Flashing blue lights atop a police car during a stop and search operation in a UK city.

A black 14 year-old boy in South London has reportedly been stopped and searched by police 30 times over two years.

It is an illustration of racial bias in police stop-and-search tactics. First introduced in England and Wales as part of the Police and Criminal Evidence Act in 1984, officers are able to stop and search anybody if they have “reasonable cause to suspect” they are breaking the law, while in the US ‘stop-and-frisk’ has been around in one form or another since 1968.

In England, Black and Asian people make up 11 per cent of the population, yet they account for 30 per cent of all stop-and-searches. There are numerous studies showing the racial bias of stop-and-search.

So why exactly is stop-and-search so racially biased? Is it down to individual officers or is it because more officers are deployed to more ethnically diverse areas which suffer socio-economic depravation and higher crime rates?

We looked to answer this question by analysing data shared by West Midlands Police, one of the largest forces in the UK. The data set covered 1,194 officers in 29 teams, 203,176 reported crimes and 36,028 stop-and-searches from April 2014 to September 2018. The research was funded by West Midlands Police and Crime Commissioner and by Warwick University.

The data provide great detail on who the officer was and the exact location and time of the stop-and-search plus information on all the crimes, incidents and suspects across the West Midlands.

We combined this with population data from the 2011 census to work out the ethnic composition of the areas where officers were deployed.

Using data science statistical modelling we were able calculate whether there was an over-searching of black and Asians by officers compared to the population and crime rates in the area, and whether there was a bias in the deployment of police to certain locations. So we could compare a police officer’s stop-and-searches of an ethnic group with their actual experience of crime involving this group. This can help us understand if the racial bias is down to the officer or the fact they are being sent to areas with more black and Asian people.

Replicating previous findings, we found that black and Asian people are over searched. Across the West Midlands, Asian people make up 25.1% of the searches but only 19.8% of the population, which means they are over searched by a factor of 1.26. For black people, the factor is larger, at 2.79. We see over searching against both the ethnic composition of the population officers patrol and against officers’ own experience of crime suspects.

However, individual officers choosing to over searching black and Asian people they encounter is not the sole cause of racial bias in stop and search. Deployment decisions also play a role.

Police officers are, on the whole, sent to more ethnically diverse areas in the West Midlands. Even if officers stopped people blind to their ethnicity, if more officers are sent to more ethnically diverse areas, stop and search will be racially biased. We find areas with high numbers of black and Asian people are over-patrolled and so contribute to the racially biased stop-and-search. The average officer patrols an area that has 1.16 times more Asian people and 1.37 times more black people than live in the West Midlands.

How the headline search factors break down into the decisions of individual officers differs vs the deployment of officers into ethnically diverse regions differs for Asian and black people. For Asian people, the headline factor of 1.26 is due mostly to officer deployment, with a smaller contribution coming from the finding that some officers over search Asian people. However, for black people, while over patrolling plays a role in the headline factor of 2.79, the larger component comes from almost all officers over searching black people.

Interestingly, we also discovered that the make-up of a team of police officers has an effect on stop-and-search biases. Teams that have more white officers stop-and-search fewer black and Asians. There are virtually no black police officers, but teams with more Asian officers actually search more black people.

This could be caused by ‘stereotype threat’, where teams of white officers might feel they will be stereotyped as racist if they stop and search ethnic minorities.

For policymakers and police chiefs, it is clear they must overhaul their deployment strategy and take into account the bias they currently suffer from. The ethnic composition of teams of police officers should also be considered as this seems to have an effect on the biased nature of stop-and-searches.

Addressing the norms and biases of police officers through training must also be considered if police forces are to reverse the racial bias of stop and searches.

The 14 year-old boy from South London was so distressed by his experience that he was afraid to go out, not because of high levels of crime in the area, but because of the police.

Proper training of officers on their racial bias and a review of deployment polices is urgently needed if more youngsters are not to grow up seeing the police as the enemy.