The Seattle Police Department just published the results of its first court-ordered Disparity Report. It attempts to find out how much racial disparity existed in different aspects of the department’s police work, like use of force.

The review uses a form of statistical analysis called propensity score matching, which attempts to control for as many variables. By doing that, SPD is trying to find out, as accurately as possible, how much racial bias exists.

The CPC is reviewing the report right now. While the results are mixed, our early analysis has identified some concerning disparities detailed below. We’ll be analyzing these numbers in greater depth over the coming weeks.

Under the SPD’s bias-free policing policy, SPD is required to work with the Community Police Commission and the Office of the Inspector General to address disparity in policing. In light of these findings, we look forward to discussing what this collaboration could look like with SPD and OIG

Subjects of color are more likely to be frisked than white subjects

Controlling for other variables, people the officers perceived to be Asian were frisked 33.9 percent more frequently than white people. That’s the most for any group, but the disparity exists for other racial groups as well.

People whom officers perceived to be Hispanic were nearly 30 percent more likely to be frisked than white people. Black people were 18.2 percent more likely to be frisked than white people, and American Indians and Alaskan Natives were around 4 percent more likely to be frisked than people perceived to be white. 

Subjects of color are less likely to be found with a weapon when frisked

Despite being frisked the least, people perceived as white were the most likely to have a weapon on them while being frisked.

For example, as the report states, “Although subjects that were perceived to be Asian were frisked 34 percent more than white subjects, they were found with weapons 21.5 percent less often than white subjects.”

Likewise, people perceived to be American Indian and Alaskan Native were 40 percent less likely than white people to have a weapon on them while being frisked. For people perceived to be black it was 34.5 percent, and for people perceived to be Hispanic it was 18.8 percent. 

People of color are more likely to have a firearm pointed at them than white people

Perhaps the most concerning finding is the disparity in people at whom police point their firearms.

Hispanic people were nearly twice as likely as white people to have a firearm pointed at them. Controlling for as many variables as possible, the study says, “white subjects has a firearm pointed at them in 10.4 percent of force reports, while Hispanic subjects had firearms pointed at them in 19.4 percent of force.”

That disparity carries over to every group. Black people were 42.3 percent more likely to have a firearm pointed at them. Similarly, for Asian people, it was 42 percent.

Much of the data needs more study, but action is needed

The Seattle Police Department writes in the review, “These results should be taken as initial and not final estimates.”

This is just the first phase of this research. Seattle Police are scheduled to start phase two soon, and they’re trying to focus on controlling for more variables to better understand the causes of these disparities.

For example, Seattle Police argue the finding that officers are more likely to point guns at people of color needs to be studied further. They say the review “does not represent well the question of weapon possession by the subject in a use of force,” and that “the decision to point a firearm will almost always follow the visible presentation of a weapon by a subject.”

There’s also some good news in the report. For example, SPD found there is little to no disparity between subjects of color and white subjects for type 1 and type 2 uses of force.

But, it’s important to note the question isn’t whether disparity exists. Seattle Police acknowledge that disparity does exist in this review. The question is what SPD is doing about it, and while their response to these disparities can become more nuanced with more data, these are serious issues that need to be addressed now.