Statistics say driving while black is real
September 9, 2014 11:25 PM   Subscribe

The Justice Department statistics, based on the Police-Public Contact Survey, show that "relatively more black drivers (12.8%) than white (9.8%) and Hispanic (10.4%) drivers were pulled over in a traffic stop during their most recent contact with police." Or, to frame it another way: A black driver is about 31 percent more likely to be pulled over than a white driver, or about 23 percent more likely than a Hispanic driver. "Driving while black" is, indeed, a measurable phenomenon.
The Washington Post's Christopher Ingraham looks at the recently released statistics about traffic stops and whether people are pulled over for driving while black.
posted by MartinWisse (27 comments total) 12 users marked this as a favorite
 
Sort of a "no shit, Sherlock" thing, but it's nice to have stats to point at when the apologists come out.
posted by NoraReed at 1:51 AM on September 10, 2014 [10 favorites]


Thank goodness we have these statistics. Because goodness knows you can't listen to us black people when we tell everyone these things.
posted by magstheaxe at 5:15 AM on September 10, 2014 [25 favorites]


Being as they represent a tiny percentage of the population, isn't the overwhelming statistic of Native Americans being the single most likely demographic by far to be pulled over not of more concern? There are around a tenth of the percentage of Native Americans compared to Black people (according to this) then it is hardly the biggest shocker from that stats spread is it?

I mean, I know it's a hot button topic and all that (and still worthy and justified in being raised in profile) but for around 1% of the population to be the overwhelmingly more likely to be pulled over by the police..... Um. Isn't that something that should be being considered (at the very least) equally newsworthy?

posted by Brockles at 5:17 AM on September 10, 2014 [1 favorite]


Yes, in hindsight I should've mentioned that in the post here. I thought about it but decided against it because I was uncertain about the language the author used to describe native Americans while the hook of the article really was Driving While Black, which is one of those things everybody knows is a real thing but can't be "proven" but now you can.

Nevertheless, the even greater harassment of native Americans should've been mentioned.
posted by MartinWisse at 5:58 AM on September 10, 2014


Why are these statistics gathered from surveys? Why do they take a sample of people, contact them, hope to get honest answers and extrapolate from there?

Why aren't there actual detailed records of every stop? The police should be required to make a log every time they pull someone over. Pulling someone over in a vehicle creates a traffic danger to one degree or another which is another reason why the event should be logged and explained.


And why so few details? How many males vs females were pulled over? If there are more than 50% males does that "prove" that there is such a thing while "driving while male". Being sarcastic there but the point is that there are dozens and dozens of other variables in who gets pulled over.

A fully logged and detailed report would seem to give a better picture.

... of course that is assuming that you'll get honest logs from the police officer which is probably too much to ask for
posted by 2manyusernames at 6:04 AM on September 10, 2014 [2 favorites]


Being as they represent a tiny percentage of the population, isn't the overwhelming statistic of Native Americans being the single most likely demographic by far to be pulled over not of more concern?
The percentages in the first chart ("Driving While Black?") don't add up to 100%, though it appears they might at first glance. So unless I'm missing something, I think that chart shows how likely a person of any particular race is to be pulled over in a given year. In other words, Native Americans don't make up 15% of all traffic stops (which would reveal mind-blowing levels of racism since they make up such a small percentage of the population) -- but a Native American was 15% likely to be stopped at some point in 2011, vs. whites being 9% likely, etc.

The Post could have made that a bit more clear, I think.
posted by whitecedar at 6:09 AM on September 10, 2014


I don't doubt for a minute that driving while Black is a thing.

But "last contact with police" seems like an odd denominator to use. And it certainly doesn't follow from "relatively more black drivers (12.8%) than white (9.8%) and Hispanic (10.4%) drivers were pulled over in a traffic stop during their most recent contact with police." that "A black driver is about 31 percent more likely to be pulled over than a white driver, or about 23 percent more likely than a Hispanic driver."

e.g. (a hypothetical example of data in which that stat true and it not true that black people are more likely to be pulled over):

A population with 1000 each non-hispanic white, non-hispanic black, hispanic (eqaul numbers make the math easier). Again for simplicity, each person who has contact with police has contact only once. 500 white people and 500 hispanic people have contact with police (50%). 50 Black people have contact with police (5%). Of those who have contact with police 12.8% of blacks (about 6 people of the thousand we started with) had their contact because they were pulled over. 9.8% of the whites who had their last contact with police because they were pulled over (about 49) and 10.4% of hispanics had their last contct with police because they were pulled over last contact was due to a traffic stop (52). In the end that gives us .6% of Black drives pulled over, 4.9% of whites and 5.2% of hispanics. That would hardly support "driving while Black" is a thing. As presented, these are not data on how likely drivers are to be pulled over.

In short, I think driving while black is a thing and it offends me. I strongly suspect that data that used "all drivers" would show this. However, these data do not show this and I am also offended by poor use of data.

Without knowing what percentage of drivers have contact with police at all, and how many contacts with police they have over a given period of time, you can't go rigorously from the statistic cited at the top of the article to the conclusions drawn.
posted by If only I had a penguin... at 6:12 AM on September 10, 2014 [5 favorites]


Why aren't there actual detailed records of every stop? The police should be required to make a log every time they pull someone over.

They do. (You've been pulled over, right? What do you think the cop does in the patrol car for minutes at a time?) But in order to get data on the drivers' races from records of traffic stops, the cop would have to do one of the following:
  • ask the driver to self-report his or her race
  • document his own (the cop's) perception of the driver's race (this is probably the one that's most salient, but is the cop's testimony reliable?)
  • take a photo of the driver, so that researchers could arrive at their own perception of the driver's race after the fact
It's hard to imagine any justification for doing any of these as part of daily police work.

But, yes—agreed that the methodology is sloppy for other reasons. (For the record, I have no doubt that DWB is a Thing.)
posted by escape from the potato planet at 6:36 AM on September 10, 2014 [1 favorite]


Thank goodness we have these statistics. Because goodness knows you can't listen to us black people when we tell everyone these things.

That's called anecdotal evidence and it's fairly bad policy to be basing anything off what people say happens. I see this in psychology articles all the time. They do a study to prove something that seems obvious to the world, and you get the incredulous folk explaining something like "They spent money on that? Everyone knows teenage girls are just and horny as the boys!" Pretty much any science has these sorts of studies where if you look at what they are proving you think, "Everyone knows that," but sometimes the study doesn't end up supporting the prevailing wisdom and then we all get to learn something new!
posted by cjorgensen at 6:37 AM on September 10, 2014 [11 favorites]


Upon further reflection (i.e., a shower—sorry, I'm still waking up), I realize that my previous comment may be slightly daft.

Do drivers self-report their race as part of the license application process? If so, then one could (theoretically) use traffic stop records to get a pretty accurate racial breakdown of stoppees, without requiring cops to collect any additional data.

(My Maryland license doesn't list a race, but then I generally decline to answer that question. I honestly can't recall whether they ever asked me, and I don't know whether other states ask.)

However, this hinges on researchers' ability to get that data. They would need (a) the driver's license number from the cop's report, and then (b) the records (including self-reported race) associated with that license. Are either of these public record?

If either one isn't, then there are big difficulties with this approach.
posted by escape from the potato planet at 7:06 AM on September 10, 2014


Oh, and consider the difference between citations (which I think maybe are public record?) versus stops (which probably aren't). It'd be interesting to see the racial breakdown of both, but stops would probably be the preferred metric.
posted by escape from the potato planet at 7:09 AM on September 10, 2014


White people are pulled over more often than Asians. Stats like this are misleading. Not to say racism doesn't exist but do these stats show a racism towards white people compared to Asians? Or is there some other explanation? As for blacks and Indians ("Native Americans") a skew in the result can be explained *in part* by poverty - poor people are more likely to be pulled over because they drive beater cars with violations, don't keep up with MVA/DMV regulations or pay fines, have higher incidents of drunk driving, etc furthermore there is (usually) a greater police presence in poor areas since that is where more crime exists and a fallout is more police are looking for violations. It would be good to see these stats controlled for poverty because that's likely a major (though not only) factor at work.
posted by stbalbach at 7:49 AM on September 10, 2014 [2 favorites]


White people are pulled over more often than Asians.

Got a source? FTA: "American Indian drivers are even more likely to be pulled over (15 percent)."

Also, it jumped out at me that "nearly five percent of blacks weren't given any reason for why they were stopped, compared with 2.6 percent of whites and 3.3 percent of Hispanics."

I'm glad that this work is being done. Nearly 50,000 people answered the survey, according to the Bureau of Justice Statistics website. The website links to the actual survey and methodology used, for those curious.
posted by nicodine at 8:21 AM on September 10, 2014


You can't always determine/ assess/ guess/ assume the race of a driver before the stop.

I'd be more interested in a breakdown of what happens after a stop; What % of stops resulted in:
  • Informal verbal warning (My shift's almost done, do us both a favour and slow down)
  • Formal verbal warning (Your DLs clean, I'm gonna let you go with a warning)
  • Written warning (Get that light fixed, then mail in this postcard)
  • Citation / Ticket
  • Arrest (and for what)
  • High speed police chase
  • Physical struggle
  • Firearm discharged (by either)
  • Injury (to either)
  • Hospitalization (of either)
  • Death (of either)
Since all but the first of these involve the police running the driver's license against SCMODS or whatever, you c'd get the breakdown based on whatever's on the that state's DL records.

Oh, and consider the difference between citations (which I think maybe are public record?) versus stops (which probably aren't)

Aaaaaaand I guess I should go off to the potato planet in your place.
 
posted by Herodios at 8:23 AM on September 10, 2014


I don't doubt for a minute that driving while Black is a thing.

But "last contact with police" seems like an odd denominator to use. And it certainly doesn't follow from "relatively more black drivers (12.8%) than white (9.8%) and Hispanic (10.4%) drivers were pulled over in a traffic stop during their most recent contact with police." that "A black driver is about 31 percent more likely to be pulled over than a white driver, or about 23 percent more likely than a Hispanic driver."


Yes, I want to amplify this.

I'm sure there is good statistical evidence for the Driving While Black phenomenon, which makes it all the more infuriating that we're being shown something with no obvious connection to that phenomenon and told that it's evidence.

The author of the piece should be fined for statistical malpractice. That he is supposed to be a "data journalist" kinda makes me wanna vomit.

What the statistic we're given actually measures is, Under what sorts of circumstances do you usually have contact with the police? Is it mostly when you call them for help? Is it when you are stopped on the street for being "suspicious"? What the quoted numbers show is that, relative to white people, black people's encounters with the police are relatively more likely to be traffic stops than other kinds of circumstances (which could be either voluntary contact with the police, or other involuntary exposures).

Since traffic stops turn out to be the biggest component of involuntary contact, this may show that, relative to white people, black people's interactions with police are more likely to be involuntary. That seems plausible to me. But this would still be an ass-backwards way of trying to measure police bias, because a larger proportion of involuntary contact could be caused not only by more involuntary contact, but by less voluntary contact -- perhaps white people feel more comfortable calling the police, etc., etc.

This kind of argument really needs to be made with numbers about aggregate police encounters, not about each person's most recent police encounter.

Some of the other arguments in the article are well-supported. For example, the point about how frequently a traffic stop leads to a search is an important one.
posted by grobstein at 9:31 AM on September 10, 2014 [2 favorites]


A note about Native Americans: in at least two states with high populations (Montana and South Dakota), license plate prefixes indicate the county in which the vehicle is registered. So, if you're from a county with an Indian reservation, you can be "tagged" as Native without the cop ever having to see your face.
posted by desjardins at 10:30 AM on September 10, 2014 [5 favorites]


Also, given the huge distances and poor infrastructure in much of Indian Country, I would not be surprised if the number of vehicle miles driven was higher amongst Natives than whites. (In other words, if you have to drive 40 miles every time you want to get groceries, your risk of being stopped is greater just because you're on the road more. This is no way invalidates the racism of traffic stops, it just shifts the cause elsewhere.)
posted by desjardins at 10:34 AM on September 10, 2014


It's also possible that the federal jurisdiction of the reservations makes some law enforcement things happen disproportionately there compared to other places; I've heard that it's much harder to prosecute some crimes if they're done on the res, which might be bullshit but is certainly a notable public perception. (Note that this doesn't make racism allegations go away but instead shifts them to be structural more than individual.)
posted by NoraReed at 1:07 PM on September 10, 2014


I know this can't be right, because racism is over in America.
posted by Mental Wimp at 3:14 PM on September 10, 2014


take a photo of the driver, so that researchers could arrive at their own perception of the driver's race after the fact

It's hard to imagine any justification for doing any of these as part of daily police work.


Body Cams on officers?
posted by radwolf76 at 3:21 PM on September 10, 2014 [2 favorites]


Another study to corroborate what everyone knows; okay, it does need to be official. Done.

No - not really; eighteen more studies will follow, half of which will agree and half disagree with these results. And a little voice from the back will want the study repeated, different criteria to be used - and it will just go on.

What we DO know is that yes, black people get pulled over more often than white ones. Now we know that hispanic people are in between blacks and whites and native people get the worst end of the deal.

If I were making up my own statistics, I'd say that people driving old beat-up cars get pulled over more often than people driving newer cars (poverty is the point). And I'd guess that young people in a group in a car are more likely to get pulled over than a single young person in a car. And if you have a bumper sticker on your car that says "Ask Me About My Two Dads" or "Queer And Proud Of It," you can pretty much guarantee you'll get stopped often. And I'd bet that the neighborhood and time of day makes a difference, too. And driving too slowly through a residential area, especially a NICE area, giving the appearance of scanning the houses for later mischief will get you pulled over if you're young, male, or black. If you're female and attractive, you'll get pulled over more often than I would; in fact, if you're a middle-aged woman, overweight, with three kids in the car, you probably don't have to worry about the police bothering you. And back to bumper stickers - they matter.

The question is: What's to be DONE about it? How? We have a golden opportunity right now to change the ways the police have been doing things because of the Ferguson murder, but that window of opportunity won't stay open forever.
posted by aryma at 8:57 PM on September 10, 2014


No - not really; eighteen more studies will follow, half of which will agree and half disagree with these results. And a little voice from the back will want the study repeated, different criteria to be used - and it will just go on.

This is pretty much the scientific process in motion. It's a wonderful thing, no? If half agree and half don't of course you repeat the tests. You refine your methodology until you can get accurate and repeatable results.

If I were making up my own statistics...

I think you are missing the point. These things are studied so one doesn't have to make anything up. Statistics aren't numbers pulled out of one's ass.
posted by cjorgensen at 7:07 AM on September 11, 2014


Got a source?

Is that some kind of willful blindness to what the linked study shows? Blinded by the white?
posted by stbalbach at 2:12 PM on September 11, 2014


Sorry, cjorgensen, instead of "making up my own statistics," I should have said something like: I'm not going to look for studies supporting or denying these thoughts (feel free to do so if you disagree), but I'd say that people driving old beat-up cars get pulled over more often ...

As far as the scientific process in motion goes, accurate results are always accurate to those doing the study and if those results are repeatable, they're accepted by the entire scientific community as accurate - until a few years later when more studies change the "accurate" to "inaccurate." Of course this is the right way to do things scientifically - what would the alternative be? But nevertheless I find it best to take the "absolutist" tone of the latest study with a grain of salt (speaking of which, up until recently salt was considered unhealthy in any measurable amount, but now the story has changed somewhat and studies suggest it's not so unhealthy after all).

But back to the point - do you disagree with the other situations I described as bait for police pull-overs?
posted by aryma at 2:51 PM on September 11, 2014


...speaking of which, up until recently salt was considered unhealthy in any measurable amount, but now the story has changed somewhat and studies suggest it's not so unhealthy after all...

This is a failure of reading comprehension, not a problem with research. Yes, the understanding of how salt affects health has changed over time (how could it not), but never to my knowledge has it ever been stated by a reputable scientist that salt was unhealthy in any measurable amount.
posted by Mental Wimp at 3:06 PM on September 11, 2014


But back to the point - do you disagree with the other situations I described as bait for police pull-overs?

In some cases yes. This is again why the studies would be appropriate.

Fr example: Perhaps it's geographically dependent, but I can't imagine a can with a "My two dads," or "LGBTQ" bumper sticker would get pulled over more often, but then I live in Iowa and we have marriage equity here, and I could be wrong. Again, why these things are actually important to study.
posted by cjorgensen at 3:51 PM on September 11, 2014


It's also possible that those bumper stickers would be self selected among other lines; maybe only queer people in "good" neighborhoods would feel comfortable putting them on their cars, so they might have more money, nicer cars and so get pulled over comparatively less. Or maybe there's a story that would explain the reverse.
posted by NoraReed at 3:55 PM on September 11, 2014


« Older Measure of the Sierra Madre   |   Missouri abortion waiting period bill: a veto... Newer »


This thread has been archived and is closed to new comments