Copyright © 1995 Andrew L-T Choo and Manda Mellors. First Published in Web Journal of Current Legal Issues in association with Blackstone Press Ltd.
This paper is about the legal regulation of undercover police operations. Traditionally, such operations were considered to have been used mainly in the context of "victimless" crimes. However, a number of recent cases, notably the celebrated cases of Keith Hall and Colin Stagg in 1994, provide instances of undercover operations which apparently have as an aim the extraction of confessions from suspects. This paper examines, against this background, the adequacy of the relevant executive and judicial pronouncements on undercover police operations. Recommendations for reform of the current position are also made. The law as stated is that available to us in mid-April 1995.
The issue of undercover police operations was, during 1994, brought dramatically to public attention by the collapse of two murder prosecutions owing to the involvement of undercover police officers. The first of these was the trial of Keith Hall in Leeds Crown Court in March 1994 (see The Independent, 11 March 1994, p 3). Hall was charged with the murder of his wife, who had disappeared without a trace and whose body was never found. The West Yorkshire Police had used a police officer (who went by the name of "Liz") to reply to a lonely hearts advertisement and to pretend to fall in love with Hall. The Crown Prosecution Service fully approved of this operation. During a meeting between Hall and "Liz", he made a confession (which was tape recorded) to having strangled his wife and burnt her body. The trial judge held that the confession should be excluded under both ss 76(2)(b) and 78 of the Police and Criminal Evidence Act 1984, and Hall was acquitted. (Section 76(2)(b) requires the automatic exclusion from evidence of a confession obtained in circumstances conducive to unreliability, while s 78, to be discussed below, confers on trial judges a discretion to exclude unfair evidence.) Interestingly, the trial judge in Hall, having excluded the confession, took the unusual step of authorising the release of a full transcript of the confession - an action interpreted by many as a public statement that he was constrained by the law into excluding a confession which should really have been admitted.
The second case was that of Colin Stagg, who was charged with the infamous murder in July 1992 of Rachel Nickell on Wimbledon Common (see news items in The Guardian, 15 September 1994, and The Independent, 15 September 1994; feature in The Independent on Sunday, 18 September 1994, p 16). Again with the full backing of the CPS, an undercover police officer, "Lizzie", pretended to fall in love with Stagg and attempted to induce him, over a seven-month period, to say things which the police wanted him to say, in order that his psychological profile would conform with the psychological profile that they had prepared of the murderer. "Lizzie" informed Stagg that she had once been involved in the ritual sexual murder of a baby and a young woman. He, in turn, invented a story that he had once strangled a girl to death in the New Forest. No confession to Nickell's murder was actually made by Stagg, in spite of attempts to induce him to make such a confession involving, inter alia, threats to end the "relationship" unless he was the Wimbledon Common murderer. He did tell "Lizzie" that he had been on Wimbledon Common when Nickell was murdered, and that he would take her to the murder scene. However, he took her to the wrong place, and informed her that Nickell had been raped, which was not the case. As is well known, the evidence of tape-recorded conversations, and letters, was excluded, and Stagg was acquitted. The trial judge described the undercover operation as "misconceived" and as "betray[ing] not merely an excess of zeal, but a blatant attempt to incriminate a suspect by positive and deceptive conduct of the grossest kind." He remarked that "[a]ny legitimate steps taken by the police and the prosecuting authorities to bring the perpetrators to justice are to be applauded - but the emphasis must be on the word legitimate." Here, there had been "a skilful and sustained enterprise to manipulate the accused, sometimes subtly, sometimes blatantly."
What the Hall and Stagg cases appear to demonstrate is that undercover police operations are now being used to a greater extent outside the area of consensual crimes (or, at least, that their use outside this area is coming to light to a much greater extent). Undercover operations are now being used in the investigation of serious offences such as murder, and apparently with a view to obtaining a confession from the suspect if at all possible. The purpose of this paper is to examine the legal regulation of undercover police operations in the light of these recent developments, focusing upon both executive and judicial articulations on the topic.
Forward to references.
Back to contents.
a. No member of a police force, and no public informant, should counsel, incite or procure the commission of a crime.
b. Where an informant gives the police information about the intention of others to commit a crime in which they intend that he shall play a part, his participation should be allowed to continue only where -i. he does not actively engage in planning and committing the crime;
ii. he is intended to play only a minor role; and
iii. his participation is essential to enable the police to frustrate the principal criminals and to arrest them (albeit for lesser offences such as attempt or conspiracy to commit the crime, or carrying offensive weapons) before injury is done to any person or serious damage to property.
The informant should always be instructed that he must on no account act as agent provocateur, whether by suggesting to others that they should commit offences or encouraging them to do so....
It can be seen that the focus of the guidelines is upon the danger that undercover operations may actually result in the creation of crime, or may encourage or stimulate crime. This is, of course, a very real danger, for as Gary Marx has pointed out in his book on undercover operations in the United States (Marx 1988, pp 126-7), there are several ways in which undercover activity may "amplify" crime:
Effectively create crime is unduly narrow. This may, of course, be a primary concern in relation to "old- style" undercover operations that are targeted at consensual crimes. Such operations are essentially facilitative operations in the sense that an opportunity is offered (and inducements may be given) to the suspect to commit an offence of a type that the police suspect is being continually committed - for example, the selling of illegal drugs. But the danger of creation of crime is clearly not a major consideration in the context of operations like those used in, for example, the Hall and Stagg cases, where the aim is to gain evidence of a single offence which has already been committed. The issue here is the more fundamental one of whether targeting individuals for the purpose of obtaining evidence is consistent with general concepts of fair play and with the presumption of innocence. There is, therefore, an urgent need for the Home Office guidelines to be revamped, updated, and extended considerably.
In addition, there may exist internal guidelines within individual police forces which deal with the issue of undercover police operations. It would appear, for example, that within the Metropolitan Police there are guidelines based on an opinion specially solicited from counsel in 1986, revised regularly in the light of new caselaw, and printed in the police pocket book (Ashworth 1994, p 96).
Forward to references.
Back to contents.
In England, after some initial uncertainty in the law, the House of Lords held in 1979, in the case of R v Sang  AC 402, that entrapment evidence could not be excluded in the discretion of the trial judge. The House of Lords reasoned that this must be the case because there is no substantive defence of entrapment in English law, and the exclusion of entrapment evidence may be tantamount to the recognition of a defence of entrapment through the back door. The fact of entrapment, the House of Lords held, could be taken into account only in the mitigation of sentence. The reasoning of the House of Lords in Sang has been subjected to much criticism which need not be repeated here (see eg Choo 1993, pp 151-3, and references cited therein). On 1 January 1986, PACE came into force in England and Wales. Section 78 of this Act provides: "In any proceedings the court may refuse to allow evidence on which the prosecution proposes to rely to be given if it appears to the court that, having regard to all the circumstances, including the circumstances in which the evidence was obtained, the admission of the evidence would have such an adverse effect on the fairness of the proceedings that the court ought not to admit it." The issue arose in a series of cases as to whether, even though the common law discretion to exclude evidence for unfairness could not be utilised to exclude evidence obtained by entrapment, the section 78 discretion could be used for that purpose. In R v Edwards  Crim LR 45 (transcript available on LEXIS) it was argued in the Court of Appeal that the police undercover agents involved in the case had been agents provocateurs and that their evidence should accordingly have been excluded under s 78. Russell LJ held that there was no evidence that the undercover officers were agents provocateurs, and that no question of s 78 could therefore arise. In any event there was, he pointed out, conflicting authority in the Court of Appeal as to whether s 78 had any application "in the event of witnesses being demonstrated to be agents provocateurs." First, there was the decision in R v Harwood  Crim LR 285 (transcript available on LEXIS), where Stocker LJ had taken the narrow view that s 78 could not be utilised in entrapment cases. Stocker LJ pointed out that there was no substantive defence of entrapment in English law, and that:
[s]ince entrapment is not a defence to a criminal offence, it would seem to us to follow that an Act of Parliament dealing with evidential matters cannot be so interpreted as to bring about the conclusion that the substantive rule of law can be abrogated by evidential means....Trying by evidential means to achieve the same effect as if a substantive rule of law did not exist does not appear to us to be a possible one, bearing in mind the construction of section 78. In our view the rule that entrapment is no defence is a matter of substantive law which cannot be evaded by a procedural device of preventing the prosecution from adducing evidence of the commission of the offence.
However, in the later case of R v Gill and Ranuana  Crim LR 358 (transcript available on LEXIS), Lord Lane CJ expressed reservations about the correctness of the observations made in Harwood. In Edwards, Russell LJ chose to remain non-committal on this issue, saying simply that "[b]ecause of our finding that there was not a shred of evidence here to demonstrate that the officers involved in the instant appeal were in truth agents provocateurs it is unnecessary for us to resolve the conflict that has arisen in the two unreported cases to which we have just referred and we do not attempt to make any such ruling preferring one case to the other."
Forward to references.
Back to contents.
The Court was careful, however, to enter the following caveat (at p 991):
It would be wrong for police officers to adopt or use an undercover pose or disguise to enable themselves to ask questions about an offence uninhibited by the requirements of the Code and with the effect of circumventing it. Were they to do so, it would be open to the judge to exclude the questions and answers under section 78 of the Act of 1984.
This idea of the inappropriateness of using an undercover police operation effectively to circumvent the requirements of Code C accounts for the decision to exclude the evidence in the Keith Hall and Colin Stagg cases. Looking at Stagg, for example, it is possible that a more vulnerable person might well have confessed falsely to killing Nickell in order to impress "Lizzie" when she specifically threatened to end the "relationship" if he was not the murderer. A number of possible reasons, associated with his or her "psychological vulnerabilities", may explain why a suspect may confess to a crime which he or she has not committed (see generally Gudjonsson 1992). Stagg may well have been more vulnerable to pressure, and more in need of the protection afforded by the Codes of Practice, than someone being interviewed at a police station and aware of the risks which answering police questions might carry.
Just one month after the decision in Christou and Wright, the Court of Appeal handed down its decision in R v Bryce  4 All ER 567. It was held (p 572) that evidence of the conversations between Bryce and an undercover police officer should have been excluded because these conversations went directly to the crucial issue of guilty knowledge, they were hotly disputed, and there was no contemporary record of them (unlike in Christou, where the existence of tape and film recordings of the entire interview eliminated the possibility of concoction). Thus, the Court of Appeal in Bryce was effectively holding that the evidence in question should have been excluded because of its possible unreliability, rather than because of considerations of extrinsic policy which do not pertain to reliability.2 Any temptation, however, to focus solely upon the reliability issue must be resisted at all costs.
The case of R v Smurthwaite and Gill  1 All ER 898 represents the most significant English case on undercover police operations to date. The Court of Appeal reiterated in this case that entrapment is not a defence known to the substantive criminal law in England. However, entrapment is not irrelevant to the application of s 78 of PACE (p 902). A non-exhaustive list of factors that might be considered by a trial judge in determining whether evidence obtained in an undercover operation should be excluded under s 78 was provided (p 903):
Was the officer acting as an agent provocateur in the sense that he was enticing the defendant to commit an offence he would not otherwise have committed? What was the nature of any entrapment? Does the evidence consist of admissions to a completed offence, or does it consist of the actual commission of an offence? How active or passive was the officer's role in obtaining the evidence? Is there an unassailable record of what occurred, or is it strongly corroborated?...[A] further consideration for the judge in deciding whether to admit an undercover officer's evidence is whether he has abused his role to ask questions which ought properly to have been asked as a police officer and in accordance with the codes.
Harwood has thus been effectively laid to rest, and the possibility is left open for entrapment evidence to be excluded under s 78, especially if the entrapment was of a serious nature. Beyond this, however, what is required is a consideration of all relevant circumstances surrounding the undercover operation: whether a confession was obtained; whether the role played by the undercover officer was "active" or "passive"; whether there exists a reliable record of what occurred (or at least strong corroboration); and whether the undercover operation was aimed at circumventing Code C. No guidance as to the relative weighting to be accorded to these factors is, however, provided.
Two post-Smurthwaite and Gill decisions of the Court of Appeal deserve attention. In the first, R v Latif and Shahzad  Crim LR 750, the appellants appealed against their convictions of being knowingly concerned in the importation of heroin on the basis, inter alia, that the evidence of an undercover officer should have been excluded under s 78. Dismissing the appeal, the Court remarked that "[t]he Parliament that enacted s 78...for the purpose of protecting the innocent might have been surprised to hear it invoked on behalf of an importer of heroin worth £3.2 million into this country in order to exclude the evidence against him on the ground that he was encouraged by an agent of the British government." Thus, the seriousness of the offence charged might well constitute another factor to be taken into account in considering the s 78 discretion.
Offence seriousness was also a significant factor in the decision of the Court of Appeal in R v Khan (Sultan)  4 All ER 426. On the day after his arrival from Pakistan, the appellant, on being interviewed at a police station, denied any offence and declined to answer most of the questions put to him. He was released without charge. Some six months later, he visited the home of a person whom the police suspected of being involved in the supply of heroin on a large scale. As a result of these suspicions, the police had installed an aural surveillance device on the exterior of the property, without the knowledge or consent of the owner or occupier of the property. The police obtained a tape recording of a conversation which took place between the appellant and others in the house, during which the appellant said things which plainly demonstrated his involvement in the importation of heroin. He was accordingly arrested and charged. Dismissing his appeal against conviction, the Court of Appeal held that the fact that the attachment of the listening device involved the invasion of privacy and a civil trespass, and caused slight damage to the property, was outweighed by other considerations, such as the fact that the police were acting in accordance with the relevant Home Office guidelines, and the fact that "what was under investigation was a type of criminal conduct of great gravity" (p 437). Thus, as in Latif and Shahzad, the seriousness of the offence charged was considered to be a factor favouring admission of the evidence. The issue of the seriousness of drugs offences was adverted to again by the Court of Appeal in R v Cadette  Crim LR 229, in holding that the trial judge had been right not to exclude, under s 78, the evidence of a telephone conversation taped in an undercover police operation. What is required now is clearer guidance from the courts as to the precise role which offence seriousness might have to play in determinations under section 78.3 It is noteworthy that, in Hall and Stagg, exclusion was ordered in spite of the fact that the offence charged in both cases was murder, whereas in Latif and Shahzad and Khan, the fact that the offences were serious drugs offences was apparently sufficient to tip the balance in favour of admission. A further point which should be noted about Khan is that the surveillance operation was not designed specifically to obtain evidence against the appellant: "[i]t was not expected or foreseen that the appellant would visit the premises" (p 429). If, however, it had been the case that the surveillance operation was embarked upon specifically to gather evidence against the appellant because the interview at the police station months earlier had yielded no results, then, in our view, the evidence should have been excluded. The Appeal Committee of the House of Lords has granted leave to appeal in Khan (p 438), and the decision of the House of Lords is awaited with interest. It would be unfortunate if, while becoming increasingly critical of undercover operations (as evidenced by the decisions to exclude evidence in Hall and Stagg), the English courts displayed no willingness, even if the circumstances of the case warranted it, to exclude evidence obtained in surveillance operations. Such unwillingness might well send a signal to the police to move away from covert operations of the former type, and towards covert operations of the latter type. One might find, for example, that if it was suspected that drug dealing was taking place on certain premises, the preference of the police would be to conduct a surveillance operation rather than to employ an undercover agent to infiltrate the organisation. The serious invasion of privacy which surveillance operations may entail must not be overlooked, and thus it should not be assumed that surveillance operations are invariably more acceptable than undercover operations.
Forward to references.
Back to contents.
Forward to references.
Back to contents.
Beyond this, however, a determination of whether evidence should be excluded on the basis of the Smurthwaite factors is an inherently difficult one, given the lack of clear guidance as to the permissible limits of undercover police activity. In particular, what exactly is meant by circumvention of the requirements of Code C? Put another way, in precisely what circumstances is it appropriate to question suspects in the course of an undercover operation, and in precisely what circumstances is it inappropriate? We would endorse the calls of writers such as Sybil Sharpe and Andrew Ashworth for the introduction of legislation, or a Code of Practice, governing undercover police operations (Sharpe 1994, p 804; Ashworth 1994, p 96). The s 78 discretion has the merit of flexibility and thus of allowing all relevant factors in the particular case to be taken into account, but it needs to be accompanied by firm guidelines as to what police conduct is or is not proper in the first place. The existence of such guidelines would facilitate exercise of the discretion in a coherent and informed manner. In other words, trial judges need to know whether the impugned conduct is proper or not, in order to be able to exercise the section 78 discretion meaningfully (see generally Bradley 1993). The existence of Codes C and D of the Codes of Practice, for example, has generated a substantial body of caselaw concerning the exclusion of evidence under section 78 for violations of these Codes (see generally Birch 1994, pp 89-90; Feldman 1990). There will naturally be disagreement as to the exact content of a Code of Practice for undercover police operations. Nonetheless, the existence of such a Code would mean that no longer would the courts (and the police) be forced to operate in the dark to the extent that they are having to at present.
Back to contents.
2. See generally Galligan 1988, p 255 for a succinct discussion of the distinction between intrinsic and extrinsic policy considerations in the law of criminal evidence. Return to text.
3. The High Court of Australia (Bunning v Cross (1978) 141 CLR 54) and Supreme Court of Canada (R v Collins (1987) 56 CR (3d) 193) have expressly considered the relevance of offence seriousness to determinations of whether improperly-obtained prosecution evidence should be excluded from criminal trials. Return to text.
Ashworth, A (1994) The Criminal Process: An Evaluative Study (Oxford: Clarendon Press).
Birch, D (1994) "Excluding Evidence from Entrapment: What is a "Fair Cop"?" 47 Current Legal Problems 73.
Bradley, C M (1993) The Failure of the Criminal Procedure Revolution (Philadelphia: University of Pennsylvania Press).
Choo, A L-T (1993) Abuse of Process and Judicial Stays of Criminal Proceedings (Oxford: Clarendon Press).
Feldman, D (1990) "Regulating Treatment of Suspects in Police Stations: Judicial Interpretation of Detention Provisions in the Police and Criminal Evidence Act 1984"  Criminal Law Review 452.
Galligan, D J (1988) "More Scepticism about Scepticism" 8 Oxford Journal of Legal Studies 249.
Gudjonsson, G H (1992) The Psychology of Interrogations, Confessions and Testimony (Chichester: John Wiley & Sons).
Marx, G T (1988) Undercover: Police Surveillance in America (Berkeley: University of California Press).
Sharpe, S (1994) "Covert Police Operations and the Discretionary Exclusion of Evidence"  Criminal Law Review 793.