An ethnic minority man aged 25 to 34 is twice as likely to be fined under the coronavirus guidelines as as a white British man of the same age, a new report by the National Police Chiefs’ Council (NPCC) has found. British Asians are the most disproportionately represented among the fined, followed by black British people, people from mixed ethnic backgrounds and the helpfully specific “other”. All are more likely to be fined than white British people.
We have a pretty good idea that men, regardless of ethnicity, are the most likely group to breach the coronavirus guidelines (or, at least, the group mostly likely to admit to being likely to breach the guidelines), so, all things being equal, we would expect men to be more likely to be fined than any other group. But there is no data-led case to expect disproportionality on ethnic lines. There is reason to be troubled by these findings, as the NPCC rightly concludes. The evidence is more complex than the headlines suggest, however. What this means is that some police forces are being painted as worse than they are, while others are getting away with real and serious abuses of power.
In some instances, the actual numbers involved are tiny. Take Cumbria, for instance. On a headline measure it is one of the worst-performing police forces, with black people considerably more likely than white people to be issued a fixed penalty notice per head of population. But they fined just three black British people.
There are a variety of ways that you can adjust to take account of that. The NPCC has run the numbers in two ways: by measuring the disproportionality in terms of the issuing police force itself, and by measuring the disproportionality in terms of the place of residence of those fined.
The advantage of doing the former is it gives you a measure of the decisions taken by individual police forces. The disadvantage is that many of the police forces issuing fines are in tourist hotspots, whose own local ethnic mix may not accurately reflect the ethnic and social mix of those breaching coronavirus laws and guidelines in an area. As the NPCC argues, neither measure is perfect.
But the differences in results between the two are large. Cumbria, North Yorkshire and Lancashire are the three worst-performing police forces when the disproportionality is measured by individual forces; in each there is a significant reduction in disproportionality when it is measured by the place of residence of people fined. North Yorkshire goes from being the second-most disproportionate issuer of fines to ethnic minority Brits to one of the least, with no statistical significance between fines issued to white people and ethnic minority Brits. London goes from having a small disparity between the treatment of ethnic minority Britons and white Brits to having no difference at all. There are some police forces that perform well on both metrics: police in London, Merseyside, Cheshire and Northumbria are among the least likely forces to disproportionately fine ethnic minorities however you decide to count. If you weight the number of fines not by the police force issuing them, but the place where the people being fined live, almost all police forces emerge with a smaller level of disproportionality between ethnic minorities and white Brits – though there remains a significant difference either way overall.
See also: Gary Younge on Covid-19’s disproportionate impact on BAME people
The NPCC argues, perfectly reasonably, that both metrics are imperfect. However, I think that we can make a persuasive case that measuring disproportionality by the place of residence of people fined is the better metric, using another topic that is frequently and rightly cited as an example of disproportionate and unfair policing: stop and search.
One under-appreciated legacy of Theresa May’s attempts to reform the police when she was home secretary is the greater use of comparisons between police forces by the Home Office. These comparisons are not always well-used and her successors haven’t been inclined to use them to drive improvements, but they do provide a good guide to which police forces are operating well and which aren’t. There are a host of comparisons that the Home Office now helpfully collects and measures. For our purposes, however, we’ll confine ourselves to stop and search and the subsequent arrest rate.
While it is not a perfect indicator, a police force’s arrest rate is a pretty good guide to how effectively they are using stop and search, and to what extent they are merely engaging in profiling, whether on race or class lines. If a police force is only arresting one person in every ten that is stopped and searched, they are visibly failing to distinguish between the innocent and the potentially criminal with the same level of effectiveness as a force that is arresting three or four people in ten. I’m not saying that three or four in ten is the best we can hope for, but I am saying that the police forces that are consistently doing this are the ones ahead of the curve and whose leadership should be a model for police forces across the UK.
Since these figures began being collated, the same police forces are consistently top of the pack as far as their arrest rates are concerned: Cleveland, Durham, Northumbria and the City of London. They consistently have higher arrest rates from stop and search than the national average and than other police forces. And the same forces are consistently underperforming the national average. The Gwent police force, for example, has arrested fewer than one in ten people following a stop and search consistently since 2016.
When you look at the police forces with the lowest amount of disproportionality in their use of fines, using the NPCC’s second method of measure, they are the same forces that get the best use out of stop-and-search. Gwent’s remains almost unchanged, falling from a disparity rate of 3.4 to 3.3, and in general, the police forces with the lowest arrest rates from stop-and-search (that is to say, the ones who are more likely to search the wrong people) are the ones most likely to have fined a disproportionately high number of ethnic minority people. This suggests that one issue at work is a disproportionate focus on one part of the community.
The criminologist Richard Berk describes the debate over criminal justice as a problem of Luke Skywalkers and Darth Vaders. Are you more concerned about building a system that maximises the chances of correctly identifying who is a Luke Skywalker – that is, someone who is either innocent or has been rehabilitated in prison? Or are you more concerned about building a system that maximises the chances of correctly identifying Darth Vaders – that is, people who have committed a crime, or have already done so and will do again? Whichever you prioritise, you will let some guilty people go free, and some innocent people will be unfairly punished. But wherever you come down on that issue, the Vader/Skywalker question is also a good yardstick of police effectiveness. When police are stopping and searching members of the public, they shouldn’t have to search more than nine Luke Skywalkers for every one Darth Vader they correctly identify.
Something is going wrong. We should read the NPCC report alongside the information we already have about which police forces are getting the balance right, or at least heading in the right direction, in order to work out what.
See also: Anoosh Chakelian and Michael Goodier on the stats and stories of stop and search