Who Watches the Watchers?
Investigating use of force by the Seattle Police Department
In 2011, the United States Department of Justice (DoJ) opened a case against the Seattle Police Department (SPD), focusing on “whether SPD engaged in a pattern or practice of unconstitutional policing through (1) the use of excessive force; or (2) discriminatory policing.”
The DoJ ultimately found that the SPD acted in violation of the Fourth Amendment nearly 20% of the time, largely due to inadequate training and lack of oversight. SPD officers were found to escalate minor incidents, particularly in encounters with those with mental illnesses or under the influence, as well as apply excessive force in a variety of inappropriate situations.
Though no legal finding was made, the DoJ also found evidence of “troubling practices that could have a disproportionate impact on minority communities,” including disproportionate stopping of people of color and incidents in which SPD officers used racially inflammatory language.
As a result of the investigation, the DoJ and SPD reached a settlement requiring increased training and oversight, as well as monitoring by the DoJ. A recent report in April 2017 claimed that the reforms have worked - SPD officers both apply force less often and at a lower magnitude. However, the monitoring team still found evidence of racial disparity in the subjects of force incidents, encouraging more data collection and research.
Our goal is to provide tools to allow anyone to research the available data, as well as present some initial conclusions that we found. However, we encourage you to play with the data and come to your own conclusions.
Use of Force Definitions
The different levels are defined as follows in the Seattle Police Department Manual:
Level 1: Force that causes transitory pain, the complaint of transitory pain, disorientation, or intentionally pointing a firearm or bean bag shotgun at a person.
Level 2: Force that causes or is reasonably expected to cause physical injury greater than transitory pain but less than great or substantial bodily harm, and/or the use of any of the following weapons or instruments: CEW1, OC spray2, impact weapon, bean bag shotgun, deployment of K-9 with injury or complaint of injury causing less than Type III injury, vehicle, hobble restraint.
Level 3: Force that causes or is reasonably expected to cause, great bodily harm, substantial bodily harm, loss of consciousness, or death, and/or the use of neck and carotid holds, stop sticks for motorcycles, impact weapon strikes to the head.
Level 3 - Officer Involved Shooting: While not defined in the manual itself, this subset of Level 3 is often separated in the data, since it is a sensitive topic.
1 - Conducted Electrical Weapon (Taser), 2 - Oleoresin Capsicum Spray (Pepper Spray)
As expected, the majority of incidents are Level 1, making up about three-fourths of total incidents:
Breakdown of Use of Force Incidents by Level (2014 - Present)
Seattle Police apply a variety of techniques - the majority of Level 1 incidents involve handcuffs, with higher levels having far higher levels of control holds and verbal force.
Percent of incidents involving Force Instrument/Option
(July 1, 2014 - October 31, 2016)
Who is affected?
From the original investigation to the most recent data, the DoJ has continually found evidence for bias against minorities in uses of force. In addition, public perception of the SPD includes a strong sense of systematic racial bias. A 2016 community survey done by Anzalone Liszt Grove Research found that about half of Seattle residents believed the SPD engages in racial profiling and treats people differently based on their race. Perceived marginalized groups include African-Americans, Latinx, and homeless populations.
Further exploring the data, we notice right off the bat that the number of incidents involving White and Black people is fairly comparable. Considering the 2015 US Census estimated the percent of White and Black persons in Seatle to be 69.5% and 7.9% respectively, this statistic is troubling. Even accounting for crime rates, the values aren't proportional.
Considering the gender field, we see a similar disparity between men and women, but this is more explainable considering differences in crime rates between men and women.
Number of Use of Force Incidents by
Stacked by (2014 - Present)
Looking at the force instruments/options used by Seattle police, we find another problematic statistic - Seattle police are far more likely to point a firearm towards racial minorities than White people.
Percent of Incidents Involving by Race
(July 1, 2014 - October 31, 2016)
Police are also more likely to determine minorities as "Unimpaired" compared to White people, who along with indigenous people, have the highest rate of being labeled "Impaired by Drugs or Alcohol."
Percent of Incidents with Subject by Race
(July 1, 2014 - October 31, 2016)
Where do use of force incidents happen?
While constructing our map visualization, an interesting trend emerged - disproportionate numbers of incidents involving Black people occurred heavily in southeast Seattle (Mt. Baker, Beacon Hill, Rainier Beach). If we hover over R2 on the map (roughly corresponding to the Mt. Baker neighborhood), looking at the pie chart in the bottom right, we see the population of Mt. Baker is over 1/3 White people and under 1/6 Black people. However, considering the corresponding bar graph on the left which displays incidents stacked by race, we see that the number of Black subjects makes up over 50% of incidents. We invite you to modify the filters to tease out similar conclusions to the sample above.
click to interact
Conclusion
We acknowledge that systematic racial bias is a problem that has no easy solution. We also acknowledge that the actions suggested by the DoJ and implemented by the Seattle Police Department are steps in the right direction.
However, the issue of racial bias in the SPD has been a potential issue since 2011, and less progress has been made towards it than should have. We urge the DoJ to continue investigating this problem, maintaining oversight until more progress has been made. We also hope that more data on the issue will be released, similar to other cities in the US. The data set provided by the SPD has limited fields, making it difficult to perform a substiantial analysis. With more fields/features, we would be able to draw more conclusive and nuanced conclusions.