Do Thames Valley Police disproportionately stop-and-search ethnic minorities?

Photo (c) Reading Borough Council

Katesgrove councillor Sophia James took the Thames Valley Police and Crime Commissioner (PCC) Anthony Stansfeld to task over disproportionate police stop-and-search at Reading Borough Council’s policy meeting on 29 October.

PCC Anthony Stansfeld was explaining the steady increase in knife crime in the Thames Valley Police (TVP) region from 60-70 per month in 2015 to about 120 per month in 2018.

“A lot of this is how effectively you stop-and-search,” said PCC Anthony Stansfeld. “If you don’t stop-and-search, knife crime goes up. This has been demonstrated very clearly in London, where stop-and-search almost ended and knife crime increased exponentially.”

“There’s a lot of cry about oh it’s not proportionate, but… west Berkshire has the most disproportionate stop-and-search [in the TVP region], and also the most effective; the reason was they were picking up drug gangs coming out of London, and they were very clear who they were.”

Councillor Sophia James at the Black History mural

Councillor Sophia James asked the PCC a question. “I was quite disappointed to see the national figures about stop-and-search, black people being nine times more likely to be stopped than any other group.”

“I’m quite appalled and surprised to hear our PCC talk about the fact that there’s a lot of cry about disproportionate stop-and-search,” she continued. “Yes, there is a lot of cry, because there’s clearly a misunderstanding about how it impacts and affects the black community. If you’re driving in your own car getting stopped-and-searched and hassled by the police; if you’re walking into a shop or down the street and getting stopped-and-searched and hassled by the police because you’re black, I think there’s absolutely a misunderstanding and I wonder how exactly is the public supposed to trust the police when there are attitudes within it like that?”

“Most of the stop-and-searches, when I was talking about disproportionality in west Berkshire, were not west Berkshire residents,” said PCC Anthony Stansfeld. “Also, the murder rate among certain communities in London is hugely higher, and it’s within their own communities these murders are taking place… I would like to see [stop-and-search] reduced a great deal, but I’d also like to see the murder rate come down a great deal.”

“For a number of years we’ve had an independent advisory group [IAG]… just dedicated to looking at our stop-and-searches at a force level,” said TVP chief constable Francis Habgood. “We audit our stop-and-search forms and the IAG randomly picks stop-and-search cases. They are all recorded with body-worn video… We’ve done a lot of training on unconscious bias, on legislation and what are reasonable grounds [for stopping and searching].”

What does the data say?

National statistics on police stop and search are published by the government. They say:

Between 2010/11 and 2014/15, the likelihood of black people being stopped and searched fell from 6 times that of white people to 4 times that of white people; it then rose again to just over 8 times more likely in 2016/17.

Thames Valley Police operate in a well-defined geographical area [note 1] and the ONS’ latest ethnic data for this region comes from the 2011 census [note 2].

Thames Valley Police regularly publish data about how many people they stop and search in their area, broken down by ethnicity, age and gender.

It is evident that by 2017/18, the total number of TVP stop-and-searches had declined to slightly more than half of its 2015/16 level [note 3].

The same data can be expressed as a percentage of all stops that year [note 4].
This suggests that as the total number of TVP stops decline over the years, the proportion of that declining number of stops affecting non-white ethnic groups increases.

The (approximate) likelihood of being stopped and searched based on ethnicity can be calculated by comparing the numbers of stop of search for each ethnicity with the population of each ethnicity  [note 5].
This suggests that the chances of being stopped and searched by TVP has declined for all ethnic groups over the last three years, except ‘mixed’ and ‘other.’

One can make a rough measure how much more likely an ethnically black person is to be stopped than other ethnic groups by creating ratios of these likelihoods  [note 6].

Period Black compared with white Asian compared with white Black compared with Asian
September 2015 – August 2016 6.49 1.87 3.47
September 2016 – August 2017 8.21 2.00 4.10
September 2017 – August 2018 9.39 2.39 3.93

This suggests that in the TVP region in 2017/18, you were (approximately) nine times more likely to be stopped by police if you were black rather than white, and twice as likely if you were Asian rather than white, and four times more likely if you were black rather than Asian. TVP’s own evidence suggests that although they have been reducing the number of stops for everybody over the last three years, this reduction is more obvious amongst the ethnically white population.

Police stop data also includes outcomes; that is, whether the stop and search resulted in action, such as arrest, caution or a fine, or if nothing was found and no action was taken. A reasonable hypothesis might be that any unfairly targeted group would have a lower action rate than other groups [note 7].
This data suggests that the success rate of TVP in correctly identifying potential criminals via stop-and-search is about 30%-40% irrespective of the ethnicity of the person being stopped.

Other parameters

TVP also offers stop-and-search data broken down by gender and age cohort; this is a useful comparison, because TVP may not be targeting people to stop just by ethnicity, but by other variables as well.

On the assumption that there are about the same number of males as females in the TVP region every year, this suggests that in 2017/18 males were about 11 times more likely to be stopped-and-searched than females [note 8].

We can estimate the chances of being stopped and searched in Thames Valley, broken down by age cohort [note 9], by using census data which tells us how many people there were in each age cohort in the TVP region in 2011 [note 10] .

Period 0-9 10-17 18-24 25+
September 2015 – August 2016 0% 0.85% 2.13% 0.30%
September 2016 – August 2017 0% 0.81% 1.57% 0.22%
September 2017 – August 2018 0% 0.55% 1.08% 0.15%

In other words, you were (approximately) seven times more likely to be stopped and searched by Thames Valley Police in 2017/18 if you were aged 18-24 than if you were 25 or over [note 11] .

Conclusion

TVP’s own data bear out Councillor James’ assertion that black people are about nine times more likely to be stopped and searched in the TVP region than white people (but not “any other group”), and the same data also suggests that the proportion of ‘innocents’ stopped by the police is much the same irrespective of ethnicity.

The same data suggest that residents of the TVP region are 11 times more likely to be stopped by the police if male rather than female, and seven times more likely if aged 18-24 rather than 25 or more.

Approximately 93% of the stops of black people were of males (this figure is 94% for Asian people and  87% for white people) and 40% of stops of black people were in the 18-24 age range (46% for Asian people and 33% for white people). Since both these parameters are higher for ethnic minorities,  TVP’s disproportionate stopping of gender and age groups may be contributing to the disproportionate ethnic minority stop numbers.

TVP chief constable Francis Habgood had said in the policy committee that police stop-and-search was intelligence-led (that is, not random), and TVP stop and search data suggests the police are identifying stop candidates based on gender even more strongly than on ethnicity. There may be other variables TVP use to identify potential stoppees that are not included in the data, such as time of day, proximity to crime hotspots, clothing, manner, criminal information and so-on.

Councillor James said:

It’s always negative when groups are disproportionately targeted. The situation with People of Colour and stop-and-search is not a surprise to any people identifying as African-Caribbean or Asian and faced with these challenges on a daily basis. With most stop-and-search relating to drugs, there is significant evidence that black people are less likely to be carrying drugs than their white counterparts, despite being more likely to be stopped, and black people receive harsher police responses and longer sentences.

There is an over-representation of men being stopped-and-searched. It’s a challenging experience for anyone subjected to it. There is some evidence to support a higher number of men being stopped-and-searched, such as the fact that men are far more likely to be the perpetrators of violent crime.

Importantly, Reading Borough Council have a really positive relationship with Thames Valley Police and work well with local officers to build links within different communities. There’s lots of work to be done to facilitate views from the community.

Notes


[1] The traditional borders of Berkshire, Oxfordshire and Buckinghamshire (including the unitary authorities now carved out of the traditional counties).

[2] Ethnic mix in the TVP region (source: dataset KS201EW, ONS & NOMIS, 2011)

Ethnic group Population
White 1,919,646
Asian 209,324
Black 69,013
Mixed 55,986
Other 15,803
Total 2,269,772

Caveat: this is the latest available census data, from 2011.  The population of the TVP region will have changed since this time.

[3] Annual stop-and-searches carried out by TVP grouped by ethnicity (source: UK Home Office via police.uk, 2015-2018)

Period Undefined Asian Black Mixed Other White Everyone
September 2015 – August 2016 210 1525 1747 2 27 7399 10910
September 2016 – August 2017 109 1223 1645 0 35 5522 8534
September 2017 – August 2018 132 909 1166 6 317 3367 5897

Caveat: there are two ethnicity measures in police stop data, officer-defined ethnicity and self-defined ethnicity. The latter value is much more varied than the former, and contains more ‘blanks’ (that is, people declining to offer any ethnicity), so officer-defined ethnicity has been used in this analysis.

[4] Annual stop-and-searches carried out by TVP, expressed as a percentage of all stop-and-searches, grouped by ethnicity source: UK Home Office via police.uk, 2015-2018).

Period Undefined Asian Black Mixed Other White
September 2015 – August 2016 1.92% 13.98% 16.01% 0.02% 0.25% 67.82%
September 2016 – August 2017 1.28% 14.33% 19.28% 0% 0.41% 64.71%
September 2017 – August 2018 2.24% 15.41% 19.77% 0.10% 5.38% 57.10%


[5] Annual stop-and-searches carried out by TVP, expressed as a percentage of ethnic populations in the TVP region (source: UK Home Office via police.uk, 2015-2018).

Period Asian Black Mixed Other White Everyone
September 2015 – August 2016 0.73% 2.53% 0% 0.17% 0.39% 0.48%
September 2016 – August 2017 0.58% 2.38% 0% 0.22% 0.29% 0.38%
September 2017 – August 2018 0.43% 1.69% 0.01% 2.01% 0.18% 0.26%

For example, in 2017/18 there were 1166 police stops of black people, and (an estimated) 69013 ethnically black residents of the TVP region; this is a ratio of 1.69%.

Caveats:

  1. the 2011 census population data is unlikely to be accurate for 2015-18,
  2. not all police stops relate to TVP region residents,
  3. individuals may be stopped more than once ( ie two ‘stop’ events might be the same person, twice),
  4. the 2011 census definition of ‘other’ may be different to the TVP 2017/18 definition of ‘other’, which could contribute to the sudden increase in ‘other’ stop rates in the latest year.


[6] For example, if (in 2017/18) a white person has 0.18% of being stopped and a black person 1.69%, then a black person is 1.69 / 0.18 = 9.39 times more likely to be stopped than a white person.

[7] Percentage of stops in the TVP region resulting in action, grouped by ethnic group (source: UK Home Office via police.uk, 2015-2018).

Period Undefined Asian Black Mixed Other White Everyone
September 2015 – August 2016 28% 31% 34% 50% 48% 33% 33%
September 2016 – August 2017 39% 39% 37% 0% 23% 38% 38%
September 2017 – August 2018 22% 32% 34% 50% 4% 36% 33%


[8] Annual stop-and-searches carried out by TVP, expressed as a percentage of all stop-and-searches, grouped by gender (source: UK Home Office via police.uk, 2015-2018).

Period Female Male Other
September 2015 – August 2016 10% 89% 1%
September 2016 – August 2017 9% 89% 2%
September 2017 – August 2018 8% 88% 4%


[9] Annual stop-and-searches carried out by TVP, expressed as a percentage of all stop-and-searches, grouped by age cohort (source: UK Home Office via police.uk, 2015-2018).

Period 0-9 10-17 18-24 25-34 35+ Unknown
September 2015 – August 2016 0% 18% 38% 25% 17% 2%
September 2016 – August 2017 0% 22% 36% 22% 17% 3%
September 2017 – August 2018 0% 21% 36% 22% 17% 4%

There are single-digit numbers of stop-and-searches for the 0-9 age group each year, which round down to 0%.

[10] Age cohorts mix in the TVP region (source dataset KS102UK via ONS & NOMIS, 2011)

Age cohort Population size
0-9 290,883
10-17 225,057
18-24 194,677
25+ 1,559,155
Total 2,269,772

Caveats:

  1. the latest census data (2011) doesn’t break down by the 25-34 and 35+ groups separately,
  2. the age mix has probably changed in the Thames Valley since 2011.


[11] ie 1.08 / 0.15 = 7.2


Links
  1. The Police and Crime Commissioner has been criticised by councillor Sophia James for playing down race relations
  2. Thames Valley Police and Crime Commissioner Anthony Stansfeld
  3. Stop-and-search at TVP
  4. Katesgrove Councillor Sophia James
  5. ONS 2011 census data: KS201EW (ethnic group) via NOMIS
  6. ONS 2011 census data: KS102UK (age group) via NOMIS
  7. Thames Valley police stop and search data
  8. Stop-and-search: ethnicity facts and figures
  9. Policy committee meeting 29 October, papers and webcast
Advertisements

Have your say

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.