A few years ago a shop owner I know was burglarized.
He asked me to help him get the surveillance tapes into a format he could present in court, so I did so. Reviewing the tapes while transferring was interesting. It was a simple burglary. A guy came in through the roof. He appeared to be high on crack or crystal meth and weighed probably 100 pounds. He tripped the alarm, panicked, ran about the store, couldn't get out and eventually exited back through the roof. This was about 3 in the morning. 15 or 20 minutes later 2 officers arrived with my friend in his pajamas looking sleepy. He let them in and they crept about the store pistols drawn. No one was there. One of the officers pointed to the drop ceiling (good call... there was a gaping hole there) and they eventually exited the store. 3 or 4 hours later the swat team arrived. They stood behind shields and put in some sort of gas bomb. The tapes filled with smoke and you couldn't see anything for a period. Then, a number of officers entered with gas masks and examined inside the drop ceiling with a camera. No one was there of course. There were probably not less than 15 of them and they stayed around until noon or so wandering about and poking stuff. At one point one of the officers noticed the surveillance camera and turned it so they couldn't be filmed further. No one was brutalized but it appeared to be extreme over reaction. The store reeked of tear gas for weeks. I have to wonder exactly what the cost was to taxpayers. This certainly isn't an extreme case but it was an eye opener. Lots of these guys just want to play GI Joe and don't care what they damage nor how much money they waste. It is time for some reform.
Intuitively I agree with this, and think militarization of law enforcement is terrifying, but this report at a glance has some issues. I want to understand the trade off here, but they fail to pose SWAT as a real trade off for officer safety in many cases. For example, they cite that of the raids where police predicted weapons, 35% of the raids turned out to have weapons, 32% did not, and for the remaining 33% it is unknown if they did or did not. The way they interpret this data is:
"We found in the course of our investigation that the
SWAT team found weapons (the overwhelming majority of
which were firearms such as handguns, but rarely assault
rifles) in just over one-third of the incidents in which they predicted finding them, which suggests the police are not particularly good at accurately forecasting the presence of weapons."
This is transparent bullshit, because in their very table they state that a third of the raids it was unknown. If we assume that the 'unknown' state is random (an assumption), ie ~half of those unknown accounts had weapons and the rest didn't based upon the known positive and negative rates being close to even, this means that no, the police didn't guess only one third of the time correctly, they guessed correctly approx 53% of the time, more than half of the time. The next step is to ask if a 50/50 chance of being right that you might be under threat of a weapon means you should err on the side of SWAT, but they don't ask that, they just say that we should "we would expect them to find weapons in nearly all of the incidents studied." Says who? Risk is a game of probabilities, and obviously there needs to be some threshold for which cops deploy SWAT with regards to the likelihood of weapons being present. Regardless of if the randomness of the unknown assumption is true, they are transparently mis-characterizing the data here.
Additionally, their section on race seems transparently broken. They are measuring absolute measures, not relative measures. To understand race bias in SWAT, you need to understand the rate at which SWAT is deployed for warrants for a given demographic. Instead, they spent most of their analysis (if this rate is in there, I missed it) on absolute metrics; but we all know already that on an absolute basis, law enforcement spends a disproportionate amount of time in minority communities. You can argue if this is due to systemic bias or not, but this is obviously a dependent variable if we're trying to answer the question "Do cops deploy SWAT more if they are dealing with minorities?", and it doesn't seem to be treated as such.
So, a third of the time there were guns (though not specified if they were being used against the officers, up in a safe, or tucked away somewhere), a third of the time there were not, and a third of the time the cops were too sloppy to accurately assess the presence of firearms?
And these are the bozos we want to let have automatic weapons, body armor, and carte blanche to show up in the middle of the night? Fuck that.
They don't specify, but keep in mind that the police departments are the ones reporting these statistics, and they have an interest in showing how dangerous the situation was and in justifying their use of force.
It's also very important to note that approximately 40%[1] of Americans report having a gun in their home.
"Officer safety" doesn't ring very true when the rate at which firearms are confirmed in the home is approximately what you would expect, given the prevalence of firearms ownership in the country.
If around half of households have at least one firearm, and officers predict that a gun will be present every time, then they'll be right... about half the time.
It's a good point but the people who legally own guns may be a largely separate population than the population who are being targeted by drug raids. So your point about them being right by default with the same % is wrong, since that assumes it's a random sample of the population.
That said I agree these stats are meaningless regardless of the position one takes so this report undermines it's credibility (and my trust that I'm reading something well researched) by trying to spin them. Which was my original point.
While I agree with much of what you say, your "safety" justification confuses me. If someone is wanted for a lesser, non-violent crime, it is arguably safer for the police to knock on the front door, as residents often mistake SWAT teams for armed robbers and grab their weapons in fear for their lives. There is almost zero incentive to shoot a cop knocking on your door unless you are facing decades in prison if caught, the purpose for SWAT raids on non-violent offenders is to prevent escape/the destruction of evidence. If the police were concerned with safety, they would wear chest mounted cameras (proven to reduce assaults on officers), but they actively resist this.
There are two separate questions that you're confusing.
You seem interested in "is the use of SWAT systematically _more_ biased against minorities than law enforcement in general."
The report seems to be addressing "is the use of SWAT systematically biased against minorities, _regardless_ of the state of the rest of law enforcement."
Both are interesting questions, but it's silly to criticize an attempt to answer one question for not answering the other. Minority individuals abused by SWAT probably won't be cheered up by "don't complain about SWAT's bias, you also get an unfair bias in jaywalking tickets."
The point is the second claim says nothing. You can also ask "is the use of blue uniforms systematically biased against minorities" or "is the use of police hats systematically biased against minorities" etc and the answer will be "yes" due to the dependent variable.
I don't think you actually read the report, but if you had, you might have seized upon a few limitations imposed upon this study. Consider this statement, which appears on page 9:
The vast majority of the incident reports considered,
however, did not consistently and systematically document
information in such an easily transcribable manner, instead
communicating or expressing answers—if any at all—to the above questions in a textual narrative (often located at
the end of the incident report). It is, of course, relatively more difficult to generate a categorical variable from purely narrative text, and, in particular, one must decide how to deal with narratives that are silent or ambiguous with respect to one or more of the questions posed above.
Or, there is this statement, which appears on page 11:
By examining impact rates, it is possible to control for racial disparities in the underlying populations impacted by SWAT deployments. Rates are expressed in terms of individuals impacted by SWAT deployment per 100,000 individuals. In particular, to calculate the white, Black, or Latino SWAT impact rate in a given jurisdiction, the number of white, Black, or Latino individuals impacted by SWAT deployments is divided by the total white, Black, or Latino population in that jurisdiction; the corresponding ratio is then multiplied by 100,000 to obtain the impact rate per 100,000.
There is more information on the data that the ACLU collected starting on page 27. Consider this statement:
The ACLU filed public records requests with more than 255 law enforcement agencies during the course of this investigation. One hundred and fourteen of the agencies denied the ACLU’s request, either in full or in part.
They assess the quality of the data received as follows:
The records that were produced revealed an extremely
troubling trend: that data collecting and reporting in
the context of SWAT was at best sporadic and at worst
virtually nonexistent. Not a single law enforcement agency
in this investigation provided records containing all of
the information that the ACLU believes is necessary to
undertake a thorough examination of police militarization.
Some agencies (e.g., Tupelo, Mississippi) provided
records that were nearly totally lacking in important
information. Others (e.g., Salt Lake City, Utah) provided
records that were quite lengthy, though still incomplete
and extremely difficult to analyze because of their lack of
organization.
Or their conclusion:
This variation has two immediate results: (1) any analysis of the data will necessarily have to contend with a large number of unknowns (as demonstrated above) and (2) it makes
systematic, thorough, and uniform collection of SWAT
data, at any level of government, impossible.
The ACLU acknowledges the flaws in this report, both in the body and in their section on methodology. And, they provide some good reasons for these flaws.
All men: 428.0
White: 347.7
Black: 2696.9
Hispanic: 664.6
There is roughly a 9x difference in incarceration rates (expressed per 100,000) in Utah in 2008. Now, the ACLU report gives numbers on three different cities in Utah.
In Ogden, black people are 39.55 times more likely to be impacted by SWAT actions than white people. In Unified, black people are 10.33 times more likely to be impacted. Meanwhile, in Salt Lake City, the number is 7.33 times.
Again, I'll remind you of this statement, which can be found on pages 9 and 10:
It is important to note that the data analysis in the report does not seek to make statistical estimates about the larger universe of SWAT deployments nationwide. Rather, the analysis is descriptive in nature, providing a general picture of SWAT deployments for this small cross section of
otherwise randomly chosen law enforcement agencies—the
information contained in the documents received is not
used to make more general, broader statements about the
use and impact of SWAT nationwide.
It is difficult to criticize a report's methodology when the authors acknowledge the shortcomings, explain why the shortcomings exist and try to express the data they receive in as diplomatic a matter as possible. Consider that this report is not meant to provide a statistical analysis on the state of SWAT in the United States. Rather, it is meant to 'shine a bright light' (to use their term) on militarization that has occurred largely without any public oversight.
He asked me to help him get the surveillance tapes into a format he could present in court, so I did so. Reviewing the tapes while transferring was interesting. It was a simple burglary. A guy came in through the roof. He appeared to be high on crack or crystal meth and weighed probably 100 pounds. He tripped the alarm, panicked, ran about the store, couldn't get out and eventually exited back through the roof. This was about 3 in the morning. 15 or 20 minutes later 2 officers arrived with my friend in his pajamas looking sleepy. He let them in and they crept about the store pistols drawn. No one was there. One of the officers pointed to the drop ceiling (good call... there was a gaping hole there) and they eventually exited the store. 3 or 4 hours later the swat team arrived. They stood behind shields and put in some sort of gas bomb. The tapes filled with smoke and you couldn't see anything for a period. Then, a number of officers entered with gas masks and examined inside the drop ceiling with a camera. No one was there of course. There were probably not less than 15 of them and they stayed around until noon or so wandering about and poking stuff. At one point one of the officers noticed the surveillance camera and turned it so they couldn't be filmed further. No one was brutalized but it appeared to be extreme over reaction. The store reeked of tear gas for weeks. I have to wonder exactly what the cost was to taxpayers. This certainly isn't an extreme case but it was an eye opener. Lots of these guys just want to play GI Joe and don't care what they damage nor how much money they waste. It is time for some reform.