Thanks for catching that right away, which is exactly what I want in putting ideas out publicly: To correct the record. I continue to ask anyone to PLEASE let me know if there’s a statistical or methodological flaw with my spreadsheet here or here. Or in this post.
For example, I found an error on my tab on “violent crime” because the FBI presents their tables in goofy ways, so I spent part of the day correcting it on this story and elsewhere. The black-white percentage difference is still huge at 48%, but violent crimes are only 4.1% of all arrests for white and 6.7% of all arrests for blacks. I had those higher.
In the specific statistical language error you mention Douglas Eckberg, it’s always confusing. Especially when one can say, for example, the following for police stops in Chicago in 2016 (the same year homicides increased 57%, from around 500 to 800) and they all mean the same thing:
- 80% less stops (Chicago in 2016)
- 6 times fewer stops
- And if going in the other direction one could say “stops increased by 600%” or are “600% higher” (Thus, I used “600% the incorrect way… um, this way, what way? Yes, now that correct way!)
Not to use it as an excuse: I’m obviously juggling a lot of data and I wanted to publish this the symbolic last day for consent decree input and that was one of my last edited sections. But that’s not an excuse.
What I was doing was re-purposing some older verbiage from my magnum opus post on Chicago policing and bad outcomes from well-intentioned but heavily misguided government interference and mixing it up in this post I finished on Friday.
OLD SENTENCE:
“Another fact I discovered — and was not reported — looking at easy-to-find data: In 2016, the arrest rate in Chicago was lower per capita than the national rate by 12%. Of course, that year police made 600% less stops and made nearly one-third less arrests. Some did report that. They just didn’t show the bigger picture.”
NEW SENTENCE (of course this post is getting longer):
Another fact I discovered — and was not reported — looking at easy-to-find data: In 2016, the arrest rate in Chicago was lower per capita than the national rate by 12%. Of course, that year police made 6 times less stops — around 600,000 stops compared to 100,000, which some analytically argued created “The ACLU Effect” of horrendous bloodshed. This, of course, led to nearly one-third less arrests. Some did report that. The ACLU, academics, cultural influencers, and political leaders cheered it on. They just didn’t show the bigger picture. They ignored it until it was too late. And no one, not even former Police Superintendent Garry McCarthy on 60 Minutes, said, “Last year, you were less likely to be arrested in Chicago than anywhere else — now that’s surprising!” What McCarthy did say was important, if not obvious: “When you have activity falling off the way it is and crime skyrocketing, that’s a huge problem.”
ANOTHER SECTION NEEDING A FIX:
“The ACLU seemed pleased with the 600% reduction in stops [in Chicago] — ignoring the possibly record-high 4000-plus shootings that were 95% non-white victims — but still had concerns about those unmoving disparities. Just like in New York City (NYC) when stop-and-frisk ended and demographic percentages were calculated by the ACLU from 2003 to 2017, there’s no change, essentially, in the percentage of white, Latino or black stops despite nearly 60 times (6000%) less stops.
NEW SENTENCE:
“The ACLU seemed pleased with the 80% reduction in stops — ignoring the possibly record-high 4000-plus shootings that were 95% non-white victims — but still had concerns about those disparities that wouldn’t budge. Just like in New York City (NYC) when stop-and-frisk ended and demographic percentages were calculated by the ACLU from 2003 to 2017, there’s no change, essentially, in the percentage of white, Latino or black stops despite nearly 60 times less stops.”
My Second Fix (Always Go Back and Double Check)
So I fixed the place where I had also misused that data drop percentage point by saying there was a “nearly” a 6000% drop. I said “nearly” because here I flubbed using the more specific “5500%” drop. It is changed now to the correct “55 times less stops” in New York City in 2016 compared to 2011.
Yeah, do people know that? Apparently, the NYPD now stops 55 times less people. Just like the CPD stopped 8 times less people in just one year, most ostensibly for discretionary reasons.
Click the launch window below and go to this section: “Far Less Stops, No Change in Demographics — How Is Stop-and-Frisk Racial Profiling Again, ACLU?” Under that data-heavy chunk of text, you’ll find these 3 highlights I copied & pasted. They make an interesting point, I think, about how these consent decrees really do make change. They make police do a lot less, and possibly subvert the record-keeping process. Of course, what doesn’t change are those nasty racial disparities — they’re largely the same — that occur almost 60 times less, at least on paper.
60 times! Some things just make ZERO sense.
OK, from last year (but I keep editing it up until today):
UPDATED PARAGRAPH:
Here’s what I mostly think: The police stepped back a great deal, but were they really stopping and interacting with 5500% more New Yorkers just a few years ago? I don’t think so. I think an activist judge’s decision and top-down pressure from the mayor is forcing them to stop recording “stop data.” How else could they have 37,568 “Dangerous Drugs Misdemeanor” arrests and 14,674 “Dangerous Drugs Felony” arrests in 2016, the same year they had 12,404 stops? I’m fairly sure all arrests didn’t go down 55 times since 2011. It doesn’t make sense. And I’ve yet to hear the ACLU, The New York Times, or anyone else call “bullshit” on that.
ELSEWHERE (yes, I re-purposed some of this):
Besides the massive shooting increases, you can have an 80% drop in stops (6 times less), and another shocking thing happens (or doesn’t happen): The demographic percentages won’t change at all. As a Chicagoan who’s heard for years about “over-policing” in minority neighborhoods, I found this incredibly surprising. If “racial profiling” was truly occurring, the demographics would logically shift dramatically after dramatic reductions in stops. This is because the widespread civil rights theory supposes stops for black and brown citizens are more arbitrary, but not as much for white citizens. The reasoning being non-Hispanic whites are more likely to live in lower-crime, more economically advantaged neighborhoods in cities and are less likely to be stopped and frisked in general — the police simply aren’t there and aren’t looking for them.
AND ALSO:
In NYC in data captured by the ACLU, whites were 12% of stops in 2003 at the beginning of stop-and-frisk and they were 10% in 2016 when it was all but over. Remarkably, there were 55 times more stops of all races in 2011 than 2016! (A class action lawsuit ended the practice in 2013, and a slowdown occurred before then.) But what’s most important, and rarely reported, is that crime suspects nearly match the demographics of police stops in NYC. Let’s look at police stop data (not all of these involve “frisks”) and victim report data to see: