The Lammy Review: will it change outcomes in the criminal justice system?

October 11, 2017 — Comment

Written by Lee Bridges

Lee Bridges, Emeritus Professor, School of Law, University of Warwick, continues our examination of the Lammy Review.

4daysinausgustHaving already written on an allied subject – the 2011 riots-[1] the Conservative government chose Labour MP David Lammy to head its inquiry into the treatment of black and ethnic minority (BAME) groups within the criminal justice system. But the government made sure the terms of reference specifically excluded the police from its remit, given the already well-known high levels of ethnic disproportionality arising from police stop and searches and arrests, which in part may explain why the Lammy Review[2] is not as hard hitting as so many had hoped. It is based to a large extent on a specially-commissioned statistical analysis[3] which sets out to measure rates of disproportional outcomes for different ethnic groups at a number of stages of the criminal justice process, from Crown Prosecution Service (CPS) decisions to bring charges, through various court outcomes, and treatment in prisons. But at each stage different bases of comparison are adopted in an attempt to isolate the marginal effect of the particular stage on overall disproportionality within the system. So, the results of CPS charging decisions are measured against the ethnic composition of those cases which have been referred by the police to the CPS for such decisions, while court sentencing decisions are measured against the ethnic makeup of those convicted at each court level (magistrates’ court and crown court). The problem with this approach is that it produces snapshots of disproportionality at selected stages in the process and an episodic and somewhat disjointed analysis of the criminal justice process, rather than looking at the overall system effect in producing differential outcomes for the various ethnic groups. It also carries an implication that those making decisions at later stages in the process have no role or responsibility for seeking to redress unfair treatment of particular groups at earlier stages. 

Excluding policing from remit

david-lammyIn this context, the decision to exclude the police from the inquiry is crucial, as by far the highest rates of ethnic disproportionality are to be found in police activities such as stop and search and arrests. The statistical analysis carried out for the Lammy review produces figures showing that overall for every 100 young white males arrested in 2014 there were 135 arrests of BAME young males[4] (disproportionality rate of 1.35), while the differential rate of arrest as between white and BAME adult males was 1 to 1.75.[5] But these overall figures mask even higher rates of disproportional arrests for particular groups and crimes. Thus, for every young white male arrested, there were 277 arrests of young black men, while for adult men the ratio of white to black arrests was 3.28.[6] Young black males were over ten times more likely to be arrested for robbery, over five times more likely to be arrested for drugs offences, and nearly five times more likely to be arrested for fraud.[7] Young black females were also arrested for robbery at five times the rate for young white females.[8] Black adult men were more than eight times more likely to be arrested for robbery, over seven times more likely to be arrested for fraud, over five times more likely to be arrested for drugs offences, and nearly four times more likely to be arrested for sexual offences, than white adult men.[9]

It is unfortunate that the Lammy review statistical analysis did not take the opportunity to compare these ethnic disproportionality rates for arrests to those for cases referred by the police to the CPS for decisions on possible charges. This might have shed some light on how the police operate their discretion either to take no further action against arrestees or otherwise to divert them from the criminal justice system through administrative penalties, such as cautions, warnings or on-the-spot fines, and how such decisions contribute to overall ethnic disproportionality. This has important implications for one of Lammy’s key recommendations, that an experimental ‘deferred prosecution’ model whereby arrestees can be offered non-criminal options without first having to admit their guilt, should be rolled out on a national basis. If the police are shown to operate the diversionary remedies currently available to them in an ethnically disproportionate way, it is unlikely that implementation of this recommendation would make a significant contribution to reducing overall disproportionality within the system.

Gang offendinglammy-review


The figures cited above for ethnic disproportionality in arrests serve to confirm just how strongly the stereotype of street robbery or ‘mugging’ being a particular black crime, and the discriminatory policing policies and priorities developed to fit with this narrative, have persisted over the past 60 years. However, Lammy makes no reference to this historical context or the wider pattern of policing black communities, choosing to relate disproportionality in stop and search and arrests, especially for drugs, to ‘suspicions of gang offending ‘.[10] This leads him to consider two important aspects of current policing and prosecutorial policies – the use of ‘gang databases’ and the law of joint enterprise in policing and prosecuting alleged gang crimes. He notes that of the 3,621 names on the Metropolitan Police’s Trident Matrix database[11] no less than 86 per cent are BAME, and goes on to caution that:

Care must be taken to ensure that information on such databases is accurate, up to date and used in the right way. It is not clear, for example, why the charge sheets passed by the police to the CPS detail whether or not an individual can be found on the Trident Matrix. The line between intelligence about people’s associations and evidence about their actions needs to be carefully guarded.[12]

Yet, Lammy’s only recommendation in respect of ‘gang databases’, which are more widely used than just in London, is that the current Mayor of London’s review of the Trident Matrix ‘should examine the way information is gathered, verified, stored and shared, with specific reference to disproportionality.’[13] Nor does he go on to consider how information of supposed ‘gang associations’ is filtered through to later stages of the criminal justice process and its impact on disproportionality in decisions on which courts defendants are to be tried in, whether they are remanded in custody to await trial and eventually convicted, and the sentences they receive. This contrasts with another of his recommendations, that ‘all identifying information should be redacted from case information passed on to them, allowing the CPS to make race-blind decisions.’[14] In this respect, Lammy appears to be guilty of the very failure to turn ‘insight into action’, and of passing the buck, of which he accuses criminal justice institutions[15] in their attitudes to evidence of ethnic disproportionality. At the very least he should have recommended a suspension in the use of ‘gang databases’ and banning any passing of such prejudicial information to the CPS and, through them, to the courts. 

Joint enterprise

Lammy’s consideration of joint enterprise is equally artificial. He notes that a ‘landmark’ judgement in 2016, which he wrongly attributes to the High Court rather than the Supreme Court, found that the law on joint enterprise had been misinterpreted for three decades and that as a result a higher threshold of proof of involvement in group crimes was now required. He goes on to acknowledge continuing concerns that ‘the line between “prohibitive” and “prejudicial” information’ is still not being drawn appropriately in evidence put before juries in such cases. However, he limits his recommendation to urging the CPS, in reworking its guidance on joint enterprise in light of the Supreme Court judgement, ‘to consider its approach to gang prosecutions in general.’[16] This again demonstrates a lack of historical understanding of the role of the CPS, and the current Director of Public Prosecutions in particular, working in collaboration with the police, in resurrecting the law of joint enterprise, pushing for its wider interpretation by the courts, and subsequently strenuously defending its use as a dragnet tool to prosecute ever wider numbers of individuals for their alleged involvement in gang or group offences.[17] Against this background and left to the CPS alone, there is little hope of achieving the wholesale reform of the law of joint enterprise that is so badly required, let along giving redress to the hundreds of BAME individuals who have been unjustly convicted under its aegis. 

Role of the Crown Prosecution Service

More generally, Lammy has concluded that CPS decision making does not contribute significantly to ethnic disproportionality in the criminal justice system. This is on the basis that both the statistical review he commissioned and the CPS’s own research found that on the whole they tended to charge equal proportions of cases referred to them across the different ethnic groups, although no analysis was carried out to indicate whether there was any evidence that CPS decisions had any effect in counter-balancing the ethnic disproportionality in police arrests and decisions to refer cases for charge. More importantly, no attempt was made by the Lammy review to investigate qualitiative differences in charging decisions across the various ethnic groups, which previous research has suggested may have an important impact on the ways in which cases are processed subsequently. Thus, if BAME cases tend to result in more serious charges than those of white defendants based on similar types of alleged criminal behaviour (e.g. charges of robbery rather than theft, or of drug possession with intent to supply instead of mere possession), this can influence at which of magistrates’ court or the crown court (the latter having more punitive sentencing powers) cases will be tried, as well as the type and length of sentence imposed if there is a conviction. Seriousness of charge can also influence decisions on whether defendants are remanded in custody prior to charge, a matter on which the CPS will have a direct input through representations made by the prosecution in court. It may also play a key role in the defendant’s decision on plea, as will be considered below. 

If Lammy is too quick to let the CPS off the hook, he approaches the courts and their processes with what can only be described as a rose-tinted and nostalgic view of their role:

The most important decisions in the justice system are made in our courts. They are where life-changing judgements are made about innocence or guilt.[18] 

Or again,

Our justice system is built on the principle that the law will be applied impartially. In the cases that involve the greatest harm to victims and the longest sentences for offenders, juries are the guardians of this principle.[19]

Pleas and outcomes

Yet, even before considering the influences of the courts, Lammy first focuses attention on the issue of pleas by defendants, which he sees as a major determinant of ethnic outcomes in the criminal justice system due to two factors. One is the system of sentence discounts for guilty pleas, which although it has long operated on a discretionary and non-statutory basis within the criminal justice system, was made more formal and institutionalised under both Conservative and Labour governments from the mid-1990s onwards.[20] Under this system a defendant who pleads guilty at the first opportunity to do so within the court process can have his or her sentence reduced by as much as a third, and one who pleads guilty at a very late stage (‘at the door of the court’) can still receive a discount of around 10 per cent on the sentence they may have otherwise been given. This formalised system of sentence discounts was introduced despite the well-known propensity of defendants from some ethnic minority groups more often to enter not guilty pleas at least initially, and therefore in the full knowledge that the system would significantly disadvantage such groups. Lammy describes the ethnic differences in pleas decisions as ‘stark’, although the evidence from the review’s own statistical study showed that rates of disproportionality for pleading not guilty (based solely on cases in the crown court) ranged, for young BAME males of different ethnicities, from 1.28 (for black and Asian defendants) to 1.46 (for mixed ethnic defendants) and, for adult BAME males, from 1.26 (for mixed ethnic defendants) to 1.61 (for Chinese defendants). Compared to rates of disproportionality in arrests for some offences (see above), these rates are much less significant. Lammy also cites recent research showing that between 2006 and 2014 BAME defendants pleaded not guilty in 40 per cent of cases as compared with 31 per cent of cases involving white defendants.[21] In other words, a majority of BAME defendants do in fact plead guilty. 

Lammy also accepts without comment or criticism the official rationale for sentence discounts, even with their effect in perpetuating ethnic disadvantage and that they operate to undermine his own view of the role of the courts and of juries in the justice system. A guilty plea is in itself legally determinant of a person’s guilt and therefore replaces any need that a court or jury will have to decide on guilt or innocence. Nor will the state, through the prosecution, be required to prove the case against the defendant (the burden of proof). Put another way, the sentence discount operates to penalise a defendant who exercises his or her right to a presumption of innocence and to require the state to prove guilt beyond reasonable doubt. 

Lammy attributes the higher propensity of BAME defendants to plead not guilty solely to a ‘[l]ack of trust in the justice system’ among BAME defendants, which at one point he seems to attribute to generational differences:

the criminal justice system … has a trust deficit with the BAME population born in England and Wales, many of whom lack their parents’ reverence for our legal system.[22]

While one may question how much ‘reverence’ what may be called the Mangrove generation had for the British legal system, Lammy clearly sees a particular issue regarding plea decisions among young BAME defendants, and particularly those from the black community, who he says might respond to their arrest “with a ‘no-comment’ interview in the police station before entering a not guilty plea.”[23] Lammy points out that such defendants may go on to change their plea at a later stage or to elect to be tried in the crown court where higher sentencing powers are available,[24] both of which courses he clearly regards as irrational as they risk potential loss of a sentence discount. He also argues on the basis of anecdotal evidence that BAME suspects, who tend more often to seek legal advice in police stations than white suspects, ‘neither trust the advice that they are given, nor believe that they will receive a fair hearing from magistrates.’[25] One of his recommendations in this regard is that 

The Home Office, the MoJ and the Legal Aid Agency should work with the Law Society and Bar Council to experiment with different approaches to explaining legal rights and options to defendants. These different approaches could include, for example, a role for community intermediaries when suspects are first received in custody, giving people a choice between different duty solicitors, and earlier access to advice from barristers.[26]

In making this recommendation Lammy appears to be unaware that it has been the policy of the legal aid authorities over a number of years to reduce the number of solicitors’ firms undertaking criminal defence work, a policy that has a particularly adverse effect of smaller defence firms, often those with BAME lawyers. 

On the whole, Lammy’s comments on plea bargaining display a conventional naivety both about the workings of the criminal justices system, where he seems to see issues of guilt or innocence as black or white rather than something that is contingent on questions of evidence and interpretation, and about the role criminal defence lawyers in that system. In practice, suspects under questioning by the police will often be rightly counselled to exercise their right to silence or to plead not guilty at the initial stages of court proceedings, precisely because there is lack of sufficient information from the police or prosecution about the case against them. Ethical codes for defence lawyers all make it clear that while they must advise a defendant regarding the availability of sentence discounts, they should not pressurise or advise them to plead guilty unless satisfied that the prosecution can establish its case.[27] Defendants and their legal advisers also need to consider the possibility that a reduction in, or even a withdrawal of charges will be offered at a later stage once the evidence on both the prosecution and defence sides has come to light and been examined, whether in court or outside. Given the highly disproportionate rates at which BAME defendants are arrested and possibly charged for very serious offences, this is a particularly relevant consideration for them. Lammy produces no evidence, for example, on how often defendants, whether BAME or white, who change their plea to guilty at later stages in the proceedings against them do so following such a reduction in charge. Of course, where a defendant fails to plead guilty at an initial stage due to lack of disclosure by the police or prosecution but does so subsequently, following receiving further information on the case against him and/or the offer of a reduction in charge, he or she should still be eligible for a full sentence discount, but it is unclear whether or not in practice this is what happens. This is an area where Lammy might well have recommended that further research is carried out, to examine the extent to which the system of sentence discounts actually works as it should in the case of BAME defendants and how it contributes to overall disproportionality in the sentences such defendant receive (see below).

Crown court v magistrates

Equally, as noted earlier, seriousness of charge is often a key factor in determining which level of court a case will be tried in. For example, robbery is one of a number of ‘indictable only’ charges which must be tried in the crown court, and most decisions to send other ‘triable either way’ cases to the crown court are the result of magistrates refusing jurisdiction because they do not consider their sentencing powers to be adequate, rather than because the defendant has elected to be tried there. Interestingly, in many such cases where there is a conviction at the crown court, it imposes a sentence which would have been available to the magistrates, which would seem to go against the conventional wisdom that the crown court is always more punitive. 

The statistical analysis conducted for the Lammy review also suggests that BAME defendants on the whole are not disadvantaged in comparison with their white counterparts as a result of their cases more often being conducted at crown court and their propensity more frequently to plead not guilty. This pattern of behaviour would suggest that they might be more likely to be found guilty at trial, whereas in fact the evidence is that jury trials result in virtually equal proportions of convictions and acquittals for both BAME and white defendants. As a result, once both guilty pleas and jury verdicts are taken into account, marginally smaller (but in most cases still statistically significant) proportions of crown court cases involving adult BAME defendants across the different ethnic sub-groups end up in convictions than is the case for white defendants. Overall, there are approximately nine convictions of BAME defendants at crown court for every ten white convictions. This again suggest, contrary to the presumption put forward in the Lammy Review, that there is a rational basis for the decisions made by BAME defendants in respect of their pleas. 


Finally, the Lammy Review draws attention to a large 2016 research study examining ethnic differences in sentencing at the crown court in three general categories of offences: those involving acquisitive violence, sexual offences, and drugs offences. Lammy summarises the findings as follows:

Of approximately 21,370 cases studied, there was no statistical link between ethnicity and the likelihood of receiving a prison sentences for the offence groups of acquisitive violence and sexual offences, but there was a strong effect within drug offences. Within drug offences, the odds of receiving a prison sentence were around 240% higher for BAME offenders, compared to White offenders. The study could not account for the impact of aggravating and mitigating factors, or for the possibility that BAME offenders may have been convicted of more serious drugs offences than their White counterparts, but it was able to take account of sex, ethnicity, age, previous criminal history and the plea decisions. The finding that, within drug offences, the odds of receiving a prison sentence were around 240 % higher for BAME offenders is deeply worrying. Many will conclude that this is evidence of bias.[28]

These finding are reflected in the statistical analysis carried out for the Lammy review, which showed that all groups of BAME adult males convicted of drugs offences at the crown court received custodial sentences proportionally more often than white men (with rates of disproportionality ranging from 1.14 for mixed ethnic males to 1.58 for Chinese males). This was also the case for black adult females convicted of drugs offences, who received custodial sentences at a rate 2.27 times greater than white females. As Lammy indicates, one factor contributing to these disproportional sentences could be the differences in the seriousness of the drug offences among the various groups, which may again relate back to initial decisions on charge. However, it is interesting that he did not make a connection between this pattern of sentencing and his earlier observations about the alleged connection between gangs and drug offences and the potential use in sentencing of information about alleged gang associations. Another factor noted by Lammy is decline in the use of pre-sentence reports (PSRs) due to cutbacks in the Probation Service, which has seen a reduction from over 50,000 to just over 13,000 in the number of full PSRs produced for sentencing in crown court each year and their replacement with ‘fast delivery’ PSRs produced on the same day as the sentence. Moreover, as Lammy points out, 

judges have received guidance discouraging them from using PSRs altogether for some offences. These offences include drug offences such as ‘possession with intent to supply class A drug’ – precisely the type of offence where the evidence suggests there are sentencing disparities.[29]

But yet again, Lammy is far too timid in his response to this situation. Rather than insisting on the restoration of full PSRs, he merely suggests that the Ministry of Justice should review the use and effectiveness of PSRs, and even then does not elevate this to one of his formal recommendations. 

Lack of representation in CJS

Beyond this, the Lammy Review provides a good deal of evidence regarding the lack of BAME representation in various criminal justice institutions, with perhaps the one notable exception of the CPS. The data he produces on the unequal treatment of BAME applicants in the judicial appointments process is particularly alarming. Many of his recommendations deal with steps that seek to remedy such under-representation (of which those relating to judicial appointments are likely to be the most strongly resisted by the legal establishment) and to improve systems for further monitoring ethnic disproportionality in other aspects of the criminal justice system. But continuous measuring of the extent of such disproportionality will not in itself bring about change, especially where the onus for such change is placed on the institutions themselves. This is essentially what is suggested in Lammy’s recommendation that:

If CJS agencies cannot provide an evidence-based explanation for apparent disparities between ethnic groups then reforms should be introduced to address those disparities. This principle of ‘explain or reform’ should apply to every CJS institution.[30]

Nor will changing the complexion of those working in the system necessarily alter its class base or bring about the radical changes in its structures, policies, processes and priorities that are required to tackle the ethnic disadvantages that it continues to perpetuate.

Related links

Read another IRR News article on the Lammy Review: Without racial justice can there be trust?

[1] David Lammy, Out of the Ashes: Britain after the riots (London, Guardian Books, 2011). For a critique, see Lee Bridges, ‘Four days in August: the UK riots’, Race & Class, 54:1 (July-September 2012). [2] The Lammy Review: An independent review into the treatment of, and outcomes for, Black, Asian and Minority Ethnic Individuals in the Criminal Justice System, September 2017. [3] Noah Uhrig, Black, Asian and Minority Ethnic disproportionality in the Criminal Justice System in England and Wales, Ministry of Justice Analytical Services, 2016. [4] Ibid., Figure 5.1, p. 9. [5] Ibid.. Figure 5.3, p. 10. [6] Ibid., Table 5.1, p. 13. [7] Ibid., Table A2.1, p. 37. The Lammy Review misquotes the drugs arrest rate figure for young black males, stating that ‘Black boys [were] ten times as likely as white boys to be arrested from drugs.’ (Lammy Review, op. cit., p. 19. [8] Uhrig, op. cit., Table A2.4, p. 41. [9] Ibid., Table A2.9, p. 45. [10] Lammy Review, op. cit., [11] Stuart Hall,, Policing the Crisis: Mugging, the state, and law and order, London, Palgrave, 1978. [12] Lammy Review, op. cit.. p. 19. [13] Ibid. [14] Ibid., p. 21. [15] Ibid., p. 14. [16] Ibid., pp. 19-20. [17] See Jason Beatie and Ben Rossigton, ‘The country’s top prosecutor denies “joint enterprise” has led to miscarriages of justice’, (Daily Mirror, 22 October 2014). [18] The Lammy Review, op.cit., p. 31. [19] Ibid. [20] For a discussion of these developments in sentence discounting, and of the ethical issues they pose for criminal defence lawyers, see Lee Bridges, ‘The Ethics of Representation on Guilty Pleas’, Legal Ethics, 9:1 (2006). [21] The Lammy Review, op.cit p. 27. Lammy fails to mention that these figures appear to relate only to crown court cases, as the study was mainly concerned with fairness in jury trials, and data on pleas is not available for cases in magistrates’ courts. [22] Op. cit.,p. 36. [23] Op. cit., p. 27. [24] Ibid. Lammy cites research, again apparently relating only to the crown court, that BAME defendants changed their plea to guilty in 21% of cases as compared to 17% for white defendants. [25] Ibid. [26] Ibid., p. 28. [27] See Lee Bridges, Legal Ethics, op. cit. [28] Lammy Review, op. cit., p. 33. [29] Ibid., p. 35. [30] Ibid., p. 14.

The Institute of Race Relations is precluded from expressing a corporate view: any opinions expressed are therefore those of the authors.


No comments yet.

Write a comment