This sounds appalling. And we (are supposed to) assume it proves institutional racism, right?

But what if the stats actually suggest the *opposite*?

Sounds crazy but stick with me for a sec... /1 https://twitter.com/BCWLindsay/status/1281850831088820226

The article presents these numbers as if an open and shut case, but doesn’t ever make a case - just insinuation. Got me wondering about those 2 key stats. I’m no expert but even at first glance they raise more questions than they answer... /2

(1) What good is the ‘equivalent of’ measure?

The one thing we can be sure of is it’s *not* the actual % (which must be lower; how much lower would be genuinely interesting to know but maybe data unavailable). But aren’t repeats massively relevant here? /3

If there’s a specific sub-set of a group who have (for whatever reason) multiple interactions with the police - by definition that isn’t representative of the whole group.

Isn’t ‘the equivalent of’ just 'over-generalisation bias' by another name? /4

[Sidebar: Interesting that the article acknowledges this but does so by saying some may have been "targeted" more than once. Jedi mind trick to make me think bad policing rather than bad statistics? Smooth] /5

PS assuming the thing you’re trying to prove used to be considered a logical fallacy

Anyway now to the main one...

(2) 80% => ‘no further action’ - how should we feel about this? /6

Implies the vast majority of stops were for no good reason. So ... must be racial profiling / discrimination? Case closed, your honour.

But hang on a sec. /7

Surely we need at the very least to know the equivalent stat for *other ethnic groups* to compare?

But the article didn’t seem to mention that. Hmmm /8

Thankfully turns out the @metpoliceuk have a searchable online public database of all stop and searches. /9

https://www.met.police.uk/sd/stats-and-data/met/stop-and-search-dashboard/

So 2 clicks later I found this. May 2020.

Look at the different ethnic groups ... what do you notice?

NB ‘positive outcomes’ = where further action *was* required, so the inverse of the 80% stat /10

[Asians & ‘Other’ might kick off at this point and say “See - we’re being discriminated against worst of all!” Is anyone saying that? No] /11

Moreover, after another 2 clicks, I learnt the Met’s TARGET for ‘positive outcomes’ is … wait for it

… 20%.

Let that sink in

The very stat (80% 'negative outcomes') that we’re supposed to be outraged by suggests the police are meeting their official target. /12

[Full disclosure: I know basically nada about policing theories or whether it’s good target but I assume it’s not inherently outrageous.]

And, crucially, what we're *not* told in the article is there seems to be no real ethnic disparity here. /13

[Sidebar 2: Of course the great @ThomasSowell taught us that disparity itself does not = race discrimination. There’s loads of other factors that could credibly account for a disparity.]

But here it seems there’s *not even* a meaningful disparity to speak of with the 80% /14

So this stat *could* be used to commend police for meeting target & using s&s impartially. Rebuild public confidence. Reassure law-abiding young black men that the police aren’t out to get them. Improve race relations in a united fight against crime & anti-social behaviour /15

But somehow that’s not the vibe I’m getting from the article or responses. Why?

I'm not sure.

But could it be we’ve been groomed as a society to be systematically biased...

to accept & reinforce a set narrative about race/policing (& much else) unthinkingly? /16

Could be wrong.

Just cant think of another reason why they’d omit the comparison stats.

=> Systemic, institutional, racial ... confirmation bias? /17

See - if the issue with stop and search etc *isn’t* simply a failure of policing, I suppose we’ll have to consider failures elsewhere, which may be far less appealing politically (and far less simple) /18

Stat 1: young black men disproportionately likely to be stopped

Stat2: these stops are just as likely to lead to arrests / further action as for all other ethnic groups

>> What should we conclude? /18

The one thing the data *doesn’t* support is Mrs May’s fear that individuals are being stopped “simply" because of the colour of their skin.

If that were the case (as maybe it was in the past) - how on earth would the 80% be more or less consistent across all skin colours?! /19

I certainly don’t claim to have the answers. But wondering how do we at least ask the right questions?

Reading the article something didn’t smell right. And if the knee-jerk outrage and confirmation/coverage bias takes hold how will this play out..?

My fear... /20

(1) Politicians jump to virtue signal, condemn police, pursue data-poor politically motivated 'reforms'.

(2) Police censured and/or nervous to use s&s for fear of being called 'racist'. /21

(3) Narrative of racial division and tensions inflamed, esp in communities with worst crime rate.

(4) Crime in these areas likely to get worse. (cf. USA)

>> Now let's think: who's likely to be disproportionately affected and suffer most from all this? /22

In summary: I think we should be suspicious of motivated reasoning / Jedi mind tricks / dodgy stats *whatever our political persuasion*.

We need to talk about race. And policing. And crime. But we need to do so carefully & with responsible data-handling and journalism. /23

Let’s pray and work to overcome, not reinforce, prejudice and division.

/END