Thanks for the hand-claps and accolades. And it’s funny you bring up Heather Mac Donald, as I came to her conclusions as a liberal, but literally have found myself making the same arguments in a vacuum (i.e. I wasn’t influenced, but our arguments overlap). I read the same story in the City Journal with Mac Donald using updated 2016 FBI statistics, which just came out to show an increase slightly less than 2015 for homicides (8.6% rather than 10.8%, combining for more than a 20% increase in the past two years). I need to update my analysis, because I’ve predicted nearly 3000 more homicides in 2016 than in 2014, and that’s exactly where we’re at. (FBI #s are sometimes smaller than “actual” because they’re working with data provided; but CDC and other organizations give larger estimates.)
The one article I would use to share with others is this one, as I’ve regularly updated it over the year — many of the updates directly refute the naysayers, like “You can’t trust the numbers.” If you can think of a place for the work, please send me direct recommendations.
How Roland Fryer’s Controversial Harvard Study on Racial Bias by Police Actually Shows Negligible…
I’ll crunch the numbers. You can comment and debunk.
And to critics who say, “Well, this is data provided by ‘honest’ police departments with nothing to hide (wink wink). How can we trust it?” The problematic police departments might not cooperate like Houston did, for example, or some police may be “cooking the books” on what they document so it doesn’t look like they’re pulling over as many brown and black people. Besides, arrest data isn’t necessarily accurate, statistics are limited and don’t reveal the whole picture, and you need qualitative ethnographic studies to truly understand what’s happening on America’s streets. PBS investigations like Frontline’s Policing the Police have full access, cooperation, and the officer-on-the-street’s POV but still show “bad stops.” Even Fryer suggests it, saying in the analysis, “It is possible that these departments only supplied the data because they are either enlightened or were not concerned about what the analysis would reveal. In essence, this is equivalent to analyzing labor market discrimination on a set of firms willing to supply a researcher with their Human Resources data!” He had indicated a desire to work with Chicago crime data, for instance.
OK, I hear that. Well, you can’t hide dead or gunned down bodies, right? Compared to killings, publically accessible use-of-force data indicates non-lethal interventions happen slightly more. By looking at both, any rational observer would say, “OK, these numbers are close and they should be.” If non-lethal use of force happened slightly less in the official data than killings by police, you might determine something fishy is going on systemically. Ditto if use-of-force instances were much lower than the percentage of arrests. It would be completely rational then to assume police records can’t be trusted. However, that is not the case. These are the percentages of the national total for Black Americans (13% of the population):
27% of all arrests (51% for murder; 32% for aggravated assault)
31% of all instances of use of force by police
25% of all lethal instances of use of force by police
43% of the persons killing police
Bottom line: If lethal and non-lethal use of force were 40% or higher and arrests remained at 27%, I wouldn’t be writing this analysis. I’d say, “See, there’s your systemic racism.” This isn’t high-level statistics, but it’s common sense.
If someone has an argument that X amount of interactions with police should be anything other than X amount of instances of use of force, I’ve yet to hear that as a compelling argument. They should be similar, like 27% and 31% above, with the 4% gap coming from the types of arrests (black arrests are higher in the more violent categories), police behavior (bias), and citizen behavior (cooperation and complying with orders).