Wednesday, September 24, 2014

The correct denominator

For the past few days, I have taken issue with the ACLU's assertion that racial disparity in traffic stops indicates alarmingly high rates of racial profiling in Lincoln. I think their analysis is shallow. I would be the first to agree, however, that we need to better understand this phenomenon. I've reached the limit of my own ability to work with these data, but I am encouraged by what appears to be a renewed interest in this field of study, as evidenced by the Department of Justice initiative announced last week.

I have a second problem with the ACLU's report, and it concerns the denominator. Their numbers are wrong. I'll give them a do-over, because the Crime Commission's data is not right, either. Here's what the ACLU says:

"Black drivers in Lincoln are stopped by the Lincoln Police Department almost three times as often as they should be: the black population of the area is 3.5%, yet black drivers were 9.6% of the stops."

Lincoln's population is not 3.5% black. In the 2010 Census, it was 3.8% black. The ACLU is using the Lancaster County population data, not the City of Lincoln. The Crime Commission's report also uses incorrect data. While they have it right in table B on page 14, and in table C2 on page 17, they have this totally incorrect statement in the executive summary on page  4:

"The Lincoln Police Department stops Blacks almost three times as their local adult population (9.6% to 3.3%)"

The bad grammar comes directly from their report, and I have no idea at all where that 3.3% figure came from. Maybe it's a cut-and-paste error from a previous year's report, but it is wrong. To be clear, in the 2010 census, the one race only black population of Lincoln is 3.8%. That may not sound like much of a difference, but it is. It is the difference between "almost three times" and "more than twice." The fact that the mistake is in the executive summary doesn't help. I suspect some people read no further.

Moreover, that was the percentage of Lincoln residents who identified themselves as one race only. But a large number of residents identify themselves as black and some other race. Beginning n 2000, for the first time you could specify more than one race. The Census Bureau began reporting both one race only data, and two or more races data. In the 2000 census data,  3.1% of Lincoln residents were black, and 3.8% were black or black and one or more other race.  In the 2010 census, this had grown to 3.8% and 5.3%, respectively. Check the data yourself with the Census Bureau's excellent tool, American Fact Finder, but be sure you get the City of Lincoln, not Lincoln County.

Based on the change from the 2000 census to 2010, It appears that more and more people are identifying themselves as multi-racial. If the Crime Commission is going to use the population of black residents as a comparison for traffic stops, they should use the 5.3% figure, or should just report both figures. Here's why: Nebraska's law requires police officers to collect data on the race of all motorists stopped, but prohibits officers from asking the driver. Race is simply the officer's best guess. In a multi-ethnic, multi-cultural world, these guesses are becoming more problematic. What would you select if you stopped Louisiana Governor Bobby Jindal breezing down Vine Street 10 over the limit?

Here are a couple of rather prominent multi-racial Americans: Barack Obama and Tiger Woods. I think that if either of them was pulled over in Lincoln for expired tags, the officer would almost certainly select "B" from the drop down, not "W" or "A". Hence, the Crime Commission should use 5.3%, which is a lot closer to 9.6% of the stops than the inaccurate 3.5% quoted by the ACLU or, worse yet, 3.3% in the Crime Commission report's executive summary. The correct denominator is important.

Still, the fact that 5.3% of the population is black or black and some other race, while 9.6% of the stops were of black motorists is of concern to me. I'd like to understand that better, so we can apply the right strategy to the portion of that disparity that may be the result of actions that are unjustified, unfair, or ineffective.

As an aside, the 726 black drivers involved in traffic crashes reported to the police in Lincoln last year would represent 5.4% of the total, which was 12,915. Maybe it's coincidence, but that's mighty close to 5.3%, and tends to confirm my belief, described in Monday's post, that traffic crash driver demographics make the best denominator for examining disparity in traffic stops.

Links to the series:



Anonymous said...

Statistics can be skewed to show just about anything, can't they! Far as I'm concerned the ACLU should be outlawed, disbanded, whatever it takes to get rid of them. Seems to me they cause a lot more problems than they solve.ntacnc

Anonymous said...

Never happen 8:05.
As much as I despise a lot of their actions they do have First Amendment protection. My pet peeve with them is how vocal they are on many issues but how silent they are when the Second Amendment is violated.

Gun Nut

Anonymous said...

When all else fails, manipulate the data.

Raquel said...

I think part of the issue today is a lack of personal accountability. If I were pulled over for speeding, I would no doubt be listed as white although I consider myself Hispanic, as fair as I may be. The point is not my ethnicity but the fact that I was speeding. I will admit that people with a less than favorable opinion of other races do exist. However, if an individual is breaking the law, that is the concern. If you break the law and are caught, the responsible thing is to accept your mistake and pay your dues. It is, in my opinion, a lazy excuse to pursue an organization such as the ACLU to fight over race when your actions are what is at fault.

Tom Casady said...


Just for the record, despite being rather disappointed in their shallow analysis of the 2013 traffic stop data published by the Crime Commission, I generally have no bone to pick with the ACLU.

In many cases, I appreciate the ACLU standing up for civil liberties. But in some cases, I think they shoot from the lip. This is one of those that just got under my skin a little bit, because a simple discussion could have corrected an error and (perhaps, at least) shed some light on the cause of racial disparity in Lincoln's traffic stop data.

It's fine to question authority; it's not good to jump to unproven conclusions. I would really prefer to have productive discussions, rather than be called out in a report without any chance to dialog about these data first.