Stop and search: are black people over-represented?

Readers of the Guardian, Daily Mail and Telegraph this morning were presented with more headlines implying police bias when using stop and search powers. The Guardian and the Mail, however, seemed to take slightly differing perspectives on the matter:

The Guardian: "Police up to 28 times more likely to stop and search black people — study"

Daily Mail: "Police are '37 times more likely to stop and search black people'."

Matters are confused further when we compare this to what the Guardian were saying in January:

The Guardian (January): "According to official statistics, in 1999-2000, a black person was five times more likely than a white person to be stopped by police. A decade later, they were seven times more likely."

So how are the public to make sense of these differing headlines?

Regular readers will know Full Fact has looked into this issue before. In January when the Guardian first reported the story, we pointed out that the main reason for differing reports was that there are a number of different laws under which a person can be stopped and searched by police.

Data is collected by the Ministry of Justice on the ethnicity of people stopped and searched by police under three different Acts. They are:

1. Section 1 of the Police and Criminal Evidence Act (1984)

2. Section 60 of the Criminal Justice and Public Order Act (1994)

3. Section 4 of the Terrorism Act (2000)

As we've pointed out before, the 1984 Act allows a search to be conducted when an officer has reasonable grounds for suspicion that a person or vehicle may be carrying stolen or prohibited items.

The 1994 Act extends this to when there is a percieved threat of serious violence. The 2000 Act extends this still further to the right to search for items that could be used in connection with terrorist acts.

Today's statistics presented in the Guardan, Mail and Telegraph concern only those stops and searches made under section 60 of the Criminal Justice and Public Order Act, following the publication of a study on the Act by the Equality and Human Rights Commission (EHRC).

The study combines data acquired from 40 police forces across England with experimental population statistics from the Office for National Statistics (ONS) for mid-2009.

Based on the responses they received, they were able to calculate the ratio of stop and search procedures against people of white ethnicity compared to people of black, Asian and mixed ethnicity. They present their findings as follows:

The most recent figures available to the EHRC confirm the Daily Mail's headline that police are 37 times more likely to stop a black person than a white person in England.

The Guardian's "up to 28 times" figure takes the West Midlands specifically as an example of how large the ratio can be at an individual force level, based upon the six forces that were specifically mentioned in the study.

However, with the results of around 35 police forces missing from the table, it isn't clear that this is indeed the worst case across the country.

But aren't we missing the bigger picture?

To be fair, all of today's reports and the EHRC study make clear that their statistics only concern stops under Section 60 of the 1994 Act. This is perfectly reasonable since this Act, as we found in January, is the most extreme case of a disproportionate white-black stop ratio.

However, when the other Acts under which a person can be stopped and searched are included, the ethnic disparities are more modest. Consider the Ministry of Justice's own figures on the number of people actually stopped and searched in 2009 under each Act:

While the figures for Section 60 stops shown here are out of date compared to the EHRC's figures, it is worth bearing in mind that the vast majority of stops and searches actually occur under Section 1 of the 1984 Act - where the white-black ratio divergence is much less pronounced.

What the EHRC report does highlight is the severity of disproportionality under Section 60 specifically. So while this morning's headlines are accurate when understood in this limited sense, we should be careful to consider all stop and searches conducted by police, a point which perhaps isn't fully communicated in the headlines themselves.

UPDATE (14 June 2012)

The Equality and Human Rights Commission have provided Full Fact with more details that were not published in the initial EHRC report.

The Guardian newspaper claimed that black people were "up to 28 times more likely" to be stopped and searched under Section 60, using the example of the West Midlands to illustrate this. However it is now clear that this isn't strictly correct.

The EHRC confirmed that in the case of Gloucestershire, for instance, the black-white disproportionality in stops and searches was 177-1, and in Avon and Somerset it was 164-1. They emphasised that this was due to the small number of stops that took place in these areas, combined with their predominantly white populations.

Of course had the Guardian chosen to report "up to 177 times more likely" this would doubtless exaggerate the picture based on a region carrying unusual demographics, so this is not to say that these examples would have been any better to report.

However given that the Guardian's headline isn't strictly accurate either, it may have been better for the newspaper to headline - as the Mail chose to - that the national ratio of black-white stops and searches is 37-1.