Judgments - Regina v Looseley

(back to preceding text)

    34. I consider that the detailed analysis of particular circumstances which would justify ordering a stay or which would not do so will be helpful in reaching a conclusion but no two cases are likely to be exactly the same in all their circumstances. I would wish to stress that the ultimate question for the tribunal facing a case in which entrapment is alleged is whether, in the words of my noble and learned friend, Lord Nicholls of Birkenhead, the state through its agents had lured the accused into committing an act or acts forbidden by law for which the state is now seeking to prosecute him. I agree that it is difficult in advance to give a precise and exhaustive definition of what the question means but after the facts have been either agreed or proved, and the helpful illustrations given by my noble and learned friends are taken into account, the tribunal must decide the case by applying the general principle on which these illustrations are based and whose meaning they elucidate.


My Lords,

    35. The question in both of these appeals is whether the English law concerning entrapment is compatible with the Convention right to a fair trial. In my opinion it is. I have had the advantage of reading in draft the reasons of Lord Nicholls of Birkenhead for reaching the same conclusion. I agree with them.

    English law on entrapment

    36. Entrapment occurs when an agent of the state - usually a law enforcement officer or a controlled informer - causes someone to commit an offence in order that he should be prosecuted. I shall in due course have to refine this description but for the moment it will do. In R v Latif [1996] 1 WLR 104, 112 Lord Steyn said that English law on the subject was now settled. It may be summarised as follows. First, entrapment is not a substantive defence in the sense of providing a ground upon which the accused is entitled to an acquittal. Secondly, the court has jurisdiction in a case of entrapment to stay the prosecution on the ground that the integrity of the criminal justice system would be compromised by allowing the state to punish someone whom the state itself has caused to transgress. Thirdly, although the court has a discretion under section 78 of the Police and Criminal Evidence Act 1984 to exclude evidence on the ground that its admission would have an adverse effect on the fairness of the proceedings, the exclusion of evidence is not an appropriate response to entrapment. The question is not whether the proceedings would be a fair determination of guilt but whether they should have been brought at all. I shall briefly enlarge upon these three points.

(a)     Not a defence

    37. The fact that the accused was entrapped is not inconsistent with his having broken the law. The entrapment will usually have achieved its object in causing him to do the prohibited act with the necessary guilty intent. So far as I know, the contrary view is held only in the Federal jurisdiction of the United States. It is unnecessary to discuss the cogent criticisms which have been made of this doctrine, notably by Frankfurter J in his dissenting judgment in Sherman v United States (1958) 356 US 369, because it has never had any support in authority or academic writing in this country. Indeed, the majority judgment of Rehnquist J in United States v Russell (1973) 411 US 423, 433, which describes the criticisms as "not devoid of appeal" suggests that its survival in the Federal jurisdiction owes more to stare decisis and its perceived constitutional and pragmatic advantages than to its intellectual coherence.

(b)     The jurisdiction to stay proceedings

    38. The court's assertion of such a jurisdiction is of recent origin. It was not even discussed as a possible response to entrapment by the Law Commission in its Criminal Law Report on Defences of General Application (1977) (Law Com No 83), which dealt with entrapment at pp 32-53. Nor was it mentioned by the House of Lords in R v Sang [1980] AC 402, when it was decided that the court had no discretion to exclude evidence on the ground that the offence had been procured by entrapment or that the evidence had been unfairly obtained. It seems fairly clear, however, that if anyone had suggested such a jurisdiction, it would have been emphatically rejected. Lord Diplock dismissed the notion of a discretion to exclude evidence of an offence procured by entrapment as a "procedural device" to evade the rule that entrapment was not a substantive defence. He would almost certainly have taken the same view of a stay of proceedings, as Mason CJ, Deane and Dawson JJ later did in their joint judgment in Ridgeway v The Queen (1995) 184 CLR 19, 40. The House in Sang's case said that the only constitutionally proper way in which the court could mark its disapproval was by admonishing the police (as Lord Goddard CJ had done in Brannan v Peek [1948] 1 KB 68, 72 and Browning v JWH Watson (Rochester) Ltd [1953] 1 WLR 1172, 1177) and by imposing a light or nominal sentence. It was for the police authorities to take disciplinary action or prosecute policeman or informants who took part in the crime.

    39. This disclaimer of court responsibility for convicting a person who was on trial in consequence of an abuse of state power was not to everyone's taste. The Canadian Supreme Court, when it came to consider the matter, thought that it was not good enough. In Amato v The Queen (1982) 69 CCC (2d) 31, 74, Estey J said:

    "The repugnance which must be experienced by a court on being implicated in a process so outrageous and shameful on the part of the State cannot be dissipated by the registration of a conviction and the imposition afterwards of even a minimum sentence. To participate in such injustice up to and including a finding of guilt and then to attempt to undo the harm by the imposition of a lighter sentence, so far from restoring confidence in the fair administration of justice, would contribute to the opposite result."

    These views accorded with much of what was said by Frankfurter J in his dissenting judgment in Sherman v United States (1958) 356 US 369. They were subsequently elaborated by Lamer J, speaking for the unanimous Supreme Court of Canada, in R v Mack (1988) 44 CCC (3d) 513. A stay should be granted not because the accused was not guilty or because he could not receive a fair trial or to discipline the police but to protect the integrity of the criminal justice system.

    40. The case which eventually took English law down the same path was R v Horseferry Road Magistrates' Court, Ex p Bennett [1994] 1 AC 42, in which the House of Lords decided that a criminal court had power to inquire into allegations that the accused had been kidnapped abroad by authorities acting in collusion with the UK police and, if it found them proved, had a discretionary jurisdiction to stay the proceedings. Lord Griffiths said that the jurisdiction was necessary to enable the courts to refuse to countenance behaviour which threatened basic human rights or the rule of law. The stay is sometimes said to be on the ground that the proceedings are an abuse of process, but Lord Griffiths described the jurisdiction more broadly and, I respectfully think, more accurately, as a jurisdiction to prevent abuse of executive power.

    41. It was on the authority of Bennett's case that the House decided in R v Latif [1996] 1 WLR 104 that in principle a stay could be granted on the grounds of entrapment. Lord Steyn said, at p 112, that the court should exercise the jurisdiction when "[w]eighing countervailing considerations of policy and justice", the judge considers that the bringing of the prosecution "amounts to an affront to the public conscience." I shall try later to analyse in more detail what this means.

(c)     The section 78 discretion

    42. The section 78 discretion enables the judge to safeguard the fairness of the trial. But the entrapped defendant is not ordinarily complaining that the admission of certain evidence would prejudice the fairness of his trial. He is saying that whatever the evidence, he should not be tried at all. The appropriate remedy, if any, is therefore not the exclusion of evidence but a stay of the proceedings. The distinction was clearly made by the Law Commission in its 1977 Report at para 5.29, by Andrew Choo in Abuse of Process and Judicial Stays of Criminal Proceedings (1993), pp. 164-166 and by Potter LJ in R v Shannon [2001] 1 WLR 51.

    43. On the other hand, if the court is not satisfied that a stay should be granted and the trial proceeds, the participation of state agents in the commission of the crime may well be relevant to the exercise of the discretion under section 78. As Potter LJ pointed out in Shannon's case at p 68, the question at that stage is not whether the proceedings should have been brought but whether the fairness of the proceedings will be adversely affected by, for example, admitting the evidence of the agent provocateur or evidence which is available as a result of his activities.

    "So, for instance, if there is good reason to question the credibility of evidence given by an agent provocateur, or which casts doubt upon the reliability of other evidence procured by or resulting from his actions, and that question is not susceptible of being properly or fairly resolved in the course of the proceedings from available, admissible and 'untainted' evidence, then the judge may readily conclude that such evidence should be excluded."

    44. This question of whether the proceedings should be stayed on the grounds of entrapment should logically be decided before the proceedings have begun. But sometimes proceedings are not conducted entirely logically and an application to exclude evidence under section 78 may be in substance a belated application for a stay. If so, it should be treated as such and decided according to the principles appropriate to the grant of a stay.

    The European jurisprudence

    45. These appeals raise the question of whether the exercise of the power to stay proceedings as affirmed in R v Latif [1996] 1 WLR 104 is sufficient to satisfy the right to a fair trial under Article 6 of the Convention. It is clear from the decisions of the European Court of Human Rights, which must be taken into account under section 2(1)(a) of the Human Rights Act 1998, that the right is not confined to a fair determination of the question of guilt. It is also a right not to be tried at all in circumstances in which this would amount to an abuse of state power.

    46. This appears most clearly from the decision in Teixeira de Castro v Portugal (1998) 28 EHRR 101 in which the court decided that "right from the outset, the accused was definitively deprived of a fair trial" (see p 116, para 39 of the judgment) because his conviction was for a drugs offence which had been "instigated" by two police officers. This is the situation of entrapment in which, in an appropriate case, an English court would order a stay of proceedings under the principle in Latif case. But Mr O'Connor, who appeared for Mr Looseley in his appeal and Mr Emmerson, who appeared for the acquitted person in the Attorney General's reference, both submitted that the principles upon which the power to order a stay was exercised in England did not satisfy the requirements of the Convention as stated in de Castro's case. In order to examine this submission, I must first analyse the current English law and then consider whether it is consistent with what the European Court has decided.

    The application of the entrapment doctrine in England

    47. At the highest level of abstraction, the English principles are easy to state. The court is concerned with whether there has been an abuse of executive power, something which is, as Lord Steyn said, in Latif's case [1996] 1 WLR 104, 112, an affront to the public conscience or, as Estey J said in Amato's case (at (1982) 69 CCC (2d) 31,61), whether the court's participation in such proceedings would bring the administration of justice into disrepute. But to leave the matter entirely at that level would incur the criticism, levelled by Lord Diplock in Sang's case [1980] AC 402, 431, that "[w]hat is unfair, what is trickery in the context of the detection and prevention of crime, are questions which are liable to attract highly subjective answers" or the reproach of Rehnquist J that the court was claiming a " 'chancellor's foot' veto over law enforcement practices of which it did not approve": see United States v Russell (1973) 411 US 423, 435.

    48. The objections to entrapment are certainly more specific than a generalised fastidiousness about police practices. The theme which runs through all discussions of the subject is that the state should not instigate the commission of criminal offences in order to punish them. But what counts for this purpose as instigation? An examination of the authorities demonstrates, in my opinion, that one cannot isolate any single factor or devise any formula that will always produce the correct answer. One can certainly identify a cluster of relevant factors but in the end their relative weight and importance depends upon the particular facts of the case.

    49. Limited assistance can therefore be gained from distinctions which restate the question rather than provide a criterion for answering it. For example, it has been said that a policeman or paid informer should not act as an agent provocateur; an expression used to signify practices employed by foreigners unacquainted with English notions of decency and fair play: see para 104 of the Report of the Royal Commission on Police Powers and Procedure (1929) Cmd 3297. But what exactly is an agent provocateur? The Royal Commission said that he was "a person who entices another to commit an express breach of the law which he would not otherwise have committed, and then proceeds or informs against him in respect of such offence." This is helpful so far as it goes, but one still has to say what amounts to enticing and what it means to say that the breach of the law would not otherwise have been committed. In other words, the definition assumes but does not define the standards of decency and fair play with which the activity of the agent provocateur is contrasted.

    Causing and providing an opportunity

    50. Many cases place emphasis upon the question of whether the policeman can be said to have caused the commission of the offence, rather than merely providing an opportunity for the accused to commit it with a policeman rather than in secrecy with someone else. There is no doubt that this will usually be a most important factor deciding whether or not the police have overstepped the line between legitimate crime detection and unacceptable crime creation. But a note of caution must be sounded. First, as Lord Steyn said in Latif's case [1996] 1 WLR 104, 111, it is important but not necessarily decisive. Other factors, some of which I shall mention in a moment, may have to be taken into account as well. Secondly, a good deal will depend upon what is accepted as evidence that the accused would have committed the offence with someone else.

    51. A good example of a straightforward application of the distinction between causing the commission of the offence and providing an opportunity for it to be committed is the case of Nottingham City Council v Amin [2000] 1 WLR 1071. Mr Amin owned a taxi which was not licensed to ply for hire in Nottingham. Two plain clothes policemen who saw him driving down a street in Nottingham in the middle of the night flagged him down. He stopped and upon request agreed to take them to the destination which they named. When they arrived, the policemen paid the fare and then charged him with the offence of plying for hire without a licence.

    52. The stipendiary magistrate excluded the evidence of the policemen under section 78 of the 1984 Act on the ground that they had been agents provocateurs. The prosecutor appealed. As I have already indicated, I think that the use of section 78 was on any view inappropriate. The accused was not saying that the admission of the evidence would prejudice the fairness of his trial. His case was that he should not have been charged with an offence which the policemen had induced him to commit. The proper remedy, if any, would have been a stay of proceedings. But the point is academic because the Court of Appeal treated the case as if it had concerned an application for a stay and applied the same principles.

    53. Lord Bingham of Cornhill CJ observed (at p 1075) that to call the policemen agents provocateurs was to express the magistrate's conclusion rather than his reasoning. The question was whether they could properly be so described. This depended on whether the case fell on one side or the other of a line which Lord Bingham formulated in the following terms at pp 1076-1077:

    "On the one hand it has been recognised as deeply offensive to ordinary notions of fairness if a defendant were to be convicted and punished for committing a crime which he only committed because he had been incited, instigated, persuaded, pressurised or wheedled into committing it by a law enforcement officer. On the other hand it has been recognised that law enforcement agencies have a general duty to the public to enforce the law and it has been regarded as unobjectionable if a law enforcement officer gives a defendant an opportunity to break the law, of which the defendant freely takes advantage, in circumstances where it appears that the defendant would have behaved in the same way if the opportunity had been offered by anyone else."

    54. In referring to whether the defendant would have behaved in the same way if the opportunity had been offered by anyone else, Lord Bingham obviously did not mean only that the defendant would have responded in the same way to someone who was not a policeman. Since the defendant in such cases ex hypothesi does not know that he is dealing with a policeman, such a condition would invariably be satisfied. What he meant was that the policemen behaved like ordinary members of the public in flagging the taxi down. They did not wave £50 notes or pretend to be in distress.

    55. The test of whether the law enforcement officer behaved like an ordinary member of the public works well and is likely be decisive in many cases of regulatory offences committed with ordinary members of the public, such as selling liquor in unlicensed quantities (Director of Public Prosecutions v Marshall [1988] 3 All ER 683) selling videos to children under age (Ealing London Borough Council v Woolworths plc [1998] Crim LR 58, DC, and operating a private hire vehicle without a licence (Taunton Deane Borough Council v Brice (DC unreported 10 July 1997). But ordinary members of the public do not become involved in large scale drug dealing, conspiracy to rob (R v Mealey and Sheridan (1974) 60 Cr App R 59 or hiring assassins (R v Gill [1989] Crim LR 358; R v Smurthwaite [1994] All ER 898. The appropriate standards of behaviour are in such cases rather more problematic. And even in the case of offences committed with ordinary members of the public, other factors may require a purely causal test to be modified.

    Suspicion and supervision.

    56. In the case of some regulatory offences, the effective administration of the law may require enforcement officers to have the power to make random tests. But normally it is not considered a legitimate use of police power to provide people not suspected of being engaged in any criminal activity with the opportunity to commit crimes. The only proper purpose of police participation is to obtain evidence of criminal acts which they suspect someone is about to commit or in which he is already engaged. It is not to tempt people to commit crimes in order to expose their bad characters and punish them.

    57. This point was made very clearly by Buxton J in Taunton Deane Borough Council v Brice (DC unreported 10 July 1997) when Mr Brice complained that two council employees had tempted him to carry them for reward in his unlicensed vehicle by offering the opportunity of a long and lucrative journey. The judge attached importance to the fact that the council were carrying out a bona fide investigation into complaints about Mr Brice's activities:

    "That decision [to use entrapment] was not made for an ulterior motive or in the hope of persuading someone who was not breaking the law to start doing so, but rather to see whether or not evidence was available upon which a prosecution could properly be mounted."

    58. These facts may be contrasted with the example given by Lamer J in R v Mack (1988) 44 CCC (3d) 513, 553 of the police officer who "decides that he wants to increase his performance in court". To this end, he plants a wallet with money in an obvious location in a park and keeps watch. This is unacceptable behaviour because the policeman is preying on the weakness of human nature to create crime for an improper purpose.

    59. The principle that the police should prevent and detect crime, not employ themselves in creating it, requires some modification of the causal test. In the case of the planted wallet, the policeman has not "incited, instigated, persuaded, pressurised or wheedled" the hapless offender into stealing it. He has provided him with an opportunity of which he has taken free advantage. Nevertheless, for a different reason, the policeman's conduct is an abuse of state power which the judicial branch of government should not countenance.

    60. Closely linked with the question of whether the police were creating or detecting crime is the supervision of their activities. To allow policemen or controlled informers to undertake entrapment activities unsupervised carries great danger, not merely that they will try to improve their performances in court, but of oppression, extortion and corruption. As we shall see, the European Court in Teixeira de Castro v Portugal (1998) 28 EHRR 101 attached great importance to the fact that the police were not acting in the course of an officially authorised investigation.

    61. The need for reasonable suspicion and proper supervision are both stressed in the Undercover Operations Code of Practice issued jointly by all UK police authorities and HM Customs and Excise in response to the Human Rights Act 1998. It deals with the employment of "undercover officers", "test purchasers" and "decoys". Undercover officers are defined as specially trained law enforcement officers working incognito "under direction in an authorised investigation" to infiltrate an existing conspiracy, arrest suspected criminals or counter a threat to national security. Test purchasers are appropriately trained law enforcement officers who seek "by means of authorised activity, to establish the nature and/or availability of a commodity or service, the possession, supply or use of which involves an offence". Test purchasers are used mainly in the drug trade. Decoys are officers who place themselves passively in a position to become a victim of crime for the purpose of arresting the offender.

    62. Undercover officers may be used only in connection with national security or serious crime and in cases in which the desired result cannot reasonably be achieved by other means: see paragraph 2.2. Authorisation must be given by an Assistant Chief Constable or a Commander in the Metropolitan or City of London Police.

    63. The use of a test purchaser must be authorised by a superintendent in the police or National or Scottish Crime Squads and paragraph 3.2 states that the authorising officer must be satisfied that a test purchase is -

    "required in support of an investigation into a criminal offence concerning the possession, supply or use of a commodity or service and that reasonable grounds have been established prior to the deployment of a test purchaser to suspect that such an offence is being committed."

    64. The authorising officer must also be satisfied that the desired result of the test purchase cannot reasonably be achieved by other means and a note for guidance (Note 3A) emphasises that -

    "Test purchase should not be used as a speculative means of search for the existence of a commodity or service where no other reasonable grounds exist to suspect that criminal offences have been or are being committed."

    65. The requirement of reasonable suspicion does not necessarily mean that there must have been suspicion of the particular person who happens to have committed the offence. The police may, in the course of a bona fide investigation into suspected criminality, provide an opportunity for the commission of an offence which is taken by someone to whom no suspicion previously attached. This can happen when a decoy (human or inanimate) is used in the course of the detection of crime which has been prevalent in a particular place. Lamer J in R v Mack (1988) 44 CCC (3rd) 513, 553 gave the example of the police planting a handbag in a bus terminal where numerous thefts have recently taken place. A real example in England was Williams v Director of Public Prosecutions (1994) 98 Cr App R 209, in which the police were investigating thefts from vehicles in Essex. They left an unattended Transit van with the back door open and cartons of cigarettes visible. When the appellants stole the cigarettes, they were arrested. Although the judgment contains (at p 213) some reference to causal reasoning ("they were tricked into doing what they wanted to do") I do not think that in such a case causation provides a sufficient answer. If the trick had been the individual enterprise of a policeman in an area where such crime was not suspected to be prevalent, it would have been an abuse of state power. It was justified because it was an authorised investigation into actual crime and the fact that the defendants may not have previously been suspected or even thought of offending was their hard luck.

    The nature of the offence

    66. The provision in the Code of Practice which requires the authorising officer to be satisfied that the desired result of deploying an undercover officer or test purchaser cannot reasonably be achieved by other means shows that the justification for such methods will partly depend upon the nature of the offence. Consensual offences such as dealing in unlawful substances or offences with no immediate victim like bribery or offences which victims are reluctant to report are the most obvious candidates for such methods. So is the infiltration of conspiracies. But the fact that the offence is a serious one is not in itself a sufficient ground for the police to ignore the provisions of the Code or the courts to condone their actions by allowing the prosecution to proceed.


    67. The United States Federal doctrine that entrapment is a ground for acquittal does not protect a person who was "predisposed" to commit the offence. Predisposition is regarded as showing that the defendant had the necessary guilty intent and was not lured by entrapment into committing an offence which would not otherwise have occurred to him. Predisposition may be proved by, among other things, previous convictions or similar fact evidence. The prospect of such matters going before the jury explains why the defence is rarely invoked in Federal criminal prosecutions. The result is that people with criminal records are fair game for entrapment.

continue previous