ANNEX B
Further supplementary note by the Home
Secretary
POLICE NUMBERS AND POLICE PERFORMANCE
When I gave my annual evidence to your Committee
at the end of last year I said I would provide information concerning
what assessments have been made of the correlation, at police
force level, between the change in number of police officers and
the change in crime levels. I also promised to let you have the
latest figures on police numbers.
Police numbers and performance on crime reduction
The last assessment of this was made in the
summer in reply to a written Parliamentary Question by Tony Wright
(Great Yarmouth). Officials in my Research, Development and Statistics
Directorate assembled information on changes in police numbers
and recorded crime rates between (1) 1996-97 and 1998-99; and
(ii) 1993-94 and 1998-99. This showed that between 1996-97 and
1998-99, of the 21 top performing forces (ie, those with most
improvement in their crime figures), 12 had falls in police numbers
over the last two years. In the other group of forces (which performed
rather less well), slightly more—14—had falls in officer
numbers. At the individual force level, however, there seemed
little obvious link between performance and police numbers. The
second five-year analysis also showed that there was a very mixed
picture overall, with the same very weak statistical relationship
between police numbers and changes in crime over the period.
My officials have now updated this analysis,
bringing in figures to March 2000. This compares (a) the change
in the number of offences in police force areas between 1996-97
and 1999-2000, compared to (b) the change in number of officers
available for duty on March 1996 (the beginning of the period)
and March 2000. Figure 1 plots the percentage change in each for
each force.
There is a slight tendency for forces with the
least good record on crime to have higher falls in numbers, but
the overall picture indicates no interpretable statistical association.
(This is measured by the r-squared value on Figure 1, which is
very low.) This analysis excludes Gwent because boundary changes
in 1997 increased officer numbers. The City of London is also
excluded because of small numbers. The analysis, then, is much
in line with that done earlier.
Table 1 presents the same data in a different
way. It ranks forces according to the size of their crime falls
(only four forces had increases in crime). It shows their rank
order in relation to change in police numbers, as well as the
actual percentage change in officer numbers. As with the previous
analysis, the forces are divided in the top 21 best performing
ones, and the remaining 20. In the top performing forces, the
average fall in police numbers was minus 1.0 per cent. In the
other forces, the average fall in police numbers was rather higher
at minus 3.3 per cent. But the number of forces in the two groups
with falls in officer numbers was virtually the same.
Although these analyses show that there is little
direct relationship between falls in police officers and poorer
force performance in reducing crime, there are a number of important
caveats. Local factors will influence police numbers, and crime
levels themselves will affect resource allocation and hence the
number of officers available. Changes in local crime rates will
also be affected by local factors, such as demographic and economic
changes.
So, it is not a matter of just counting helmets.
What the police do is as important as how many of them are available
to do it. Publications such as Getting the Grease to the Squeak—research
lessons for crime prevention, published in the Home Office's
Crime Detection and Prevention Series, show that effective policing
can drive down crime. And it is why we have asked police authorities
to set up challenging targets under Best Value to reduce burglary,
vehicle crime and robbery—and why we are providing better
technology to deliver more and better policing. The number of
police on the streets, as you know, is also vitally important
to the public.
Latest police numbers
Figures were published at the end of August
showing the total police officer strength on 31 March 2000. The
most up-to-date figures relate to 30 September 2000. There was
a revision to these made on 22 December, due to a mistake in the
figures provided by the Metropolitan Police Service. The figures
are below.
| Total excluding secondments (1)
| Secondments | Total police strength
|
| 31 March 2000 | 121,956 |
2,214 | 124,170 |
| 30 September 2000 (2) | 122,230
| 2,384 | 124,614 |
(1) Excludes secondments outside the police service in England
and Wales (eg, to the private sector, or to law enforcement agencies
overseas).
(2) Includes revised figures for the Metropolitan Police Service,
issued on 22 December. The total police strength in the MPS was
24,695 (revised from 24,244).
I am placing a copy of this letter and related figures in
the Library.
Jack Straw
9 January 2001
Excludes Gwent and the City of London.
Changes to the counting rules, introduced in April 1998,
have been taken into account at force level.
TABLE 1
CHANGES IN RECORDED CRIME RATES 1996-97 TO 1999-2000 AND
CHANGES IN POLICE NUMBERS MARCH 1996 TO MARCH 2000
| % change in crime
1996-97 to 1999-2000
| Rank of change in police numbers
March 1996-March 2000
(1 = biggest increase)
| % change in police numbers
March 1996-March 2000
|
| Kent | —25 | 9
| 2.7 | rise |
|
| Northumbria | —23 | 6
| 3.3 | rise |
|
| Lancashire | —21 | 17
| 0.2 | [rise] |
|
| Lincolnshire | —18 |
26 | —2.6 | FALL
| 1 |
| Cleveland | —17 | 19
| —1.2 | FALL | 2
|
| South Yorkshire | —16 |
8 | 2.9 | rise |
|
| Durham | —15 | 1
| 11.2 | rise |
|
| South Wales | —15 | 31
| —3.3 | FALL | 3
|
| Avon and Somerset | —14
| 21 | —1.6 | FALL
| 4 |
| Cumbria | —14 | 28
| —2.8 | FALL | 5
|
| Devon and Cornwall | —13
| 24 | —2.0 | FALL
| 6 |
| North Yorkshire | —13 |
29 | —3.1 | FALL
| 7 |
| West Mercia | —12 | 35
| —6.4 | FALL | 8
|
| Gloucestershire | —12 |
22 | —1.6 | FALL
| 9 |
| North Wales | —11 | 11
| 1.9 | rise |
|
| Nottinghamshire | —10 |
32 | —4.9 | FALL
| 10 |
| Humberside | —10 | 33
| —5.3 | FALL | 11
|
| West Yorkshire | —10 |
34 | —6.2 | FALL
| 12 |
| Dorset | —10 | 5
| 3.4 | rise |
|
| Cheshire | —10 | 16
| 0.6 | [rise] |
|
| Cambridgeshire | —10 |
18 | —0.1 | [fall]
| [13] |
| Hampshire | —9 | 10
| 2.2 | rise |
|
| Dyfed Powys | —8 | 3
| 4.9 | rise |
|
| Leicestershire | —8 |
4 | 4.5 | rise |
|
| Warwickshire | —6 | 38
| —8.1 | FALL | 1
|
| Bedfordshire | —6 | 41
| —8.8 | FALL | 2
|
| Hertfordshire | —6 |
7 | 3.2 | rise |
|
| Essex | —6 | 27
| —2.7 | FALL | 3
|
| Wiltshire | —5 | 40
| —8.3 | FALL | 4
|
| Northamptonshire | —5 |
30 | —3.1 | FALL
| 5 |
| Norfolk | —5 | 20
| —1.4 | FALL | 6
|
| Surrey | —4 | 2
| 8.6 | rise |
|
| Derbyshire | —3 | 13
| 0.8 | [rise] |
|
| Staffordshire | —2 |
23 | —1.8 | FALL
| 7 |
| Sussex | —1 | 39
| —8.2 | FALL | 8
|
| Merseyside | —1 | 37
| —7.4 | FALL | 9
|
| Suffolk | —1 | 15
| 0.7 | [rise] |
|
| Metropolitan Police | 2 | 36
| —6.8 | FALL | 10
|
| Thames Valley | 4 | 12
| 1.8 | rise |
|
| West Midlands | 5 | 14
| 0.7 | [rise] |
|
| Greater Manchester | 9 | 25
| —2.1 | FALL | 11
|
| Top 21 forces | —15 |
| —1.0 | |
|
| Bottom 20 forces | 1 |
| —3.3 | |
|
City of London and Gwent are excluded.
The changes to the counting rules, introduced in April 1998, have
been taken into account at force level.
|
