Select Committee on Home Affairs Written Evidence


2.  Memorandum submitted by Dr Marian FitzGerald, Specialist Adviser to the Committee

EXECUTIVE SUMMARY

Background

  This paper is based on statistics provided specially for the Committee by four police services:

    —    The Metropolitan Police Service (MPS);

    —    Greater Manchester;

    —    The West Midlands; and

    —    Nottinghamshire.

  The forces vary in the size and internal make-up of their "Black" populations. The largest Black group in London is of African origin. Although Black Caribbeans are the largest single Black group in the other areas, only in the West Midlands do they account for over half of the Black population (at 56%). Among young people, the proportion is much smaller and the growth of the African and "Mixed" groups is becoming increasingly important for the future.

  The police figures are patchy but they were able to provide them at the level of Basic Command Units (BCUs) and for the financial year 2005-06, whereas the most recently published Home Office figures are at force level only and date back to 2004-05. The figures were supplemented by Youth Justice Board statistics for the force areas.

Victimisation

General

  The Home Office does not require forces to keep ethnic information on victims of crime reports; but figures kept by the MPS suggest the following:

    —    Black people are slightly overrepresented as victims compared to their presence in local populations; but

    —    the lower figure for Black males compared to other groups confirms that they may be less likely to report crimes against them to the police.

    —    Black overrepresentation as victims is higher in the younger age ranges; and

    —    they are more likely to be victims of violence (including serious violence and rape).

Homicides

  Despite the requirement to keep ethnic information on homicide victims, a surprising number of cases show ethnicity as "unknown" in some forces. The London data are more complete, though; and the total figures for six years are large enough to show that:

    —    by far the largest numbers of homicides in the Black group were of males aged 21 to 30; but

    —    by far the highest level of "disproportionality" was at younger ages where Black males accounted for nearly two thirds of all murders of 10 to 17 year olds.

  More detailed information for 2005-06 suggests that, overall, murders with sharp instruments in that year were more than three times as common as those using firearms. In the case of Black men, though, the figure reduces to two-to-one; and in the case of Black male victims aged 15 to 30, 10 were murdered with guns compared to 13 stabbings.

Entry into the Criminal Justice System

Searches

  In London, police searches under s1 of PACE fall disproportionately on Black people in all 32 boroughs; but in most this closely reflects the suspect descriptions recorded by the police. Compared to other groups, s1 searches fall particularly heavily on Black men and a higher proportion of these are in the younger age ranges.

  The arrest rate from s1 searches of Black people is similar to or higher than that for Whites in all areas; and this too could be cited by the police as evidence that the searches are justified. However, most s1 searches do not result in an arrest; so even if there are legitimate reasons for the level of disproportionality, this would still mean that Black young men going about their lawful business are disproportionately searched. And, as the Committee repeatedly heard, this is a major source of deep-seated resentment.

  The problems are far more acute in the case of searches under s60 of the 1994 Criminal Justice and Public Order Act. These do not require officers to have "reasonable grounds" for any search, so the arrest rate is much lower; and their impact on Black people is more disproportionate in. Generally, s60 searches are much less common than s1 searches; but, in the West Midlands, they account for about half of all searches on Black people—and in the BCU with the largest Black population they are twice as common as s1 searches.

Arrests

  Most arrests are not the result of police searches. So it is mistaken to claim that searches significantly increase the numbers of Black people in the criminal justice system. Black people are overrepresented in the arrest figures too; but the level of disproportionality varies by area and actually appears to be lower in London than in the three other forces.

  Where further analyses were possible these show important differences in the offences for which different groups are arrested as well as differences by age, although there are also local variations within this picture. Black young people are generally more likely than Whites to be arrested for robbery and drugs offences, but a higher proportion of young White people in the same areas enter the criminal justice system for burglary and criminal damage. However, robbery tends to be more common among young Black people aged 10 to 17 whereas by age 18 to 20 a higher proportion are arrested for drugs offences. Not only are drug offences a more important route into the system for Black young people than for Whites, the offences they are accused of tend to be more serious.

  In Greater Manchester and Nottingham, figures were available for young people of `Mixed' origins and these further highlighted the growing importance of this group for the future. Young people of Black Caribbean/White heritage in Nottingham actually outnumber those of Caribbean origin as such among arrestees aged 10-17.

Disposals

  Since the CPS and the courts are not required to collect ethnic statistics, there is little published information on the proportion of arrests of different ethnic groups which result in a disposal of some sort. However, both Greater Manchester Police and the MPS were able to provide some information on this. They differ in the overall proportion of all arrests which do not result in a formal disposal; but, in both areas, arrests of young Black people are slightly more likely to result in no further action compared to Whites. At the same time, Black young people are slightly less likely to receive a caution or warning.

  In London, the proportion of young Black people charged with an offence is higher than for Whites in both the 10 to 17 and the 18 to 20 age groups; but the figures are also large enough to break this down by offence type. In the case of robbery, the vast majority of offences result in a charge and there were no ethnic differences within this; but the overrepresentation of Black young people in robbery will itself increase the overall proportion of Black young people charged with an offence. Cases of violence against the person, though, are more likely to result in young Black people being charged compared to Whites but the opposite is true in cases of theft and handling. Reliably to interpret these contrasting findings would require information on the relative seriousness of the offences involved and on other factors such as admission and criminal histories.

Additional insights from Youth Justice Board statistics

General

  Significant numbers of young people who receive a reprimand or formal warning from the police appear not to be known to local Youth Offending Teams, although the scale of the problem varies by area. Black young people are disproportionately represented among the 10-17 year olds who are known to the youth justice system but the overall level of "disproportionality" is lower than in the police statistics. This may be because the published police figures refer only to "notifiable" offences, but many young people in the YOT statistics have committed summary offences and/or have breached a previous order (although the prevalence of some of these categories varies by area).

Patterns of offending

  Despite their differences from the police figures, the YOT data show similar ethnic differences in offending patterns, even within the same areas. Again, White young people are more likely to have received a disposal for criminal damage and burglary offences while Black young people are more likely to enter the system for robbery and drugs offences. However, the pattern for the `Mixed' group is different from either.

Remands and disposals

  Both Black people and those of Mixed heritage, though, are consistently less likely to receive pre-court disposals. They are more likely to be remanded in custody rather than bailed and they are also more likely to receive custodial sentences. The YOT data do not provide any information on the level of seriousness of their offences, whether the offences are admitted or other relevant factors like previous convictions. However, both Black and Mixed heritage young people receive longer sentences than Whites and a higher proportion of these are sentenced for offences which, in the case of adults, would be eligible for custodial sentences of 14 years or more

Issues arising from the analyses

  Police forces hold much richer and more up-to-date ethnic information than is apparent from the figures published by the Home Office. Importantly, these are susceptible of analysis at sub-force level, by age and in many cases also by gender. As such, they have considerable potential for informing local policy and this is essential for several reasons.

  There are local variations in the make up of minority populations, their particular social, economic and policy context and issues of crime and policing related to these. Interventions need to take account of these factors and to be delivered in ways which are locally appropriate. Their impact will then also need to be monitored at local level.

  That is, top-down solutions may not be appropriate and one-size-fits-all approaches may not work in many situations. In most areas the numbers of ethnic minorities in forces' data will be too small for meaningful analysis at BCU level; but an over-emphasis on statistics (driven by an imperative to make the figures look less "disproportionate" in the absence of any agreement on what fairness should look like) may actually inhibit forces from finding other ways of exploring the issues, discovering with precision where any problems are arising and taking any necessary action.

  In any case, no progress is likely without local "ownership" of the issues; but the relative absence of information in some forces even on items where they are formally required to keep ethnic statistics suggests that the s95 figures are currently seen as yet another set of statistics forces have to collect to provide returns to the Home Office. Even where the data provided to the Committee have provided valuable new insights, the forces which provided them may not have been aware of these or their possible policy implications.

  Similarly, the YJB data appear very comprehensive; but the published figures do not provide any breakdown within ethnic group by age or gender. Nor do they include information on other factors which would need to be taken into account before any inferences of discrimination could be drawn. However, it is also worth exploring why the youth justice figures show less disproportionality than the police statistics. One possible reason is that the police figures count incidents rather than individuals; and this may itself exaggerate the impression of disproportionality since the same individual may come repeatedly to police attention and any ethnic differences in the rate at which this occurs could significantly affect the total for each group.

  Collecting ethnic statistics cannot be justified if they are not analysed with rigour and the findings used to inform policy and practice; and this is especially true in the present context. For there is an increasing danger that the publication of figures which repeatedly and crudely highlight Black people's overrepresentation in crime compared to their presence in the population may eventually serve only as evidence to support racist stereotypes.

1.  BACKGROUND

Origins of this report

  The published statistics on ethnic minorities and the criminal justice system are patchy in their coverage. Also, those for individual police forces cover the whole of any force area, even though the distribution of minorities varies considerably at sub-force level, as do overall levels of crime, the deployment of police resources in relation to this and the use of police powers. In addition, for the purposes of the inquiry, the available data are often limited by the absence of any breakdown by age; and no information is available on the ratio of males to females within different ethnic groups at most key points in the criminal justice system.

  The Committee therefore approached four police forces to ask what additional information they could provide from the data they already routinely hold which might shed light on these issues. The forces were:

    —    the Metropolitan Police Service (MPS);

    —    Greater Manchester;

    —     the West Midlands; and

    —    Nottinghamshire.

  The first three of these are large urban forces, whereas Nottinghamshire is a much smaller force, much of which comprises rural areas and medium size towns. However, it has recently suffered levels of gun crime and drug-related activity which are more usually associated with the inner city areas of metropolitan forces; and here, as in the other three forces, young Black people have been victims of this violence. The forces vary in the size and the ethnic make-up of their resident Black populations (see further below); but between them, they account for 75% of the Black population served by the 43 police forces in England and Wales. That is, any supposedly national average figures for Black people in the published figures may largely be determined by what happens in these four forces.

  The information received from the forces varied in its coverage, in the extent to which this recorded ethnicity and also in the form in which the data were provided, which in turn had implications for the extent to which further analyses were possible. The forces were, though, most helpful in responding to our requests in a very short space of time and, taken together, the material provides important additional insights to the statistical information available to the Committee to date. It also indicates the potential of force level data both for exploring the factors behind the overrepresentation of Black people (and young people in particular) in the criminal justice system and for considering their implications for policy and practice.

  This report begins by setting out information on the Black population of each area, set in the context of the force's Basic Command Units (BCUs). It then looks in turn at:

    (a)  victimisation;

    (b)  entry into the criminal justice system; and

    (c)  subsequent decisions;

  before concluding with some brief, personal reflections on possible implications for the inquiry.

  .

  The police are no longer responsible for the decision to charge arrestees; and the CPS are unable to provide any ethnic data on this point. Nonetheless, the data from forces offered some useful insights under c) and this was supplemented by statistics on 10-17 year olds in the same local areas from Youth Offending Team (YOT) returns to the Youth Justice Board.

The local context in four forces

  Most police data are still presented using the traditional "4+1" codes adopted when mandatory ethnic monitoring was introduced in 1996 (ie White, Black, Asian, Other and Not Known/Recorded). This system was based on visual classification by police officers and was effectively a condensed version of the six point (PNC) classification traditionally used by the police to classify suspects by ethnic appearance. The currently available published data, for the most part, still use this four point classification. However, officers have formally been required since 2003 to ask any individual they stop, search and/or arrest to classify themselves according to the 16-point Census classification. For the purposes of this paper, it is assumed that the "Black" group in the police statistics comprises the following Census categories:

    —    Mixed White-Black Caribbean.

    —    Mixed White-Black African.

    —    Black Caribbean.

    —    Black African.

    —    Any other Black group.

  Table 1 shows the total size of this Black population in each of the force areas and its internal composition. Both vary considerably, with Black Africans the predominant group in London but people of `Mixed' White and Black Caribbean heritage accounting for more than a quarter of all Black people in Greater Manchester and the West Midlands, and fully a third in Nottinghamshire. It is only in the West Midlands that Black Caribbeans still account for more than half of the total Black population; but even here the figure is no more than 56 per cent.

Table 1

Size and composition of "Black" population in each force area[2]


Of which
% residents "Black"
Mixed White- Black Caribbean
Mixed White-Black African
Black Caribbean
Black African
Any other Black group

12.4
8.0
  3.8
38.7
42.7
6.8
Greater Manchester
  1.9
27.5
10.2
34.0
21.5
6.8
West Midlands
  2.8
27.0
  2.5
55.8
  8.1
6.6
Nottinghamshire
  2.3
33.3
  3.9
48.8
  8.1
5.9


  The MPS additionally provided an age breakdown for the different groups in the London population and this provides further evidence that, even in London, the population of "Mixed" origins will form an increasingly important component of the Black youth population for the future. Meanwhile, the Black Caribbean population of the capital will continue to shrink and already comprises less than a third of the total Black group in the 10-17 age range (see Figure 1).

Figure 1


  However, the distribution of the Black group within each force area is very skewed. As Figure 2 shows, it does not follow the distribution of the population as a whole between the component BCUS.[3]

Figure 2a


Figure 2b


Figure 2c


Figure 2d


  In each force, the areas in which Black people tend to be concentrated also tend to have higher than average levels of crime. However, the high crime areas in each force are not exclusively those with larger than average Black populations; for crime levels tend to be closely associated with two types of area characteristics. On the one hand, victim surveys confirm that people living in very deprived areas are more likely to be victims of crime. White people account for the majority of residents in many deprived, high crime areas within each of these forces, and in many forces, such areas will be almost exclusively White. On the other hand, some of the highest levels of recorded crime in any force area are in town and city centres or other neighbourhoods which have relatively small resident populations but which attract large numbers of non-residents for shopping, leisure, business or for other reasons.[4] Some of these reasons may be crime-related—for example in areas with significant drugs markets and red light districts.

  In this latter type of area many (if not most) of the victims and the perpetrators of any crimes reported will not be local: they may come from adjacent, more residential areas or from other parts of the same city; but some will not live within the force area at all. Hence, the borough/CDRP of Westminster dominates the picture for the MPS, accounting for over 7% of all London crime but under 3% of the capital's population. Similarly, the North Manchester BCU within the Greater Manchester CDRP accounts 12% of the force's total crime but for 5% of the force population.

  In all forces, the distribution of uniformed operational officers varies between BCUs with higher than average numbers assigned to the highest crime areas. Typically, town and city centres are the most intensively policed and have the most visible police presence; but this will also apply to some degree to the deprived high crime areas which is where Black people are more likely to live.[5]

2.  VICTIMISATION

General

  As anticipated, the data kept by all forces on victimization in general had large numbers of cases in which ethnicity was not recorded. This is in part because the police are not required to record victim ethnicity in cases other than homicide. In addition, though, some recorded crimes are either "victimless" (such as drugs offences) or they involve corporate victims rather than individuals. In view of both the large proportion of cases where ethnicity was not recorded and the relatively small size of their Black populations, no victimisation data are presented here for Greater Manchester, the West Midlands or Nottinghamshire.

  In the MPS, the level of non-recording for victims varied considerably by borough, averaging 16.5% overall but covering a range from 44% down to less than 2%. Table 2 shows Black people as a proportion of all victims of recorded crime, compared with their presence in the local population for boroughs where cases with no ethnicity recorded were 12% or lower. On average Black people were slightly more likely to be victims of crime than their presence in the local population would predict but there are a few cases where the figures are much higher. More work would be needed to discover whether, for example, the higher victimization figures in Barking, Greenwich and Wandsworth relate to the presence of particular groups of refugees and asylum seekers who might not have been included in the 2001 Census figures. It may, nonetheless, be significant that in the areas where Black people account for over 20% of the population there tends to be more parity between the population and victimization figures.

Table 2

Black people as victims of crime* selected London boroughs 2005-06


As % all victims of recorded crime
As % resident population

Richmond-Upon-Thames
  2.5
  1.6
Kingston-Upon-Thames
  4.6
  2.2
Sutton
  5.3
  3.4
Bromley
  6.7
  3.7
Hounslow
  8.1
  5.4
Bexley
  8.5
  3.4
Tower Hamlets
10.4
  7.7
Harrow
11.0
  7.1
Wandsworth
14.9
11.2
Barking & Dagenham
15.8
  8.2
Greenwich
18.4
12.5
Brent
22.8
21.6
Newham
23.5
23.5
Lambeth
27.7
28.6
Southwark
29.5
28.1

*  Figures based only on cases with ethnicity recorded


  Further analyses of the victim data for London as a whole suggest that, where ethnicity is known, there are further ethnic differences in the gender and average age of victims of crimes recorded by the police, as well as differences in the type of crime involved. In interpreting these figures it is important to bear in mind that the level of recorded crime is in some measure determined by the willingness of victims to report to the police in the first place. So the lower figure for Black males as victims of crime shown in Figure 3 could reflect under-reporting. For the idea that Black men have a lower underlying rate of victimization would be surprising in view of the homicide statistics for this group discussed further in the next section.

Figure 3


  Notwithstanding the possibility of under-reporting, in cases which do come to the attention of the police, Black people account for a higher than average proportion of victims in the younger age ranges (Figure 4).

Figure 4


  Black victims are also significantly more likely to be victims of crimes of violence (Figure 5); and within this very broad offence category, they account for a higher than average proportion of victims of both murder (see further below) and grievous bodily harm (GBH). Similarly, while only a small proportion of any group is recorded as the victim of a sexual offence and the overall figure is no higher for Black people than for any other group, Black victims accounted for a quarter of all the recorded reports of rape in London where ethnicity was known. However, contrary to assertions made by several witnesses to the committee, robbery does not account for any larger a proportion of Black victims than it does in other groups.

Figure 5


Homicides

  Homicides are rare events; so the total numbers will be small, particularly for minority groups within the population and it was for this reason that forces were asked for their homicide data for six full years, from 1999-2000 to 2005-06. Nonetheless, Nottinghamshire had only recorded 69 homicides over that period, of which 15 had no recorded ethnicity and only two were shown as Black.[6] In the West Midlands, the numbers were much larger (390); but ethnic information was missing in nearly two thirds of these cases.

  Missing data was less of a problem in Greater Manchester; but at 20%, this still makes inter-ethnic comparisons uncertain when there were only 24 recorded Black victims out of a total of 339 homicides[7] in six years (ie 7% of all homicides, rising to 9% of cases where ethnicity was recorded). What can safely be inferred is that Black people are disproportionately victims of homicide in Greater Manchester, since only 2% of the force population is Black. Caution is needed in breaking down these figures any further since the total is already small; but it is worth noting that this included only one Black female victim. Of the 23 Black males who were victims of homicide, by far the largest numbers were aged 19 to 25 (nine) and 26 to 35 (eight).

  By contrast, not only are the numbers of homicide cases in the MPS very much larger than anywhere else (with a total of 1,337) only four cases had no recorded ethnicity; and 422 of the total were Black, thus allowing scope for further analyses by gender and age.

  It is important to bear in mind that child murders are rare in any group; but Figures 6 and 7 suggests that while Black girls are slightly more at risk than Black boys through infancy and primary school, the number of Black boys who are victims of homicide rises sharply from the age of 10 onwards.

Figure 6


  In all age groups, Black people account for a disproportionate number of homicide victims of both sexes. However, by far the highest level of disproportionality is in the 10 to 17 age range where Black boys and young men account for nearly two thirds of the London total (see Figure 7).

Figure 7


  Additional information provided by the MPS for the financial year 2005-06 provides case-by-case details of all homicide victims in that year and includes information on method of killing. The overall figures highlight the much greater prevalence of firearms-related deaths in the case of Black victims compared to all other groups (Figure 8a). In fact, Black victims accounted for 15 out of the 22 cases of homicides involving firearms in that year.

Figure 8a


  Given the small number of Black deaths attributable to different methods of killing, further analyses must be treated with caution. However Figure 8b may be illustrative in that it shows that 10 out of the 11 firearms killings of Black men were of young men aged between 15 and 30. It should, nonetheless be borne in mind that a higher number of victims in this age group were actually killed with sharp instruments (ie probably a knife of some sort).

Figure 8b


3.  ENTRY INTO THE CRIMINAL JUSTICE SYSTEM

Searches

  Contrary to the claims made by many witnesses, police searches have little influence on the total numbers of young Black people who enter the criminal justice system. However, as the Committee has repeatedly heard, they are a major source of tension between the police and young Black people. Most of these searches are conducted under s1 of the 1984 Police and Criminal Evidence Act (PACE) which requires officers to have "reasonable grounds for suspicion" that the person they search is carrying illegal or prohibited goods before they can search them. However, the power under s60 of the 1994 Criminal Justice and Public Order Act allows officers to search anyone without needing "reasonable grounds for suspicion" and this power has also been used quite extensively in some areas in recent years. Each of these types of search is considered in turn.

s1

  As Figure 9 shows, s1 searches tend broadly to track the level of crime within different BCUs in both Greater Manchester and London. Certainly the use of the power is highest in the city centre BCUs which also record the highest level of recorded crime. Indeed, the second "spike" in the Manchester data (which interrupts an otherwise steady rise related to the level of crime) is in the Metropolitan BCU which is adjacent to the city centre. In London, though, while the level of searches on average tends to be higher in BCUs with higher levels of crime, there is considerable local variation. Further work would be needed to discover whether this was in any way correlated with the types of crime in different areas; but extensive experience of working with search data in different police forces suggests that differences between BCUs which do not appear to be crime-related tend to reflect local custom and practice rather than any strategic, shared force-wide understanding of how and where the power should be used.

  Nottinghamshire did not provide search figures broken down by BCU; but the pattern in the West Midlands (Figure 9c) stands out from the other two forces in that it does not show the expected peak in searches in the BCU with the highest crime levels.

Figure 9a


Figure 9b  Greater Manchester


Figure 9c  West Midlands


  When considering the ethnic breakdown of the s1 figures for 2005-06, it must be borne in mind not only that the police are now required to ask everyone they search to assign themselves to a Census ethnic classification (see earlier). Following recommendation 61 of the Macpherson report, they are now also supposed to record details (including ethnicity) for the far more numerous occasions where they stop and question a member of the public. Forces have designed a common form for recording both types of encounter; but this is necessarily far more complex than the previous search forms. As yet, no figures have been published for stops, suggesting that these returns are incomplete and/or otherwise deemed unreliable. Meanwhile, anecdotally, the additional requirement regarding stops, combined with the change in the form, has affected both the level and the quality of s1 recording in ways which may distort inter-ethnic comparisons.

  Most forces have continued to use officer classification alongside self-recorded ethnicity in the case of searches; and the MPS, Greater Manchester and Nottinghamshire provided returns on this basis—although the Greater Manchester returns are not entirely comparable with the other areas since these now include a "Mixed' category in addition to the "White", "Black" and "Asian" classifications.[8] However, returns from the West Midlands were based on the full Census breakdown and this may, in part, explain the fact that ethnicity was not recorded in 22% of cases in the West Midlands, compared to 4% in Nottinghamshire and less than 2% in both London and Greater Manchester. For the sake of comparability, "not recorded" cases have been excluded from the following analyses but this does mean that the figures for the West Midlands must be considered less reliable than for the other areas.

  Figure 10 confirms that the proportion of Black people in the s1 figures considerably exceeds the proportion of Black residents in the local population in all four force areas, although the extent of "disproportionality" varies between the forces. The figure is three times as high in London but five times as high in Greater Manchester and the West Midlands and six times as high in Nottinghamshire.

Figure 10


  There are important ethnic differences by age and gender within this picture; and these are broadly consistent across all forces. A higher proportion of s1 searches in the Black group fall on the younger age range, compared to Whites—in particular the group aged 18-20, although in Greater Manchester (where the average age of people searched tends to be lower for all groups) there are also marked ethnic differences among 10-17 year olds (Figure 11a).[9] Also, while searches on women are relatively uncommon in all groups, the gender ratio for the Black group is also different from that for Whites, with a higher proportion of all searches falling on men (Figure 11b).

Figure 11a


Figure 11b


  In sum, the overall level of Black disproportionality in searches is inflated by the higher than average levels of search among young Black men, as illustrated by the figures for Greater Manchester in Figure 12.

Figure 12


  To some degree the figures may also be inflated by the fact that Black people live in higher crime areas which tend to be more intensively policed and where the s1 power may be used more. However, as Figure 9 has illustrated, this relationship is by no means clear cut. The figures for London show an overrepresentation of Black people in all 32 boroughs relative to their presence in the local population. Uniquely, though, the MPS also keeps data on the ethnicity of suspects in reported crimes where the victim or witness can provide a description. As Figure 13 shows, with few exceptions, in London the proportion of suspects described as Black corresponds closely with the proportion of s1 searches on Black people.

Figure 13


  Suspect descriptions, though, are far more likely in the case of contact crimes. In 2005-06, the average number of suspect descriptions for each recorded case of burglary or theft/handling in the MPS was 0.4 and 0.3 respectively; but there was one for every recorded case of violence and the figure rose to 1.7 in the case of robbery (where the police may be more likely to have statements both from victims and witnesses). So involvement in contact crime—and robbery in particular—will itself incur a greater likelihood of being the target of an "intelligence-led" police search.

  The arrest rate from s1 searches is slightly higher than average for Black people in all three forces which provided these data[10] (Figure 14), although there are also local variations within this. Arrest rates for all groups appear to be lower than the force average in the city centre areas of both Greater Manchester and London (at 8% and 11% respectively).

Figure 14


  The arrest rate may also vary by age. However, in terms of the groups with whom the Committee is most concerned, further analyses of the London figures suggest that arrest rates are somewhat higher than average for young Black men, not only overall but compared with their White peers of the same age (Figure 15).

Figure 15


  Inasmuch as the arrest rate provides the only prima facie evidence of the extent to which searches are justified, therefore, it suggests that the police are equally likely to have "reasonable grounds" for suspicion when they use their s1 powers on Black people as they do in the case of White people. However, the s1 power of its nature sweeps up very large numbers of people who at the time of the search are not committing any offence, including large numbers who may not even have any history of offending.[11] Insofar as Black people are very disproportionately subject to s1 searches, this inevitably means that completely innocent Black people are indeed disproportionately targeted by the police; and this (as the Committee has repeatedly heard) constitutes a major source of grievance and alienation—even though the police may reasonably claim that their targeting does not constitute discrimination but is largely justified pre-hoc on intelligence grounds and post-hoc by arrest rates.

S60

  s60 of the 1994 Criminal Justice and Public Order Act allows officers to search anyone without needing "reasonable grounds for suspicion", albeit subject to the constraint that the order can only be applied in a designated area "in anticipation of serious violence" for a limited period on the authorisation of a senior officer. Because it does not require grounds for suspicion, the arrest rate from s60 is very much lower than for s1, averaging less than 3%. So the impact of s60 searches on innocent members of the public going about their legitimate business is much greater still than that of s1.

  The power was introduced on the basis that it was needed in connection with specific events where large crowds were expected to gather and there was the possibility of disorder, such as a football match or a demonstration. However, in a Panorama programme in 2000, West Midlands police claimed to be using the power as an effective tool for tackling street crime; and the unusual extent to which it is used in this force relative to others has been the subject of an inquiry by the Independent Police Complaints Commission, the fact of which is publicly known although no report has formally published.

  As Table 3 shows, in 2005-06, West Midlands police used the s60 power at a very much higher rate than either of the other two major urban forces.

Table 3

Use of s60 by selected forces 2005-06


s60 totals
Population
Rate per 000
population

MPS
11,541
7,164,990
1.6
Greater Manchester
  1,989
2,539,043
0.8
West Midlands
12,268
2,555,426
4.8


  However, as Figure 16 shows, the use of the power is very uneven within each of the force areas and is highly concentrated in a minority of BCUs. Also, it is only in Greater Manchester that one of the peaks for s60 corresponds with the BCU with both the highest crime levels and the highest rate of s1 searches (see Figure 9 above).

Figure 16

Use of s60 in different BCUs

Figure 16a  MPS


Figure 16b  Greater Manchester


Figure 16c  West Midlands


  In terms of the impact of s60 on Black people, the problem of missing ethnic data is again acute in the West Midlands where the figure for cases where ethnicity was "not known" rose to over 30% in the case of s60 compared to 22% for s1. By contrast, the "not known" figures for the MPS and Greater Manchester are comparable with those for s1, at around 2%. When these missing data are excluded, it is apparent that the extent of disproportionality is higher again in s60 searches than it is in s1 searches in each of these three force areas (Figure 17). It is in the West Midlands, though, that s60 has by far the greatest impact on Black people. For, even with the uncertainties created by the large number of non-recorded cases, in the force overall, the absolute number of s60 searches on Black people was at least equal to the number of s1 searches; and in the K2 division—where the highest number of s60 searches were recorded—the number of s60 searches was actually double that of s1 searches on Black people in 2005-06.

Figure 17


  Importantly for the committee, the use of s60 in the West Midlands impacts far more on the youngest age groups than s1. Whereas around 16% of White and Black people searched under s1 were aged 10-17 (see earlier), in the case of s60, the figure rose to nearly 40% for both groups.

Figure 18


  Finally, the West Midlands data did not include information on arrests from s60 searches; but the arrest rate for both the MPS and Greater Manchester was about 4% on average and was the same for Black people.

Arrests

  The Home Office s95 publication only provides information on arrests for notifiable offences. Analyses of the figures provided in the most recent publication suggest that, in 2004-05, arrests from s1 searches contributed less than 10% to the total arrest figure.[12] As Figure 19 illustrates, the contribution of s1 arrests was much higher than average in London, where the recorded use of the s1 power has always tended to be greater than elsewhere. In all areas, though, s1 arrests accounted for a slightly higher proportion of total arrests for Black people. The figure is still little more than a tenth overall, though; and the size of the disparity in the proportion of Black people arrested as a result of s1 searches in the figure for England and Wales is simply a reflection of the fact that the MPS accounts for 25% of all arrests from s1 searches but 69% of all s1 searches of Black people were conducted in London.

Figure 19


  In considering Black people's entry into the criminal justice system, therefore, it is more important to consider patterns in the totality of arrests than to focus on searches. Rather, the concerns associated with searches pertain more to people who are not arrested (see earlier) than as a route of entry into the system.

  Compared to their presence in the local population, Black people were overrepresented in the arrest figures provided by all four forces to the inquiry (see Figure 20). There was, nonetheless, some variation in the extent of this disproportionality;[13] and the pattern broadly followed that for s1 searches in that the level of Black overrepresentation tended to be slightly lower in London than elsewhere.

Figure 20


  Also, while a higher proportion of these arrests occur in areas with larger Black populations, the disproportionality (as with searches) occurs across all BCUs, irrespective of their level of crime or the size of their Black populations, as the picture in London at Figure 21 illustrates. The extent of this overrepresentation appears much higher than average in a small number of boroughs, though; and the reasons for this might usefully be explored.

Figure 21


  It is only in the MPS that Black arrestees tend to be younger than Whites (Figure 22a); and again this overrepresentation of Black young people appears to be amplified by the higher proportion of Black young men who are arrested whereas there is less disparity in the arrest rates of Black and White women. Males accounted for 86% of 10-17 year old Black arrestees in London, compared to 82% for Whites in the same age group.

  Also, the White people arrested in Greater Manchester and in Nottinghamshire appear to be younger than those in the Black group; but the figure for 10-17 year olds in the "Mixed" group in Greater Manchester in particular is a further reminder of the growing importance of this group for the future. This is also illustrated by the ethnic make up of the different age bands in Nottinghamshire (Figure 22b) where the arrest data were presented on the 16 point Census classification. In addition, the aggregated figures used here for the West Midlands can similarly be unpacked into the Census categories, showing that 35% of arrestees of Mixed White-Black Caribbean heritage were aged 10-17, compared to a below average figure of 18.5% for the Caribbean group.

Figure 22a


Figure 22b


  The offences shown in Figure 23, it should be noted, cover arrests for notifiable offences only. The detailed statistics provided by several forces included figures for non-notifiable offences as well. In the MPS these accounted for about a third of all arrests on average; and although the figure was lower in the younger age ranges (at around 15% for both Black and White arrestees aged 10-17) this has some implications for the youth justice statistics which are discussed further below.

  With regard to notifiable offences, there are important ethnic differences in the offences for which people are arrested; but these also vary between areas, in part reflecting the particular crime problems of each force. Also, as the figures for Nottinghamshire and Greater Manchester illustrate, there are both similarities and differences between the "Mixed" and the "Black" groups which may have implications for the future. In all areas, though, the Black—and to a varying extent the "Mixed" group—are more likely to have been arrested for drugs and for robbery offences than other groups.

Figure 23a


Figure 23b


Figure 23c


  However, these broad offence categories themselves also mask a wide range of specific offences, some of which will be more likely to culminate in a custodial offence if a charge is brought and successfully prosecuted (see also next section). The "drugs" category is particularly interesting in this context, not least since drugs offences are "victimless" so the numbers in any force area are largely dependent on proactive policing.

  A specific additional request to the MPS for a breakdown of drugs arrests sheds further light on ethnic differences within the overall "drugs" category; and the detail provided within the Nottinghamshire figures provides additional, important insights.

  As Figure 24a illustrates, White people arrested for drugs offences in London are more likely to be charged with personal possession. By contrast, Black people—even at the youngest ages—are more likely to be charged with possession with intent to supply. These differences may go a long way to explaining the fact that Black people have not equally benefited from the change in policing policy with regard to drugs in the MPS in recent years (Figure 24b) The additional information provided by the MPS to the inquiry shows that the fall since 2003 has largely been driven by a fall of 50% in arrests for personal possession, so White people have benefited very much more from this. Arrests for possession with intent to supply, however, actually rose at the same time by over 20% and this will have had a disproportionate impact on the Black group, offsetting any reductions in those arrested for personal possession.

Figure 24a


Figure 24b


  In Nottinghamshire, a more refined breakdown is available within the drugs arrest category than in any of the other forces' data. This shows that, in the case of Black arrestees, arrests for supply are also more likely to involve class `A' drugs (Figure 24 c). In addition, they highlight a more general issue which may be of relevance to the inquiry, which is that, even within the same ethnic group, patterns of offending tend to change as people get older.

Figure 24c


  As Figure 25 shows, notwithstanding important area differences in the extent of their overall involvement in drugs offences and in robbery, in both London and Nottinghamshire robbery is more common among Black juveniles. As they get older, though, young Black people appear to move out of robbery and into drug-related offending.

Figure 25a


Figure 25b


  Their disproportionate involvement in both robbery and drugs offences has further implications for the type of disposal young Black people face relative to others at the post-arrest stage.

4.  DISPOSALS

  There is a significant gap in ethnic data at the disposal stage due to the absence of ethnic monitoring data published by the CPS. However, all four forces were able in principle to supply information on disposals by ethnic group. In the event these data were received from only three forces. Of these, however, Nottinghamshire provided information where the number of cases of "unknown" ethnicity was consistently very much larger than the ethnic minority total (and in some instances larger than the figure for White people). So it has only been possible to use the disposal data provided by Greater Manchester and the MPS.

  The two forces use slightly different categories for disposals and both issue some caveats about the coverage of the figures.[14] The overall picture which emerges from both areas (Figure 26) suggests that a higher proportion of all arrestees in Greater Manchester than in London were subject to "No Further Action".[15] However, they also suggest some differences between the two forces in the disposals of Black arrestees relative to Whites. In London, Black people are more likely, on the one had, to be charged with an offence rather than cautioned or given a formal warning but, on the other, to have no further action taken against them (shown as "not proceeded with"). Importantly, though, Greater Manchester again highlights the extent to which differences may be more acute as between the White and Mixed group, suggesting that any figures for the Black group may look different where individuals of Mixed heritage are excluded from the totals. In Manchester it is the Mixed group which has the highest rate of "no further action"; and a lower proportion are charged with any offence.

Figure 26a


Figure 26b


  This broad overview picture may tell us little, though, and could actually be misleading inasmuch as it does not reflect a number of factors which should properly affect disposal decisions, including the following.

  Eligibility for a caution or warning depends on:

    —    the arrestee's admitting the offence;

    —    their previously criminal history (including whether they may not be eligible because they have previously been cautioned); and

    —    the seriousness of the offence.

  The decision to charge, in turn, will be influenced by these factors; but, in addition, where the arrestee does not admit the offence, the police and CPS will require sufficient evidence—not least in terms of the active co-operation of victims and witnesses—to ensure a reasonable chance of conviction.

  The likelihood of being charged with an offence is lowest among younger people who are de facto less likely to have any previous convictions or warnings. This is well illustrated by the Greater Manchester figures which have the further advantage that they subdivide the 10-17 age range. As Figure 27 shows, the age profile of the Mixed group here is very much younger and they are actually far more numerous than the Black group among 10-17 year olds. It is also a reminder that the total figures for disposals in Figure 26 above are determined by decisions for adults. They bear little relation to what is happening to people in the age ranges of particular concern to the inquiry. However, the numbers for 10-13 year olds in Manchester are so small, especially for the minority groups, that any further breakdown by disposal type would be unreliable.

Figure 27


  In the 14-17 age group, Black young people appear more likely to have no further action taken against them, whereas the Mixed group is more likely to be charged with an offence. For both groups, though, the rate of cautioning is lower than for Whites in each age band (Figure 28).

Figure 28a


Figure 28b


  The figures for London do not distinguish people of Mixed heritage from the Black group; but they are large enough to be disaggregated by offence type. As Figure 29 shows, some of the overall differences in disposals between Black and White young people may be related to differences in the type of offence for which they are arrested in the first place. In particular, over 80% of young people arrested for robbery are charged, including those in the lowest age range; so their overrepresentation in the robbery figures already increases the higher charge rate for young Black people overall and it lowers their eligibility for a caution. However, there are also ethnic differences within offence categories, some of which are replicated in both age bands. In general, the proportion of cases which result in no further action ("not proceeded with") is slightly higher for Black arrestees, with the exception of 18-20 year olds arrested for violence against the person.

Figure 29a


Figure 29b


  In both age groups, though, the proportion of Black people who are charged following an arrest for violence is much higher than for White arrestees; but further work would be needed[16] in order to establish whether this was because the offences involving Black young people are more serious, because those involved are less likely to admit the offence, because they are ineligible for a caution or warning owing to previous convictions or for any other reason and combination of reasons. By contrast, young Black people tend to be far less likely than their White counterparts to be arrested for theft and handling offences; but those who are are more likely to receive a caution or warning. Again, more detailed work would be needed to establish whether, for example, these offences are more likely to be first time offences for Black young people, and/or whether the nature of the thefts involved is different. For the "theft and handling" category is very large, spanning shoplifting, theft from the person and the main types of notifiable motor vehicle crime (that is, theft both of and from motor vehicles).

5.  10-17 YEAR OLDS KNOWN TO THE YOUTH JUSTICE SYSTEM

General

  In addition to the police data which is the basis of the earlier sections of this paper, the Youth Justice Board (YJB) provides a range of information on the offences committed by young people aged 10-17 as well as on the bail/remand decision and disposals. This is based on standardized returns from the Youth Offending Teams (YOTs) to the YJB and is available on the practitioners' portal within the YJB website broken down by region and by individual YOT within this. It is potentially a very rich dataset (subject to the caveat that the ethnic breakdowns are not available by gender or age) and only limited analyses have been undertaken for this paper.

  The YJB data do not map readily onto the police statistics in two particular ways; and these are worth clarifying before looking at any further insights the YOT data offer the inquiry.

  In the first place, there appears consistently to be a shortfall between the number of "cautions" the police record and the number of young people known to the YOTs who have been subject to any form of "Pre-court decisions"' (ie police reprimands and final warnings). The figures in Table 4 are taken from the s95 publication, in view of the absence of age-related disposal data from two of the four police forces who responded to the inquiry's request for further information. They therefore refer to the financial year 2004-05; but the information we received from Greater Manchester shows a similar picture for 2005-6, when the police recorded 8,682 cautions of 10-17 year olds whereas the YOT total for pre-court decisions was 4,717.

Table 4

Pre-court disposals of young people aged 10-17 2004-05 Police vs YOT data


National
MPS
Greater Manchester
Nottinghamshire
West Midlands

YOTs
85,370
6,397
4,346
2,062
4,442
Police
101,202
10,819
5,926
2,388
5,186


  If, as appears to be the case, the youth justice system is unaware of significant numbers of young people dealt with by the police, this may of itself be a particular matter of concern. This is not least because, inasmuch as they are eligible for this type of disposal, many—if not most—of the young people in question may be first time entrants to the system; but this might imply that some young people who have already been reprimanded or warned by the police for an offence only come to the attention of the YOT once they have re-offended. That is, they may miss out on being included in relevant programmes to prevent their becoming further involved in crime.

  The second area of apparent mismatch between the police figures—and, in particular, those published by the Home Office—is that a substantial number of young people known to the YOTs have received disposals for what appear to be non-notifiable offences (see Figure 31). These will include the various breach categories, several offences within the "criminal damage" and "public order"' categories, as well as many "motoring offences" (as distinct from vehicle theft). The latter account for a significant number of the offences which bring young people into the system and they will include driving without insurance, allowing oneself to be carried in a stolen vehicle, various vehicle defects and minor road traffic violations.

  Together these imply that we have a less than complete picture of young people of all ethnic origins at the earliest stages of their involvement in crime and the criminal justice system. This may mean that inter-ethnic comparisons are less reliable than has been assumed;[17] but, importantly, it may also be an obstacle to effective early intervention and limit the scope for evaluating the effectiveness of preventative work. These factors may also help to explain why the YJB figures give an impression of `disproportionality' which is far less stark than the picture considered thus far. Figure 30 suggests that Black young people are represented within the juvenile justice system at about twice the rate of their presence in the population at large and that there is no overrepresentation of the Mixed group.

Figure 30


Patterns of offending

  The YJB data routinely present figures for "Black" young people separately from an omnibus "Mixed" category. Overall, they confirm that there are not only patterns of ethnic differences in the offences for which young people come to the attention of the criminal justice system but also important area differences.

Figure 31

Offence characteristics of 10-17 year olds subject to disposals 2005-06

a.  MPS


Figure 31b.  Greater Manchester


Figure 31c.  Nottinghamshire


Figure 31d.  West Midlands


  A common feature is the overrepresentation of the Black group in the robbery figures; but this still accounts for only a minority of all the offences for which this group was subject to disposals. Added to this (although the numbers are smaller still) is an overrepresentation for drugs offences relative to other groups, especially in Nottinghamshire. White young people, by contrast, are much more likely to have entered the system as a result of anti-social behaviour related offences such as criminal damage and public order. The Mixed group often occupies an intermediate position between the two but tends to be more comparable with White people in the relatively high proportion of cases involving motoring offences in all areas. Finally, the higher proportion of young people supervised by the YOTs in Greater Manchester for breaches of statutory orders may reflect the much higher than average use of ASBOs in that city. However, the proportion who are breached is noticeably higher for the Mixed group than for White young people in the West Midlands.

Remands and disposals

  The YJB figures provide a stark picture of ethnic differences at opposite ends of the disposal spectrum, which is broadly consistent across all areas (Figure 32). Not only is the Black group less likely to receive a pre-court disposal and more likely to be sentenced to custody, this is even more true of the Mixed group, despite (as noted above) the fact that their offending profiles are more similar to those of White young people in the system.

Figure 32


  Of those not dealt with pre-court, a higher proportion of both Black and Mixed heritage young people are remanded in custody (Figure 33)—a fact which, of itself, will increase the likelihood of a custodial sentence if they are found guilty.

Figure 33


  These YJB figures do not provide any further breakdown for disposals by offence type; and still less is it possible to take account of other potentially relevant variables, such as seriousness within offence type, age and previous convictions. However, in addition to their greater likelihood of being sentenced to custody, both Black and Mixed heritage young people tend to receive longer custodial sentences; and this figure is boosted by the higher proportion who are sentenced under section 90 and 91 powers which apply in cases where an adult would receive at least 14 years in custody.

Figure 34


  Finally, it is only in the context of custody that information provided specially to the inquiry by the YJB provides some insight into the ethnic make-up of the Mixed heritage group. So the picture in Figure 35 may be accurate only for the minority of young people convicted of the most serious offences. However, it does suggest that people of Caribbean group account for a higher proportion of young Black people in custody than would be expected from their presence in the population at large, while individuals of White and Caribbean heritage predominate among the `Mixed' young offenders sentenced to custody.

Figure 35a


Figure 35b


6.  ISSUES RAISED

  The analyses made possible by the additional information provided by the four forces have clarified a number of issues of particular relevance to the inquiry—not least insofar as their figures included age and BCU-level breakdowns which are often unavailable in the published data. In addition, the Committee may wish to consider four more general issues, the first two of which are related.

  Firstly, it is evident that, in principle, police forces themselves have sufficient ethnic data to explore some of the possible reasons for the overrepresentation of young Black people within their own force area. That is, they should be able to use their data to monitor whether there are any prima facie areas for concern for the force—whether in terms of officer decision-making or the impact of force operational practices—and then to pinpoint exactly where within the force these may be occurring. Qualitative approaches will then be needed to follow these up and, as necessary, take any remedial action. In practice, however, drilling down into the statistics in this way soon means looking at numbers which may be too small for reliable inter-ethnic comparisons, even in areas where the total numbers, force-wide are fairly large. This should not preclude any force from adopting this type of approach, though; for where the numbers are small, this would make it possible to compare individual cases on a like-for-like basis.

  It is by no means evident, however, that forces have been using their data in this way, despite the fact that they have now been required to keep ethnic statistics on specified items for over ten years. Ironically, many appear routinely to be adding an ethnic marker to other statistics as well; yet if ethnicity has not been recorded in a high proportion of cases, this simply seems a waste of effort. Especially where the `not known' category is very much larger than the total for minorities, any attempt to analyse the figures could produce seriously misleading results. The issue of missing data is of particular concern with regard to items where forces are actually required by the Home Office to keep ethnic information—though in the case of police searches this may be an unhelpful side effect of the requirement to use self-classification and additionally to record stops.

  Secondly, the figures provided by the four forces have illustrated some broad, general patterns (many of which were already known); but they have also highlighted important local differences. These include differences in the composition of the "Black" population and the ways in which this is itself changing over time, as well as differences in the impact of policing practice on local populations which may have specific implications for young Black people. At the same time, they have also highlighted apparent area differences in patterns of offending by young people which are reflected in the local patterns of offending by different ethnic groups.

  This adds to the importance of forces taking ownership of exploring their own data in the light of local circumstances and the force's particular policies and practices. At the same time, these grounded understandings of what is happening to different groups of young Black people need to feed upwards to inform guidance on policy and practice from the Home Office, ACPO, HMIC etc. For national, one-size-fits-all approaches may not be relevant to many local situations and it is even arguable that these may further be counter-productive, especially if they appear to take away any local responsibility for exercising quality control over the data in order to analyze and use them. It seems highly likely that in many areas, s95 data are seen simply as figures which have to be collected to meet a Home Office requirement, albeit in a context where the Home Office has set great store on reducing disproportionality as an end in itself without first exploring with any rigour the many factors which contribute to disproportionality and the relative importance of each, including many socio-economic and demographic factors which significantly impact on the likelihood of different groups coming to the attention of the criminal justice system and over which the police themselves have no influence. Yet, as the Home Office itself has acknowledged to the Committee, this has left police forces without any idea of what the figures should look like; for, even if they themselves were behaving fairly towards all sections of the local population, the police would still come across some groups within that population more than others both as victims of crime and as suspects.

  Unsurprisingly in these circumstances, levels of disproportionality have proved intractable over the past decade, despite huge investments in training, tightening of procedures and lots of centralized guidance, especially with regard to stop and search, since these have all been premised on an assumption that disproportionality has largely been driven by conscious or unconscious discrimination on the part of the police. New approaches are needed to understanding the reasons for disproportionality which, as this paper has illustrated, are likely to be complex and may vary by area. Only in this way will it be possible to identify and isolate whether, how and where discrimination is contributing to the overall level of disproportionality and to take any necessary action. And this will require local forces to take the lead.

  Thirdly, the Youth Justice Board figures, when juxtaposed with the police data, have highlighted the fact that the published s95 figures give an incomplete picture of what is happening to young people of different ethnic origins as they first enter the criminal justice system. They also suggest the need for improvement in the exchange of information between the police and local YOTs on this age range. Unless both of these issues are addressed, it will inhibit any coherent approach to prevention and diversion in this critical age range.

  In addition, while the data available from the Youth Justice Board are very rich, their potential for informing policy and practice with regard to minorities will remain limited unless it is possible additionally to analyse them taking account of relevant variables—in particular age, gender and offence type/seriousness within the broad categories used in the published figures. Some YOTs (as the Committee's visit to Southwark illustrated) have taken this approach within their own local data and, as with the police, all should be encouraged to do so. However, the Board itself needs to have this information and to use it intelligently if it is to develop a coherent approach to understanding the reasons for any ethnic differences within the youth justice system and to monitoring the effectiveness of any interventions to address these.

  Consideration also needs to be given to the reasons for and implications of the lower levels of disproportionality in the YJB data compared to the police statistics which are more commonly cited in this context. One possible reason suggested above is that the picture may look different when non-notifiable offences are included. The fact that both Black and Mixed heritage young people have disproportionately worse outcomes once they are in the system may in part be because they are more likely to be subject to disposals for notifiable offences than their White peers (although this could only be established if further breakdowns were available within the main offence categories, as suggested above). However, another factor may be that the YJB figures are more likely to reflect the number of individuals of different ethnic origins within the juvenile justice system at any given time.[18] For an important caveat about the police data which is often overlooked (including in earlier sections of this paper) is that these count incidents not individuals. In practice the same individual may show up in the same set of search and arrest statistics more than once within a 12 month period; and this may tend especially to be the case for individuals who are already "known" to the police. In other contexts it has become a commonplace to assert that most crime is committed by a relatively small group of individuals. So if there are any significant ethnic differences in the numbers of "repeaters", this too could give a misleading impression of the level of "disproportionality".

  Importantly, if the search and arrest figures are indeed inflated for some groups by a disproportionate number of repeaters, this would not only raise possible questions about policing practice. It would also imply that discussing these ethnic statistics as if they referred to individuals may have a distorting effect on perceptions of the group as a whole, both within the police service and in the public at large. Over time, especially as the considerable investment in reducing "disproportionality" appears to have had no impact, there is a real danger that the published figures will simply reinforce negative stereotypes of criminality among Black young people in general which, as has repeatedly been referred to in their testimony to the Committee, is understandably a major cause of resentment.

  This, in turn, adds weight to the need for better quality ethnic data, as well as for more intelligent approaches to analysing and using it both nationally and locally. In addition to identifying the weaknesses in what currently exists and the scope for improvement, this paper also serves to illustrate its considerable unrealised potential.

Marian FitzGerald

Visiting Professor of Criminology

Kent Crime and Justice Centre, University of Kent

21 March 2007







2   The West Midlands police, unlike the other forces, did not provide population figures using the full 16 point Census classification, including only a composite "Mixed" group which will include unknown numbers of White-Asian and other Mixed heritage individuals. The figures shown in Table 1, therefore, are taken from the 2001 Census for the West Midlands Metropolitan district where the Black population forms a somewhat lower proportion of the total (2.8%) than in the figures provided by the police (3.7%). It is probably safe to assume, though, that the internal composition of the force's Black population is broadly similar to that shown in Table 1. Back

3   Greater Manchester supplied a full ethnic breakdown only for local authorities within the force area; but the Manchester City Council area (which covers over half of the force's total Black population) is covered by three BCUS-North Manchester, Metropolitan and South Manchester). Back

4   In this context it is worth noting that the "Airport" division in Greater Manchester and Heathrow in London are two examples of the "pure" case of BCUs which contribute to the crime totals for the force but effectively have no resident population. Back

5   Data from the MPS, however, suggest that the allocation of uniformed officers is slightly lower than might be expected in Hackney and Lambeth which are two of the boroughs with the largest Black populations. Back

6   There is no record, for example, in the Black female category of the case of Danielle Beccan who was shot dead in October 2004 and whose murder featured prominently in the national press. Back

7   These exclude the victims of Harold Shipman. Back

8   For the purposes of analysis here, only the "Black" figure has been used; but this necessarily under-estimates the experience of Black people in Manchester relative to the other forces since a high proportion of the "Mixed" group will also be Black and their inclusion might be expected to add up to 40% to the Black total (see Table 1). Back

9   No further analyses are included for Nottinghamshire in this section of the report, in view of the much smaller numbers in Nottinghamshire, the fact that their data were received later and some unresolved queries about the search tables. Back

10   The White arrest rate in London seems improbably high and may provide support for the anecdotal evidence for one aspect of the fall off in recording quality referred to earlier which may, in turn, have implications for "disproportionality". Sources in different forces have spontaneously suggested to me that officers have now reverted to recording more conscientiously those searches where they feel they may be open to scrutiny. This tends to include those which involve minorities and/or which result in some further action. Those which are less likely to be recorded, therefore, will include searches of White people which do not result in an arrest and where officers may feel less vulnerable to complaint. Back

11   In my study of searches in London in 1999, I asked for checks to be conducted on a large sample of s1 searches which had not resulted in an arrest to see whether the individuals targeted had any criminal record. An equal proportion of both Black and White subjects of these searches (50%) were not known ever to have offended. Back

12   The publication does not make clear whether the s1 arrest data are similarly limited to notifiable offences. However, if-as seems likely-they also include non-notifiable offences, their contribution to these published arrest figures will be even more modest. Back

13   In the case of Greater Manchester and Nottinghamshire account must additionally be taken of the "Mixed" group for whom arrests figures were given separately; so the population figure for Black people in Figure 20 is based only on the main Black groups. In Greater Manchester these accounted for 3.3% of arrestees, compared to 1.3% of the population; and the figures in Nottinghamshire were 3.3% compared to 1.2%. Back

14   Thus the MPS provided comprehensive data on everyone who passed through the force's custody suites in 2005-06, including people who had not been arrested-for example because they were en route to or from prisons or had been taken to police stations as a "place of safety". Of those who had been arrested, by no means all had a recorded disposal, in part because final decisions may not yet have been made or entered onto the system at the time the figures were provided. Back

15   This may to some degree be affected by the much higher numbers of cases with no recorded disposal in the MPS data. Back

16   This could not be done using the aggregate data provided for the purposes of the current paper, detailed though that is. It would require analysis of the detailed information in a sample of individual cases. Back

17   A further complication in this context is that many of the figures cited refer to incidents (whether of stop/search, arrest or disposal) but they are discussed as if they related to individuals. This point is discussed further in the concluding section of the paper. Back

18   In practice, of course, it is possible for a young person to re-offend and re-enter the system in the same year; and those shown as "breaches" will de facto tend to be double counted. So too will young people on orders of more than a year, or whose order spans two years since they will show up in the totals for successive years. For this reason, it would also be useful for the YJB to present figures for new entrants to the system in any given year, as with the prison statistics which show "admissions" separately from the prison population. Back


 
previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2007
Prepared 15 June 2007