the conditions mentioned in section 2(1) are fulfilled, namely “(a) that the person is a sex offender; and (b) that the person has acted, since the relevant date, in such a way as to give reasonable cause to believe that an order under this section is necessary to protect the public from serious harm from him”. The Divisional Court held that such an application was properly characterised as a civil, not a criminal, proceeding and that the justices were accordingly required to apply the civil standard of proof. As to that standard, however, Lord Bingham CJ, giving the leading judgment, went on to observe:

“30.

It should, however, be clearly recognised, as the justices did expressly recognise, that the civil standard of proof does not invariably mean a bare balance of probability, and does not so mean in the present case. The civil standard is a flexible standard to be applied with greater or lesser strictness according to the seriousness of what has to be proved and the implications of proving those matters: Bater v Bater [1951] P 35, Hornal v Neuberger Products Ltd [1957] 1 QB 247, and R v Secretary of State for the Home Department, Ex p Khawaja [1984] AC 74.

31.

In a serious case such as the present the difference between the two standards is, in truth, largely illusory. I have no doubt that, in deciding whether the condition in section 2(1)(a) is fulfilled, a magistrates' court should apply a civil standard of proof which will for all practical purposes be indistinguishable from the criminal standard. In deciding whether the condition in section 2(1)(b) is fulfilled the magistrates' court should apply the civil standard with the strictness appropriate to the seriousness of the matters to be proved and the implications of proving them.”

It should be noted that, although earlier cases were referred to in para 30 of Lord Bingham’s judgment, neither In re H nor Rehman was cited to the court.

The subject matter of Gough v. Chief Constable of the Derbyshire Constabulary [2002] QB 1213 was an application to a magistrates’ court for a banning order under section 14B of the Football Spectators Act 1989. The Court of Appeal held that proceedings under that section were civil in character. As regards the standard of proof, however, Lord Philips MR, giving the judgment of the court, stated:

“90.

It does not follow from this that a mere balance of probabilities suffices to justify the making of an order. Banning orders under section 14B fall into the same category as antisocial behaviour orders and sex offender orders. While made in civil proceedings they impose serious restraints on freedoms that the citizen normally enjoys. While technically the civil standard of proof applies, that standard is flexible and must reflect the consequences that will follow if the case for a banning order is made out. This should lead the justices to apply an exacting standard of proof that will, in practice, be hard to distinguish from the criminal standard: see B v Chief Constable of Avon and Somerset Constabulary [2001] 1 WLR 340, 354 and R (McCann) v Crown Court at Manchester [2001] 1 WLR 1084, 1102.

91.

Thus the necessity in the individual case to impose a restriction upon a fundamental freedom must be strictly demonstrated. The first thing that has to be proved under section 14B(4)(a) is that the respondent has caused or contributed to violence or disorder in the United Kingdom or elsewhere. Mr Pannick conceded that the standard of proof of this is practically indistinguishable from the criminal standard.

92.

The same is true of the next requirement, that imposed by section 14B(4)(b), though this is less easily derived from the language of the statute. The court must be ‘satisfied that there are reasonable grounds to believe that making a banning order would help to prevent violence or disorder at or in connection with any regulated football matches’. In practice the ‘reasonable grounds’ will almost inevitably consist of evidence of past conduct. That conduct must be such as to make it reasonable to conclude that if the respondent is not made subject to a banning order he is likely to contribute to football violence or disorder in the future. The past conduct may or may not consist of or include the causing or contributing to violence or disorder that has to be proved under section 14B(4)(a), for that violence or disorder is not required to be football related. It must, however, be proved to the same strict standard of proof. Furthermore it must be conduct that gives rise to the likelihood that, if the respondent is not banned from attending prescribed football matches, he will attend such matches, or the environs of them, and take part in violence or disorder.

93.

These matters are not readily susceptible of proof. We can well understand the practice that is evidenced by this case of using a football intelligence service to build up profiles of ‘football prominents’. Such a practice may well be the only way of assembling evidence sufficiently cogent to satisfy the requirements of section 14B(4)(b). Those requirements, if properly applied in the manner described above, will provide a satisfactory threshold for the making of a banning order.”

Lord Phillips referred in para 90 to the decision of the Court of Appeal in R (McCann) v. Manchester Crown Court. McCann was the subject of a further appeal to the House of Lords, whose decision is reported at [2003] 1 AC 787. The case concerned an application to the magistrates’ court for anti-social behaviour orders under section 1 of the Crime and Disorder Act 1998. Again this was held to be a civil proceeding. Nevertheless their Lordships went on to hold that the standard of proof to be applied to allegations about the defendants’ past behaviour was the criminal standard. Lord Steyn stated:

“37.

Having concluded that the relevant proceedings are civil, in principle it follows that the standard of proof ordinarily applicable in civil proceedings, namely the balance of probabilities, should apply. However, I agree that, given the seriousness of the matters involved, at least some reference to the heightened civil standard would usually be necessary: In re H (Minors) (Sexual Abuse: Standard of Proof) [1996] AC 563, 586D-H, per Lord Nicholls of Birkenhead. For essentially practical reasons, the Recorder of Manchester decided to apply the criminal standard. The Court of Appeal said that would usually be the right course to adopt. Lord Bingham of Cornhill has observed that the heightened civil standard and the criminal standard are virtually indistinguishable. I do not disagree with any of these views. But in my view pragmatism dictates that the task of magistrates should be made more straightforward by ruling that they must in all cases under section 1 apply the criminal standard. …”

Lord Hope gave the following reasons for endorsing the approach of applying the criminal standard:

“82.

Mr Crow for the Secretary of State said that his preferred position was that the standard to be applied in these proceedings should be the civil standard. His submission, as it was put in his written case, was that although the civil standard was a single, inflexible test, the inherent probability or improbability of an event was a matter to be taken into account when the evidence was being assessed. He maintained that this view was consistent with the position for which he contended, that these were civil proceedings which should be decided according to the civil evidence rules. But it is not an invariable rule that the lower standard of proof must be applied in civil proceedings. I think that there are good reasons, in the interests of fairness, for applying the higher standard when allegations are made of criminal or quasi-criminal conduct which, if proved, would have serious consequences for the person against whom they are made.

83.

This … was the view which the Court of Session took in Constanda v. M 1997 SC 217 when it decided that proof to the criminal standard was required of allegations that a child had engaged in criminal conduct although the ground of referral to a children’s hearing was not that he had committed an offence but that he was exposed to moral danger. There is now a substantial body of opinion that, if the case for an order such as a banning order or a sex offender order is to be made out, account should be taken of the seriousness of the matters to be proved and the implications of proving them. It has also been recognised that if this is done the civil standard of proof will for all practical purposes be indistinguishable from the criminal standard: see B v Chief Constable of Avon and Somerset Constabulary [2001] 1 WLR 340, 354, para 31, per Lord Bingham of Cornhill CJ; Gough v Chief Constable of the Derbyshire Constabulary [2002] QB 1213, 1242-1243, para 90, per Lord Phillips of Worth Matravers MR. As Mr Crow pointed out, the condition in section 1(1)(b) of the Crime and Disorder Act 1998 that a prohibition order is necessary to protect persons in the local government area from further anti-social acts raises a question which is a matter for evaluation and assessment. But the condition in section 1(1)(a) that the defendant has acted in an anti-social manner raises serious questions of fact, and the implications for him of proving that he has acted in this way are also serious. I would hold that the standard of proof that ought to be applied in these cases to allegations about the defendant's conduct is the criminal standard.”

The other members of the House in McCann agreed with Lord Steyn and Lord Hope.

Finally in this chronological survey, reference should be made to the decision of the Privy Council in Campbell v. Hamlet [2005] UKPC 19. In that case there had been a finding of professional misconduct by an attorney. One of the issues on the appeal was the correct standard of proof to apply to such proceedings. At para 14 of the judgment, Lord Brown stated that their Lordships entertained no doubt that the criminal standard was the correct standard. He continued:

“17.

It has, of course, long been established that there is flexibility in the civil standard of proof which allows it to be applied with greater or lesser strictness according to the seriousness of what has to be proved and the implications of proving those matters.”

He then referred with evident approval, at paras 17-19, to passages that we have already cited from B v. Chief Constable of Avon and Somerset Constabulary, Gough v. Chief Constable of the Derbyshire Constabulary and R (McCann) v. Crown Court at Manchester in which it was indicated that, in the situations there under consideration, the civil standard of proof would for all practical purposes be indistinguishable from the criminal standard.

We have set out those authorities in chronological order because, in our judgment, it helps to show that Mr Bowen’s submissions seek to impose an unwarranted straitjacket of classification upon them. Certainly there are differences in the language used and the rationalisations given over time; but the essential point that runs through the authorities is that the civil standard of proof is flexible in its application and enables proper account to be taken of the seriousness of the allegations to be proved and of the consequences of proving them.

Whatever differences in expression there have been over time, it was laid down clearly by the House of Lords in In re H and Rehman that in English law the civil standard is one single standard, namely proof on the balance of probabilities (or preponderance of probability). The other standard is the criminal standard of proof beyond reasonable doubt. There is no intermediate standard, nor is the civil standard to be broken down into sub-categories designed to produce one or more intermediate standards. (We leave out of account for the purposes of this analysis the standards applicable in certain specific statutory contexts, such as a “reasonable likelihood” of persecution in asylum cases. The present case is governed by the general rules.)

We reject Mr Bowen’s contention that the decision in In re H might have been different if Khawaja had been cited in it. Lord Nicholls in In re H considered his approach substantially to accord with that in Hornal, which was approved in Khawaja and which Mr Bowen would place in the same category as Khawaja. In any event Khawaja was cited in Rehman, where In re H was followed. The point might equally be made that In re H, in turn, was not cited in B or Gough, though it was in McCann, and that Rehman was not cited in any of those three cases. In our view, however, this does not mean that successive cases have been decided per incuriam or that the law has been getting into a state of confusion. Full citation of earlier authorities has not been necessary in these cases, for the very reason that their broad thrust has been the same (albeit that, as we explain below, slightly different language might sometimes have been used if In re H had been cited in some of the more recent cases).

Although there is a single civil standard of proof on the balance of probabilities, it is flexible in its application. In particular, the more serious the allegation or the more serious the consequences if the allegation is proved, the stronger must be the evidence before a court will find the allegation proved on the balance of probabilities. Thus the flexibility of the standard lies not in any adjustment to the degree of probability required for an allegation to be proved (such that a more serious allegation has to be proved to a higher degree of probability), but in the strength or quality of the evidence that will in practice be required for an allegation to be proved on the balance of probabilities.

The flexibility that exists in the application of the standard is clear from In re H itself, where Lord Nicholls, whilst affirming the existence of a single civil standard, stressed that it had “a generous degree of flexibility” in respect of the seriousness of the allegation (page 586F – see also his reference, at page 587E, to “the in-built flexibility already mentioned”).

It is true that the rationalisation put forward in In re H and followed in Rehman focused on the seriousness of the allegation rather than on the seriousness of the consequences if the allegation is proved. The reasoning was that the more serious the allegation the less likely it is that the event occurred, and that the inherent probability or improbability of an event is itself a matter to be taken into account when weighing the probabilities and deciding whether, on balance, the event occurred. In general, the seriousness of an allegation is a function of the seriousness of its consequences, and vice versa, so that the rationalisation in In re H and Rehman will take due account of the seriousness of the consequences if an allegation is proved. We accept Mr Bowen’s submission, however, that there will be cases where proof of an allegation may have serious consequences even though it cannot be said that the matter alleged is inherently improbable. It seems to us that the same general approach must apply in such cases, even though the rationalisation put forward in In re H does not readily accommodate it. The more serious the consequences, the stronger the evidence required in practice to prove the matter on the balance of probabilities.

In terms of outcome, and however difficult it may be to accommodate it within his reasoning, that conclusion accords with the way in which Lord Nicholls evidently thought of the matter in In re H. The alternative, intermediate standard which he formulated for consideration but then rejected was “commensurate with the gravity of the allegation and the seriousness of the consequences” (page 587E, emphasis added). In rejecting it, however, he doubted whether in practice it would add much to the balance of probability standard. Moreover Morris LJ in Hornal, with whose judgment Lord Nicholls considered his own approach substantially to accord, was plainly of the view that the seriousness of the consequences, whether for liberty or reputation, was a matter to be taken into account when deciding on the balance of probabilities (see the latter part of the passage quoted at para 39 above).

The more recent cases provide uniform confirmation of the need to take account of the seriousness of the consequences or implications of what has to be proved when applying the civil standard: see B v. Chief Constable of Avon and Somerset Constabulary at para 30, Gough at para 90 and McCann at para 83. The relevant passages are set out above and we do not need to repeat them.

Some of the language used in those cases might be thought to have the potential for confusion. For example, Lord Bingham said in B v. Chief Constable of Avon and Somerset Constabulary that the civil standard of proof “does not invariably mean a bare balance of probability” (para 30), and Lord Steyn referred in McCann to “the heightened civil standard” (para 37). Such observations are an echo of references in some of the earlier cases to higher degrees of probability being required where the allegations are more serious. In Rehman, however, Lord Hoffmann took issue with that kind of language, which does not sit comfortably with a single standard of balance of probabilities. If Rehman had been cited in the later cases, or indeed if In re H had been cited in B v. Chief Constable of Avon and Somerset Constabulary, it may well be that different words would have been used. But in our view it is only a matter of words. We do not think that anything said in the later cases affects the substance of the position as set out in In re H and Rehman.

In all of this we take comfort from Morris LJ’s observations in Hornal, quoted at para 39 above, that it may well be that no clear-cut logical reconciliation can be formulated in regard to the authorities, but perhaps they illustrate that “the life of the law is not logic but experience”; and that the words which are used to define the approach must be the servants not the masters of meaning.

Although there remains a distinction in principle between the civil standard and the criminal standard, the practical application of the flexible approach demonstrated in the authorities means that they are likely in certain contexts to produce the same or similar results. Indeed, there are exceptional situations in which, for reasons of policy or pragmatism, the actual criminal standard is used in civil proceedings, as in contempt of court (In re Bramblevale Ltd. [1970] 1 Ch 128), the making of anti-social behaviour orders (McCann) or certain disciplinary contexts (Campbell). These are exceptions to the general rule. Mr Bowen has not argued before us that the mental health context constitutes a further exception in which the criminal standard applies, and we are sure he is right not to have done so.

To recapitulate, we do not accept Mr Bowen’s submission that the House of Lords in In re H somehow took a wrong turn, reflecting only one strand in the earlier case-law and resulting in the erroneous adoption of a fixed standard instead of a flexible standard; or that there is an irreconcilable conflict between In re H and later cases such as McCann. In re H is in line with the overall thrust of the case-law, whilst analysing the matter in terms of a single civil standard (balance of probabilities) which is to be applied flexibly according to the seriousness of the matters to be proved and the consequences of proving them. The reasoning in support of the adoption of a single standard may not provide a complete explanation of that flexibility in its application, but the seriousness of the consequences if a matter is proved is nonetheless a factor to be taken into account when deciding in practice whether the evidence is sufficiently strong to prove that matter on the balance of probabilities.

We have not spent time on the Commonwealth authorities, such as Briginshaw, to which Mr Bowen referred us, but due account is taken of them in the English cases that we have examined. Although the US cases are of no direct assistance, since they were decided in the somewhat different context of the "due process" clause of the US Constitution and they lay down an intermediate standard which, as In re H makes clear, is no part of English law, they do focus on a similar problem and it seems to us that the “clear and convincing evidence” test in Addington v. Texas, the language of “reliably shown” or “convincingly shown” in the ECHR context, and the flexible application of the balance of probabilities standard may all lead to much the same result in practice.

Given that the standard of proof is flexible in its application, there remains the question whether evidence of an especially high strength or quality is required to meet the standard in the context of sections 72 and 73 of the Act (to the extent that the issues arising under those sections are susceptible of proof to a defined standard at all). We take it as axiomatic, and it is not in dispute, that cogent evidence will in practice be required in order to satisfy the tribunal, on the balance of probabilities, that the conditions for continuing detention are met. But we would not put it any higher than that.

As submitted by Mr McCullough on behalf of the tribunal, the mental health context is very different from other situations where individual liberty is at stake. The unwarranted detention of an individual on grounds of mental disorder is a very serious matter, but the unwarranted release from detention of an individual who is suffering from mental disorder is also a very serious matter. The decision of the US Supreme Court in Addington v. Texas, quoted extensively by Munby J at paras 51-57 of his judgment, is valuable for its discussion of the competing interests of the individual and the State in this area. The court concluded that the individual’s interest was of such weight and gravity that clear and convincing evidence was required to justify his involuntary confinement. Although the court dealt with the matter in terms of an intermediate standard of proof, its reasoning is relevant to the strength of the evidence required in the flexible application of the English balance of probabilities standard. In our view it supports a requirement of cogent evidence but does not compel a more demanding evidential requirement than that.

Several considerations weigh against pitching the evidential requirement unduly high. One of the points made by Burger CJ in rejecting the criminal standard of proof was that: "One who is suffering from a debilitating mental illness and in need of treatment is neither wholly at liberty nor free of stigma …. It cannot be said, therefore, that it is much better for a mentally ill person to 'go free' than for a mentally normal person to be committed." Furthermore the consequences that may flow from the release of a person suffering from mental disorder include not only a risk to the individual's own health and safety (e.g. self-harm, even suicide), but also a risk of harm to other members of the public. It is not to be forgotten that a person whose case is being considered under section 73 was detained in the first place pursuant to a hospital order under section 37 following conviction for a criminal offence, often an offence of violence: the appalling facts of AN’s own case are very much in point. Some of those whose cases are considered under section 72 will also have been detained pursuant to section 37. All those factors are taken into account in the operation of the statutory machinery under the 1983 Act.

The matter is expressed very well by Munby J at para 71 of his judgment. In the continuation of the passage quoted at para 18 above, the judge said:

“I agree with [Mr McCullough] that to raise the standard of proof above the ordinary civil standard of proof would subvert the obvious purpose of the Act, which seeks both to protect the interests of the individual whose ability to act in his own best interests is impaired and at the same time enable a proportionate balance … to be struck between individual and public interests. It would, as Mr McCullough submits, relegate the interests of the patient, as objectively ascertained, and of the public, to a position subsidiary to the principle of personal autonomy – an approach for which there is no principled basis. And it would thereby create a heightened risk to patients and the public – contrary, as it seems to me, to the very scheme and purpose of the Act.”

That was said in support of the judge's conclusion that the applicable standard of proof is the ordinary civil standard of proof on the balance of probabilities; but it also tells against demanding an especially high evidential requirement in order to meet that standard.

Turning to the actual decision in the present case, the tribunal was faced with competing expert opinions as to whether AN was suffering from a psychopathic disorder within the meaning of the Act. Directing itself by reference to the balance of probabilities, it gave a reasoned basis for preferring one group of experts over the other, concluding in effect that the diagnosis of psychopathic disorder was more likely than not to be the correct one. The body of evidence it preferred was plainly cogent and it is clear from the tribunal's reasons that its preference was reasonably clear-cut. The tribunal's approach cannot be faulted. The position would have been the same, however, even if both sets of competing opinions could fairly have been described as cogent and the decision between them had been a finely balanced one. We do not think that more is required than that the decision is based on cogent evidence which is accepted as correct on the balance of probabilities. It is not necessary, for example, for one body of evidence to have a much higher degree of cogency before it can be accepted on the balance of probabilities.

The same general considerations apply to the determination of other disputed issues of fact such as whether the patient has behaved in a particular way or said a particular thing in the past. The tribunal was wrong to adopt a standard of proof "akin to the criminal standard" (though, as Munby J held, it was an error favourable to AN and cannot help him in these proceedings). The tribunal's approach not only expresses the standard of proof incorrectly, but also suggests an evidential requirement higher than is appropriate for the proper application, in the present context, of the correct standard of proof. As regards the cogency of evidence required, we see no reason to disagree with the guidance given by Munby J at paras 121 et seq. of his judgment, where he dealt in particular with the evaluation of hearsay evidence; though we should stress that the detail of the points covered in that part of the judgment has not been the subject of argument before us.

In our judgment, the conclusion we have expressed above is in full conformity with the requirements of the ECHR.

In determining whether a detention on grounds of mental illness complies with article 5, the ECtHR has consistently applied the test laid down in Winterwerp v. Netherlands (1979) EHRR 387. For example, in HL v. United Kingdom (application no. 45508/99, judgment of 5 October 2004), the court stated at para 98:

“It is recalled that an individual cannot be deprived of his liberty on the basis of unsoundness of mind unless three minimum conditions are satisfied: he must reliably be shown to be of unsound mind; the mental disorder must be of a kind or degree warranting compulsory confinement; and the validity of continued confinement depends upon the persistence of such a disorder (the Winterwerp judgment, at para 39 …).”

The court examined the material on which the applicant’s detention had been based, and found that “the applicant has been reliably shown to have been suffering from a mental disorder of a kind or degree warranting compulsory confinement …” (para 101).

It seems to us that full effect is given to the Winterwerp test by the application of a standard of proof on the balance of probabilities and a recognition that cogent evidence will in practice be required to meet that standard. We note, too, that in Reid v. United Kingdom (2003) 37 EHRR 9, where the court repeated the Winterwerp test (see para 70 of the judgment), there was no suggestion that the standard of proof on the balance of probabilities was inappropriate or incompatible with the requirements of the ECHR, even though the court gave specific consideration to the two cases containing obiter dicta in support of that standard, namely Reid v. Secretary of State for Scotland and R (H) v. London North and East Region Mental Health Review Tribunal.

Mr Bowen has cited cases concerning the standard of proof required in other ECHR contexts. For example, in R (N) v. M [2003] 1 WLR 562 it was common ground, in the light of the decision of the ECtHR in Herczegfalvy v. Austria (1992) 15 EHHR 437, that the standard of proof for the purposes of determining whether medical treatment to which the patient does not consent is compatible with article 3 is that the medical necessity has been “convincingly shown”: see the judgment of the Court of Appeal at para 17. At para 18 the court rejected a contention that the test was in effect the same as the criminal standard of proof, stating that no useful purpose is served by importing the language of the criminal law, that the phrase "convincingly shown" is easily understood and that the standard is a high one but it does not need elaboration or further explanation. In R (B) v. Dr SS [2005] EWHC 1936 (Admin), Charles J considered, but left open, the relationship between the “convincingly shown” standard adopted in R (N) v. (M) and the decision of the House of Lords in In re H as to the civil standard of proof in English law. He proceeded on the basis of the "convincingly shown" standard, treating it (as the parties had apparently agreed) as lying between the English civil standard and criminal standard.

Although the "convincingly shown" standard is arguably different from “reliably shown” and it is not necessary to decide whether that is so or not for the purposes of this appeal, it seems to us, as indicated above, that they are not likely to produce materially different results from each other or from the “clear and convincing evidence” test in Adddington v. Texas. Indeed, it seems to us to be desirable, so far as possible, for one single (though flexible) approach to be adopted for these different problems in the civil law.

However that may be, the language of "reliably shown" has been consistently applied under the ECHR for many years in the context of detention on grounds of mental health. It must be taken as the correct test, and the application of the standard of proof on the balance of probabilities in the way we have described is capable of meeting it.

All parties accept that Munby J was right to hold that, whatever standard of proof is determined under issue (1) to be the correct standard in this context, it can and should be applied to the question whether a person is suffering from a mental disorder within section 72(1)(b)(i) (including, as we understand the concessions, the nature and degree of that disorder) and to any issue of past conduct that the tribunal may have to determine.

Mr Bowen submits, however, that the judge was wrong to hold that the remaining issues under sections 72 and 73 are not susceptible to proof to a defined standard but are to be determined by a process of evaluation and judgment. His case is that the same standard of proof is applicable to the tribunal’s determination of all the issues that arise under those sections.

This question is not governed by any binding authority, though Mr Bowen draws support from obiter dicta in two cases. In Reid v. Secretary of State for Scotland, Lord Clyde evidently considered there to be a burden on the patient to satisfy the decision-maker on the balance of probabilities as to each of the conditions contained in the Scottish equivalent of section 72 (see p.539F). The case pre-dated the amendment to the Act which placed the burden of proof on the detaining authority; but whether an issue is susceptible to proof to a particular standard should not depend on where the burden of proof lies. There was, however, no argument on the standard of proof or its scope. Similarly, in R (H) v. London North and East Region Regional Mental Health Review Tribunal the Court of Appeal referred in general terms to the relevant conditions being demonstrated on a balance of probability (see para 32 of its judgment), but again this was obiter and there was no argument on the point.

In developing his submissions, Mr Bowen takes as his starting-point the proposition that there is a burden on the detaining authority to satisfy the tribunal that the conditions for detention are met. The existence of that burden is common ground and is supported by R (H) v. London North and East Region Mental Health Review Tribunal and by the decision of the ECtHR in Reid v. United Kingdom (at least to the extent that the issues that arise under sections 72 and 73 are also within the scope of article 5 ECHR). The existence of the burden is unaffected by the fact that aspects of the tribunal’s procedures are inquisitorial in nature (for example, the requirement of a medical examination by the medical member of the tribunal and the power to require the attendance of witnesses and to call for further information).

Mr Bowen’s essential argument is that where there is a burden of proof, there must be a corresponding standard of proof. If no standard applies, then there is in truth no burden of proof. It makes no sense to talk in terms of a burden of proof unless there is a standard by reference to which the matter must be proved; and if a standard does apply, it must be the same standard as that held to apply to the determination of issues of fact.

We are not persuaded by the logic of that argument. It is inherent in the statutory language that the detaining authority has to satisfy the tribunal that the relevant conditions are met: “the tribunal shall direct the discharge of a patient … if they are not satisfied” as to the appropriateness and necessity, etc., of continuing detention. If the authority fails to satisfy the tribunal as to any of the conditions, the tribunal must order discharge. Thus the authority has the burden of persuading the tribunal to form the requisite judgments. Munby J preferred to describe this as an onus or persuasive burden (see para 106 of his judgment). Whether or not one prefers that language, the point of substance is the same. The existence of a burden does not mean that there must be a particular standard of proof in play.

The point can be illustrated by reference to McCann, upon which the judge relied as authority at the highest level against Mr Bowen’s argument. We have already explained that McCann concerned applications for anti-social behaviour orders. Such an application is made by a local authority to a magistrates’ court. Section 1(4) of the 1998 Act provides that “if, on such an application, it is proved that the conditions mentioned in subsection (1) are fulfilled”, the court may make an order. The conditions in subsection (1) are “(a) that the person has acted … in an anti-social manner …; and (b) that such an order is necessary to protect persons … from further anti-social acts by him”. Having held that pragmatism favoured the adoption of the criminal standard of proof under section 1, Lord Steyn continued (at the end of para 37):

“If the House takes this view it will be sufficient for the magistrates, when applying section 1(1)(a) to be sure that the defendant has acted in an anti-social manner …. The inquiry under section 1(1)(b), namely that such an order is necessary to protect persons from further anti-social acts by him, does not involve a standard of proof: it is an exercise of judgment or evaluation” (original emphasis).

We have set out previously the passage in para 83 where Lord Hope expressed a similar view (see para 55 above). We have also pointed out that the rest of their Lordships expressed agreement with both Lord Steyn and Lord Hope.

Mr Bowen seeks to deflect the force of McCann by submitting, as a matter of inference, that their Lordships considered that neither a burden nor a standard of proof was applicable under section 1(1)(b). That cannot be right. It is plain from the terms of section 1(4) that there is a burden on the local authority to prove that the conditions in section 1(1) are met. The effect of McCann is that, in order to prove that the condition in section 1(1)(b) is met, the local authority must persuade the court to form a judgment that an order is “necessary”. Standard of proof does not come into it, but there is still a burden of proof.

There are other statutory contexts in which it has been held that the concept of standard of proof is inapplicable or unhelpful in relation to questions of judgment or evaluation, especially as regards the assessment of future risk.

Munby J, at paras 91-93 of his judgment, cited passages from the decisions of the Court of Appeal and the House of Lords in Rehman. A sufficient flavour is given by this passage from the speech of Lord Hoffmann (at para 56):

“In any case, I agree with the Court of Appeal that the whole concept of a standard of proof is not particularly helpful in a case such as the present. In a criminal or civil trial in which the issue is whether a given event happened, it is sensible to say that one is sure that it did, or that one thinks it more likely than not that it did. But the question in the present case is not whether a given event happened but the extent of future risk. This depends upon an evaluation of the evidence of the appellant’s conduct against a broad range of facts with which they may interact. The question of whether the risk to national security is sufficient to justify the appellant’s deportation cannot be answered by taking each allegation seriatim and deciding whether it has been established to some standard of proof. It is a question of evaluation and judgment, in which it is necessary to take into account not only the degree of probability of prejudice to national security but also the importance of the security interest at stake and the serious consequences of deportation for the deportee.”

It must be borne in mind that in Rehman the primary decision-maker was the Secretary of State, who had to form an executive judgment as to whether there was a danger to national security, and that the House of Lords was considering the correct approach of the Special Immigration Appeals Commission when reviewing the Secretary of State’s decision. So there is not a precise parallel with the present situation, where the detaining authority has the burden of satisfying the primary decision-maker, the tribunal, that certain conditions are met. Nevertheless, Rehman is important for the contrast it draws between matters of fact, which the Secretary of State had to establish to the civil standard of proof (see also per Lord Slynn at pp.184H-185A), and matters of judgment and evaluation, in relation to which the concept of standard of proof was considered to be not particularly helpful.

Munby J also referred to a number of cases concerning the approach of the Parole Board in determining whether it is safe to release a prisoner. In R v. Lichniak [2003] 1 AC 903, at para 16, Lord Bingham doubted “whether there is in truth a burden on the prisoner to persuade the Parole Board that it is safe to recommend release, since this is an administrative process requiring the board to consider all the available material and form a judgment”. In R (Sim) v. Parole Board [2003] EWCA Civ 1845, [2004] QB 1288, at para 42, Keene LJ accepted that “the concept of a burden of proof is inappropriate when one is involved in risk evaluation”. That was followed in R (Brooks) v. Parole Board [2004] EWCA Civ 80, where Kennedy LJ stated at para 28 that what the Parole Board must do is to decide in the light of all the relevant material placed before it whether, in the terms of section 28(6)(b) of the Crime (Sentences) Act 1997, it “is satisfied that it is no longer necessary for the protection of the public that the prisoner should be confined”. Although Munby J considered this to be the most obvious analogy with the present case, we accept Mr Bowen’s point that the relevant function of the Parole Board is treated as an administrative decision-making process in which the burden of proof has no part to play, so that it does not help directly with the question whether a standard of proof applies where there is a burden on the detaining authority. Nevertheless, it seems to us that the reasoning in such cases provides support also for the view that a risk evaluation of the kind engaged in by the Parole Board is not susceptible to proof to a defined standard.

On the other hand, it would seem that in B v. Chief Constable of Avon and Somerset Constabulary Lord Bingham considered that the magistrates’ court could apply the civil standard of proof (“with appropriate strictness”) in deciding whether the defendant “has acted … in such a way as to give reasonable cause to believe that an order … is necessary to protect the public from serious harm from him” (para 31, quoted above). A similar point may be made in relation to Gough v. Chief Constable of Derbyshire (see paras 90-93, quoted above), where the magistrates’ court had to be satisfied inter alia “that there are reasonable grounds to believe that making a banning order would help to prevent violence or disorder …”. In each case, as para 92 of Gough brings out, the focus is on evidence of past conduct, in relation to which there is no difficulty in applying a standard of proof. But in each case there is also an additional element of judgment and evaluation of risk. Yet no distinction was drawn in either case between that element and the fact-finding element as regards the applicability of the civil standard of proof.

Against that background we turn to consider the conclusion reached by Munby J at paras 101-102 of his judgment, to the effect that issues under sections 72 and 73 as to the appropriateness and necessity of continuing detention (as well as the alternative question under section 73 as to the appropriateness of the patient remaining liable to be recalled) are not susceptible to a defined standard of proof.

We agree with the judge that, in relation to those issues, the tribunal "is not … concerned so much with finding facts which are capable of exact demonstration but rather with a process of judgment, evaluation and assessment" (para 102). We also agree that this is very similar in nature to the process of judgment or evaluation referred to in cases such as McCann, Rehman and the Parole Board cases.

We would accept that the concept of a standard of proof is "not particularly helpful" (per Lord Hoffmann in Rehman, with emphasis added) in relation to such a process. But we would not go so far as to hold that there is no room for its application at all. An opinion on the appropriateness or necessity of continuing detention may in principle be held with different degrees of certainty, and it may be important for the tribunal to know what degree of certainty is called for. Under sections 72 and 73 the tribunal has to be "satisfied" as to the relevant matters. As Lord Lloyd observed in In re H (at p.576D-G), "is satisfied" is an expression with a range of meanings covering the criminal standard of proof ("satisfied so as to be sure"), through the civil standard ("satisfied on a balance of probabilities") to being a synonym for "concludes" or "determines" and therefore having an entirely neutral function. We see no absurdity in a tribunal having some doubt as to the appropriateness or necessity of continuing detention, yet being satisfied on the balance of probabilities that it is appropriate and necessary. Accordingly, as it seems to us, the standard of proof has a potential part to play in the decision-making process even in relation to issues that are the subject of judgment and evaluation. In practice, we would expect the tribunal generally either to form the requisite judgment or not to form it, without needing to have specific regard to any standard of proof. But the standard of proof provides a backdrop to the decision-making process and may have an important role in some cases.

Analysis of this issue is not helped by the fact that “proof” in the phrase “standard of proof” and “probabilities” in the phrase “balance of probabilities” are words which go naturally with the concept of evidence relating to fact, but are less than perfect with evaluative assessments. That is why the courts have started to speak of the “burden of persuasion”. Where a court has to be satisfied “on balance” in evaluative matters, it needs to be satisfied on the balance of the argument, where the argument depends in part on evidence (there is always going to be some factual substratum) and in part on evaluation. Since the evidence cannot be divorced from the argument, and since there is also argument on pure issues of fact, it is perfectly acceptable to refer to the whole process as one in which the court has to be satisfied on the balance of probabilities. In relation to the evaluative part of the process that may involve an element of shorthand, but it gives rise to no conceptual or practical difficulty.

A related consideration is that one of the issues arising for determination in the present context, namely the question under section 72(1)(b)(i) whether the person is suffering from a mental disorder “of a nature or degree which makes it appropriate for him to be liable to be detained in a hospital for medical treatment”, is a mixed question of fact and judgment or evaluation. The nature and degree of the mental disorder is a question of fact and is accepted to be susceptible to proof on the balance of probabilities. There is a certain artificiality in applying a standard of proof to that question but not to the related question whether the nature and degree of the disorder make detention appropriate. For example, a modest parallel can be drawn with the apparent view of the court in B v. Chief Constable of Avon and Somerset Constabulary that a standard of proof should be applied to the entire question whether the defendant “has acted … in such a way as to give reasonable cause to believe that an order … is necessary to protect the public from serious harm from him”.

We bear in mind too that, although the observations about the standard of proof in Reid v. Secretary of State for Scotland and R (H) v. London North and East Region Regional Mental Health Review Tribunal were obiter, in each case the court evidently saw no difficulty in applying the standard of proof to the full range of issues to be decided by the tribunal.

We also think it likely that the tribunal’s task will be made easier if, instead of dividing up the issues into matters that are susceptible to proof to a defined standard and those that are not, it approaches the entire range of issues by reference to the standard of proof on the balance of probabilities, whilst recognising that in practice the standard of proof will have a much more important part to play in the determination of disputed issues of fact than it will generally have in matters of judgment as to appropriateness and necessity.

For all those reasons, we respectfully differ from the conclusion reached by Munby J on this issue. We would hold that the tribunal should apply the standard of proof on the balance of probabilities to all the issues it has to determine. We would not, however, expect the difference between that approach and the approach favoured by Munby J to have much practical significance, given the limited role that the standard of proof will have in relation to matters of judgment and evaluation. Nor does the difference affect the outcome of the present appeal, since the tribunal appears to have applied the standard of proof on the balance of probabilities to all issues save the factual issues to which it applied a standard akin to the criminal standard.

This issue arises out of the passage at paras 117-120 of Munby J's judgment which we have quoted fully at para 24 above. In effect, the judge states in that passage that the tribunal is obliged to discharge a patient under section 72 or section 73 only when the detaining authority fails to satisfy it both as to the section 72(1)(b)(i) criterion and as to the section 72(1)(b)(ii) criterion.

As all parties before us accept, that is an error. The correct position is that those two criteria are cumulative and the tribunal is obliged to discharge a patient if the detaining authority fails to satisfy it as to either of them: see Reid v. Secretary of State for Scotland at pages 539-540.

Nothing, however, turns on this slip. The judge's observations about the default position were unnecessary for his decision and do not impact on the reasoning in the main part of the judgment. All that is needed, therefore, is to note the error so as ensure that it does not give rise to any confusion in the future.

Although our reasoning differs in some respects from that of Munby J, we are in agreement with his conclusion that the only misdirection by the tribunal in the case of AN was favourable to the patient and that AN’s judicial review application fell to be dismissed. It follows that this appeal must also be dismissed.