I’ve become interested in a rather ghoulish question, viz:
If a cop started beating someone to death in the presence of other cops, what percentage of the cops watching on would either just let it happen, or even join in?
I’m serious. I want to know the percentage, and I think we can have a reasonable go at estimating it, though the level of statistical acumen required is beyond me.
Tyre Nichols
There were five officers, apparently all from the SCORPION street crime unit who, to various degrees, murdered Tyre Nichols(1). Afterwards, the city insisted that the name of SCORPION had been sullied by the actions of a few, which were not representative of a majority:
In general, after an event like this, people will claim that a majority of officers in a unit are good people, there are just, unfortunately, a few bad apples.
If this theory is true, we would expect a majority of members of the SCORPION unit would have tried to prevent the murder had they been present, and certainly not participated in it.
SCORPION had 30 officers,
5 of them were present.
Assume the most modest and therefore statistically plausible case for the “actions of a minority” view, viz that 16/30 officers in the SCORPION unit would have tried to prevent the brutality had they been present- that 16 out of 30 were “fresh apples”.
The odds of five randomly selected officers being rotten apples if 16 out of 30 members of the squad as a whole are fresh apples and 14 are rotten apples is:
(14/30)*(13/29)*(12/28)*(11/27)*(10/26)=1.4% or about 1 in 70.
A friend did some Bayesian modelling. Obviously, this stuff is sensitive to priors, but they found that there’s very slightly less than a 95% chance that at least 20/30 cops in the SCORPION squad were bad apples.
Of course, this approach is riddled with limitations. It assumes that the 5 officers present were a random reflection of the SCORPION unit in general. (Edit: added emphasis) Also, even if this is true, since the SCORPION unit may not be reflective of the police force as a whole, it can’t tell us all that much about the Memphis police force, let alone the United States police in general.
Still, we could repeat similar modelling with George Floyd, Eric Garner etc. In my view, a pattern starts to emerge.
This a challenge for an econometrician looking to make a name for themselves.
Looking to make a splash and not afraid to get your hands dirty with some bleak modelling? I think that a skilled econometrician, through the use of logistic regression, could probably come up with a reasonable estimate of what percentage of American police would stand by and watch as someone was killed (or, for that matter join in). If you are such an econometrician, I think you should try it.
Why though? Why set this impossibly ghoulish task? Is this just leering at American necrosis? I don’t think so.
There’s a lot that could be said about policing in the United States- many police seem to hold the insane view that it is never incumbent upon them to take any risk with regard to being shot- that if there is even the slightest risk that someone is reaching for a gun they are entitled to shoot first. All too often the law backs them up. And cowardice, rather than aggression, does seem to be the dominant motive- consider Uvalde.
There’s a debate about whether the police are reformable. Some hold we can make the police better, some hold that the whole institution needs to be ripped out root and branch and replaced, some hold that the whole institution needs to be ripped out root and branch and not replaced, some hold that the institution is essentially broken and not much can be done to fix it, but it would be too dangerous to rip it out.
If we suppose, for a moment, that at least some of the people engaged in this debate are engaging in good faith, then I think that a condition of the debate going anywhere is for us to have a sense of just how rotten the institution is. What better figure to sum that up than the percentage of police who would, or wouldn’t, intervene to stop brutality- or for that matter, the percentage who would join in?
I think I already know roughly what the results would be if the estimate were done well. I think that a lot of people, even liberals, know by now in their hearts of hearts that most cops would, at best, just stand around while they were killed, so long as their attacker was a police officer, and had at least a thin pretext for their actions (that last part may not be necessary).
But I could be wrong. Maybe 80% would intervene. Or maybe I’m right, and the vast majority wouldn’t, at least if the officer-assailant had even a fig leaf of legitimacy. But if liberals and conservatives who say that just a few tweaks, nudges and leadership overhauls are enough are being earnest, this is exactly these are the sort of questions they need to honestly face:
Do you think most police would do the right thing in a situation of police brutality, and if not, how can it be true to say the police only need a touch-up and a paint job?
Alas, I think for some, the idea that the police are fundamentally good is not a claim that refers to empirical reality. It’s a ground condition of their understanding of political reality, something they must believe, in order to understand in their favoured terms.
We deserve guardians who won’t become our murderers or their accomplices on a dime. If our society prevents that, our society must change.
Footnote:
(1): A sixth officer, present at the first scene but not the second, appears not to have been part of SCORPION? I am unsure.
I think you raise good points here. We have unclear statistics on the issue and that leads to unclear ideas on what to do about it.
I want to offer one possible confounder when taking the raw numbers of 5 out of 30 officers as a random sample: like-minded coworkers are more likely to be working together than those that aren't. So if Bad Apple #1 was present, it's likely to find Bad Apples #2-5 present too. That's an optimistic reasoning for how many were there, but I think it's worth considering.
I think some would say that people who would stand aside and not intervene in current conditions can be conditioned to intervene - that this is not some overriding principle that they never intervene but some thin balance of incentives that can easily be shifted in most. (This is obviously related to the debate about how many Russians support Putin's politics and how many of them would stop doing this if they weren't literally programmed to by television and similar sources.)