Diminishing due process: the growing influence of crime control values
Recent development of Code D is consistent with a general trend that some commentators have identified in the evolution of the criminal procedure comprised in the Police and Criminal Evidence Act 1984 and the Codes of Practice issued under the Act. Sanders and Young53 have argued that the Act was introduced at a time when criminal justice was steadily becoming more illiberal in line with a general drift towards a ‘law and order’ society54 and was seen as being necessary to deal with discontent and disorder created within large disadvantaged sections of the community.55 Against this backdrop they suggest that PACE could not have been intended to inhibit the ability of the police to deal with crime and disorder.
Though subject to some criticism on the grounds that they represent an over-simplification of what is, in fact, a complex structure,56 Packer's crime control and due process models of criminal justice remain useful evaluative concepts.57 Briefly, the emphasis of crime control is on achieving a high rate of apprehension and conviction of offenders. In order to achieve this, a premium is put on efficiency in terms of speed, which depends on informality. It is assumed that the police will exercise judgment to filter out innocent suspects at an early stage of the process, and those who are probably guilty pass quickly through the remaining *INTLJEVIDENCEPROOF 113 stages of the process, in this respect it is a managerial model. If there is confidence in the reliability of informal fact-finding in the early stages of the criminal process, the remaining stages can be performed relatively perfunctorily. The due process model is the procedural antithesis of crime control. Its premise is that no such confidence can be placed in informal non-adjudicative fact-finding. It is susceptible to error and demands the prevention and elimination of mistakes to the extent possible. The process exists as much to protect the factually innocent as to convict the factually guilty and efficiency in terms of throughput of cases must be tempered by measures that reduce the risk of error.
During the period to which their observations are related, Sanders and Young claimed that, although at every level of criminal justice system due process and crime control values seemed ‘intertwined often in contradictory ways’, it appeared that the conflicts were gradually being resolved by a movement in a crime control direction.58
The first version of the Code (1986) placed the police under a duty to hold a parade whenever the suspect made a request that one be held.59 The second revision (1995) imposed a duty to conduct procedures whenever the suspect disputed identification and consented to procedures being held, irrespective of whether or not the suspect requested that they be conducted.60 This version of the Code might be described as paternalistic and can be viewed as the high watermark of procedural due process.
The most recent revision of the Codes of Practice (2003) represent an extension of inquisitorial powers that have been conferred on the police in recent years with ‘the accused … increasingly seen as the subject of a police inquiry, … the product of which is available for use by the prosecution for adversarial purposes’.61 The changes effected in Code D bear many of the hallmarks of Packer's crime control model. Many of the amendments were previewed in temporary modifications to the Code issued in 2002.62 These were accompanied by a letter explaining that the changes had been made with the objective of speeding up the process:
It is becoming increasingly apparent that problems in arranging identification parades within a reasonable period of time have *INTLJEVIDENCEPROOF 114 contributed significantly to delays in processing cases and such delays can impact upon the quality of evidence and compromise the whole process of justice. Ministers are understandably concerned and want to take immediate measures to tackle these problems.63
With regard to the assertion that recent reform of Code D is firmly grounded in crime control values, this statement might be seen as ambivalent if viewed in isolation. Its references to reducing delay, or in other words increasing efficiency in terms of speed (crime control values), are offset by a reference to improving the ‘quality of evidence’, which presumably means ensuring that reliable identification evidence is obtained (a due process value, of course). However, under the current version of the Code, the police may decline to hold procedures where doing so would ‘serve no useful purpose in proving or disproving whether the suspect was involved in committing the offence’. This casts the police in the role of arbiter of whether the objective of adjudicative accuracy at a later stage in the proceedings would be furthered by holding an identification procedure. The drafters of the Code have turned away from the imposition of a procedural burden imposed for the purpose of reducing the risk of error, to a managerial process in which the police are given considerable discretion to dispense with a burdensome and time-consuming procedural hurdle. The merits and implications of this development cannot be considered at length here,64 but worthy of note is that this involves the police in a speculative, probabilistic exercise: whether the police are best equipped to undertake such an exercise is something we might reasonably question.
The changes implemented in the most recent revision of Code D (2003) might be a precursor to a period of rapid and significant growth in the influence of crime control values in shaping the contours of English criminal procedure. In 2002, the Home Office and Cabinet Office undertook a review of the Police and Criminal Evidence Act and the Codes of Practice. Its terms of reference were to identify changes in procedure that could (1) simplify police procedures, (2) reduce procedural or administrative burdens on the police; (3) save police resources, and (4) speed up the process of justice.65 In the Report of the Review, which was *INTLJEVIDENCEPROOF 115 the subject of vehement criticism,66 it was suggested that the Codes of Practice ‘could be given a structure based on outcome rather than process, leaving the police with more scope to use their professional judgment’.67 The congruence of this suggestion and the notion of ‘efficiency’ expounded in Packer's paradigm crime control model is both glaring and unsurprising, given that the Review's terms of reference amounted to an agenda for reform driven by the values underlying the model.
Procedural due process requirements provide a means of ensuring the protection of a suspect's fundamental interests. If one rejects as illusory the idea that it is possible to ‘balance’ the competing value systems68 in some sort of equilibrium, then one must also accept that there is a danger that any shift in emphasis towards the crime control values of procedural informality in the guise of reducing procedural and administrative ‘burdens’ can only be achieved by some diminution in protection against wrongful conviction (or least a significant risk that such protection will be eroded).
Policing adherence to the pre-trial regulatory regime
Atomistic evaluation of the adequacy of safeguards against the risk of mistaken identification is of limited value. It has been argued above that pre-trial procedures which ensure that identifications are procured under conditions that minimise the risk of error are the most effective safeguards that we have at our disposal.
If pre-trial procedure provides the most effective available means of guarding against the risk of mistaken identification, then once its prescriptions are developed to a satisfactory state, their importance ought to be reflected in the means of ensuring compliance with them. In an article that appeared in this journal, Ian McKenzie suggested that guidance recently issued to police officers in the United States69 would be of limited effect in the absence of sanctions for non-compliance. Code D was used as a comparator in respect of which it was claimed that a breach might be followed by a number of sanctions. It was asserted that any police officer responsible for a breach might be subjected to disciplinary proceedings, that any breach might be admissible in criminal or civil proceedings *INTLJEVIDENCEPROOF 116 and that a serious breach would ‘unfailingly’ lead to evidence being excluded. The position in English law might not be as straightforward as that which McKenzie appears to set out.
Although s. 67(8) of PACE originally provided that an officer would be liable to disciplinary proceedings for any breach of the Codes of Practice, this provision was repealed by s. 37 of the Police and Magistrates' Courts Act 1994.70 As for the threat that breaches of the Codes of Practice might be admissible in evidence in subsequent criminal or civil proceedings, s. 67(10) of PACE makes clear that a breach of the Codes will not of itself render the officer who commits the breach liable to criminal or civil proceedings. In the context of Code D, it is not at all obvious how evidence of a breach might be relevant to future criminal proceedings against an officer.71 Moreover, beyond the possibility that a breach of the Code's provisions concerning covert identification procedures giving rise to a breach of Article 8 of the European Convention on Human Rights (the right to respect for private and family life) might result in damages being awarded, it is difficult to envisage circumstances in which a breach of Code D might give rise to an actionable tort. This leaves the claim that ‘serious breaches will unfailingly’ lead to evidence being excluded.72
The provisions of Code D that set out how formal procedures conditions are to be carried out are admirably clear and detailed. Its provisions concerning when these procedures ought to be held are vague, and those concerning street identifications are inadequate. No one has yet undertaken a thorough review of appeal cases involving submissions that evidence of identification ought to have been excluded following some breach of the Code. However, a cursory review of recent cases suggests that the circumstances surrounding the conduct of formal procedures are few and that the greater proportion concern the reliability of street identifications73 and the failure of the police to conduct formal procedures.74
In respect of the exclusion of evidence of street identifications, the courts, recognising the difficulties faced by the police in these circumstances, appear *INTLJEVIDENCEPROOF 117 reluctant to find a breach of the Code's provisions. Even where such a breach is found identifications conducted under what appear to be the most suggestive of circumstances will not necessarily be excluded. It is not unusual for the appellate courts merely to recite factors in the circumstances surrounding an identification before declaring that the identification of the defendant was untainted or free of the risk of unreliability. Generally no comment is made on whether the factors stated have been considered relevant to the issue of reliability, whether or not they have been taken to be indicative of reliability or unreliability, and the weight that has been accorded to them.
For example, in R v Williams,75 a case recently heard by the Court of Appeal, a woman waiting at a bus stop became the victim of an attempted robbery during which the culprit struck up a conversation which became more menacing and culminated with a demand for cash and jewellery and a threat to stab her with a hypodermic needle. She managed to escape and call the police. She provided a description by telephone and then went to the house of a relative. That description was of a clean-shaven man with a London accent wearing black or dark blue trousers, a black jacket with black zip and brown shoes. The police making their way to the scene saw and gave chase to a man who was eventually detained on the roof of a factory. This man spoke with a distinctly different Glasgow accent, had two to three days stubble on his face, was wearing black trousers and a black jacket which had a brass zip, and black, not brown, shoes. The officers who detained the suspect were told of the culprit's threat to stab the victim with a hypodermic needle. As a result he was handcuffed. The victim was driven to a place where she saw the handcuffed suspect standing by a police car surrounded by police officers. She observed him from a distance of about 10 metres and stated that she was positive that he was the culprit. The court accepted that there had been a breach of the Code. It ‘mattered not’ whether this was because the police had a ‘known suspect’, in which case formal procedures ought to have been held, or that they had breached the provision stating that ‘care must be taken not to direct the witness's attention to any individual unless, taking into account all the circumstances, this cannot be avoided’.76 The court acknowledged the suggestiveness of the procedures adopted, stating that ‘a clearer way of pointing out to her who it was that the police sought her to identify could not have been imagined’.77 Notwithstanding this, it was held that the trial judge's decision not to exclude the evidence was correct and the conviction was safe.
*INTLJEVIDENCEPROOF 118 This case, in which the prosecution case rested on identification evidence of questionable reliability following breaches of the Code, is not atypical. It is suggested that a thorough review of the authorities would not reveal that what might be thought of as ‘serious’ breaches of the Code ‘unfailingly’ lead to the exclusion of evidence. The basis of exclusion is the effect that the evidence on which the prosecution seek to rely would have on the fairness of the proceedings. The problems of accurately reconstructing past events to establish whether witnesses recollections have been distorted by suggestive conduct have been alluded to above. However, Williams illustrates that even evidence obtained in breach of Code D, in circumstances that are clearly suggestive, which are attended by a high risk of error, will not necessarily be excluded. Although exclusion for any breach of the Code, however minor, is not advocated here,78 decisions not to exclude evidence in circumstances such as those in Williams indicate an insensitivity to importance of pre-trial procedures in reducing the risk of mistaken identification. Moreover, if repeated such decisions might have an insidious effect on police compliance with the provisions of the Code.
The Court of Appeal in R v Forbes79 stated explicitly that the s. 78 discretion should not be exercised as a means of disciplining the police for failing to hold a formal procedure.80 However, in absence of using the discretion for this purpose, there is no effective sanction attached to a breach. The significance of this, as Schauer explains, is that the strength of any regulatory provision lies in the conditions surrounding its applicability, acceptance and performance.81 In other words its strength comes from the sanctions that attach to any violation of it. If the police perceive that the courts usually allow the prosecution to adduce identification evidence obtained in breach of the Code the option of non-compliance becomes more available and attractive. Why should they spend the time and effort necessary to comply with the provisions if it is likely that no adverse consequences will flow from any breach? Salembier observes; ‘In such circumstances a participant's actions cease to be guided by the rules and instead become dependent on its knowledge of enforcement proclivities.’82 The result is an undermining of the regulatory regime.
*INTLJEVIDENCEPROOF 119 It is necessary to add one significant caveat. The greater any trend towards replacing the Code's bright-line or hard-edged rules with provisions that confer on the police wider scope to use their professional judgment, the less relevant any consideration of enforcing compliance with the Code will become. Enforcing compliance with the provisions of the Code is a worthwhile pursuit only where following them serves to improve the chances of securing reliable witness identifications. It was suggested earlier that this is objective is more likely to be met through issuing prescriptions to the police than the conferring discretion upon them.
It is important that we should avoid the seductive notion that because considerable efforts have gone into developing some aspects of pre-trial procedure that our response to the problem of mistaken identification is satisfactory and miscarriages of justice are now a remote possibility. The nature of memory illustrates the importance of the regulation of pre-trial procedure, particularly at an early stage of the investigative process. It has been suggested above that identification procedures conducted under controlled procedures represent the most effective means of safeguarding against the risk of miscarriages of justice resulting from mistaken identifications. However, their effectiveness can be undermined by inadequate regulation of events prior to the formal procedure. While recent reform of English procedure has provided greater regulation of what might occur prior to formal identification procedures, the source of that regulation, Code D, implies that what is provided remains inadequate. The Code provides for admirable formal identification procedures that appear sensitive to the empirical findings of psychologists and where employed are likely to reduce significantly the risk of mistaken identification. However, changes to the provisions regulating when such procedures must be held simply gives rise to an increased risk at another point of the process. The importance of pre-trial procedure as a safeguard requires that departures from it be taken seriously. Unfortunately, the position in England and Wales is that the response of the Court of Appeal to such occurrences is at best ambivalent. Viewed from a broad perspective, the spectre of miscarriage of justice resulting from mistaken identification remains.
I am indebted to David Ormerod and Diane Birch for their insightful comments on earlier drafts of this work. I remain solely responsible for errors and deficiencies.
E. & P. 2004, 8(2), 100-119
W. Twining, Rethinking Evidence: Exploratory Essays (Northwestern University Press: Evanston, IL, 1994) (first published in 1990 by Basil Blackwell Ltd).
Ibid. at 153.
Ibid. at 156.
D. Carson and R. Bull, ‘Psychology and Law: Future Directions’ in R. Bull and D. Carson (eds), Handbook of Psychology in Legal Contexts, 1st edn (Wiley: Chichester, 1995) 646.
See generally, P. Ainsworth, Psychology, Law and Eyewitness Testimony (Wiley: Chichester, 1998) 35-48; A. Kapardis, Psychology and Law, 2nd edn (Cambridge University Press: Cambridge, 2003) 36-47; B. Cutler and S. Penrod, Mistaken Identification: The Eyewitness, Psychology, and the Law (Cambridge University Press: Cambridge, 1995) 79-136.
It appears that identification accuracy is poorest among the relatively young and older witnesses. The findings of a number of studies indicate a significant decline in the performance of those aged 60+, see e.g. T. O' Rourke, S. Penrod, B. Cutler and T. Stuve, ‘The External Validity of Eyewitness Identification Research: Generalising Across Subject Populations’ (1989) 13 Law and Human Behavior 385.
Studies indicate greater accuracy where witness and subject are of the same ethnic appearance; see R. Bothwell, J. Brigham and R. Malpass, ‘Cross-Racial Identification’ (1989) 15 Personality and Social Psychology Bulletin 19; G. Wells and E. Olsen, ‘The Other-Race Effect in Eyewitness Identification: What Can We Do About It?’ (2001) 7 Psychology, Public Policy and Law 230. For analysis of US procedural safeguards in this respect, see R. Natarajan, ‘Racialized Memory and Reliability: Due Process Applied to Cross-Racial Eyewitness Identifications’ (2002) 78 New York University Law Review 1821; and in England and Wales, see T. Valentine, N. Harris, A. Colom Piera and S. Darling (2003), ‘Are Police Video Identifications Fair to African-Caribbean Suspects?’ (2003) 17 Applied Cognitive Psychology 459.
Although there have been conflicting findings there is evidence that high levels of stress experienced during observation of an event can have an adverse effect on memory, see generally Kapardis, above n. 5 at 40-4.
Kapardis, above n. 5 at 38, has cautioned against the common-sense assumption that the identification accuracy is likely to increase in line with duration of observation on the grounds of selective attention. It appears that witnesses who experience stress during observation tend to over-estimate the duration of the event, see I. Sarason and R. Stroops, ‘Test Anxiety and the Passage of Time’ (1978) 46 Journal of Consulting and Clinical Psychology 189; B. Cutler, S. Penrod and T. Martens, ‘The Reliability of Eyewitness Identifications: The Role of System and Estimator Variables’ (1987), 11 Law and Human Behavior 223.
See Cutler, Penrod and Martens, above n. 9; E. Loftus, G. Loftus and J. Messo, ‘Some Facts About “Weapon Focus”’ (1987) 11 Law and Human Behavior 55; N. Steblay, ‘A Meta-Analytic Review of the Weapon Focus Effect’ (1992) 16 Law and Human Behavior 413.
G. Wells, ‘Applied Eyewitness Research: System Variables and Estimator Variables’ (1978) 36 Journal of Personality and Social Psychology 1546; G. Wells, Eyewitness Identification: A System Handbook (Carswell: Toronto, 1988) 13-23
Wells (1988) above n. 11 at 13.
R. Atkinson and R. Shiffrin, ‘Human Memory: A Proposed System and its Control Processes’ in K. Spence and J. Spence, The Psychology of Learning and Motivation: Advances in Research and Theory, vol. 2 (Academic Press: 1968).
R. Christiansen, J. Sweeney and K. Ochalek, ‘Influencing Eyewitness Descriptions’ (1983) 7 Law and Human Behavior 59.
See G. Kohnken, ‘Interviewing Adults’ in R. Bull and D. Carson (eds), Handbook of Psychology in Legal Contexts, 1st edn (Wiley: Chichester, 1995) 218.
See, e.g., G. Mazzoni, M. Vannucci and E. Loftus, ‘Misremembering Story Material’ (1999) 4 Legal and Criminological Psychology 93; see generally G. Davies, ‘Contamination of Witness Memory’ in A. Heaton-Armstrong, E. Shepherd and D. Wolchover (eds), Analysing Witness Testimony (Blackstone: London, 1999).
H. Crombag, W. Wagenaar and P. van Koppen, ‘Crashing Memories and the Problem of “Source Monitoring”’ (1996) 10 Applied Cognitive Psychology 95.
See E. Loftus and G. Loftus, ‘On the Permanence of Stored Information in the Human Brain’ (1980) 35 American Psychologist 409. See also D. Hall, E. Loftus and J. Tousignant, ‘Post-event Information and Changes in Recollection for a Natural Event’ in E. Loftus (ed.), Eyewitness Testimony (Cambridge University Press: Cambridge, 1984).
See J. Searcy, J. Bartlett and A. Memon, ‘Influence of Post-Event Narratives, Line-up Conditions and Individual Differences on False Identification by Young and Older Eyewitness’ (2000) 5 Legal and Criminological Psychology 219.