By Paul Kiefer
At Tuesday’s city council public safety committee meeting, Seattle University Crime and Justice Professor Jacqueline Helfgott presented the results of the 2020 Seattle Public Safety Survey, a collaboration between her university and the Seattle Police Department (SPD) conducted annually since 2015. The goal of the survey, Helfgott told PubliCola, is to monitor the public’s perceptions of crime and policing year-by-year, with the goal of shaping SPD’s neighborhood-level response to public safety concerns. The survey results were part of a series of presentations about the city’s efforts to stand up non-police public safety programs.
According to Helfgott, the survey began as a tool for SPD to tailor its neighborhood-level presence; an SPD captain in the North Precinct, for instance, might direct officers patrolling in Ballard to bear in mind residents’ concerns about “aggressive panhandling.”
Responding to a question from Councilmember Lisa Herbold about how the survey related to the city’s plans to shift public safety away from police, Mayor Jenny Durkan’s senior public safety advisor, Julie Kline, told council members that the survey results could “help the city choose how to prioritize non-police public safety investments” that can address Seattle residents’ public safety concerns.
The survey itself primes its participants to view Seattle through the lens of enforcing public order: in one question, the survey presents children skipping school, disrespecting adults and playing loud music as threats to the “social cohesion” of a neighborhood. I
What are those concerns? For the more than 11,000 Seattle residents who responded to the survey, the most common concern was city government itself, followed by “public order crimes,” property crimes, reduced police capacity, and homelessness. With the exception of the outsize frustration with city politics, respondents to the previous six surveys had almost the exact same priorities.
There were some variations by neighborhood. Respondents living north of the ship canal, for instance, perceived homelessness as a more significant public safety threat than residents elsewhere in Seattle, while respondents in Hillman City expressed lower confidence in SPD than residents of any other neighborhood.
Most people who took the survey expressed confidence in SPD; the department’s favorability ratings were highest among Black, Latinx, Asian, Indigenous, and older respondents, and lower among white respondents and those younger than 25. White people, particularly white women, were strongly over-represented in the survey. Overall, the survey showed show only a slight decrease in confidence in SPD across all neighborhoods since 2019.
Helfott’s presentation to the council did not clarify the survey’s role in shaping the city’s efforts to rethink public safety. On a granular level, the survey is still calibrated to shape SPD’s approach to policing Seattle, not how the city should invest in alternatives to policing. And the survey itself primes its participants to view Seattle through the lens of enforcing public order: in one question, the survey presents children skipping school, disrespecting adults and playing loud music as threats to the “social cohesion” of a neighborhood. In contrast, most of the existing non-police public safety programs in the city aren’t intended to enforce public order.
Helfgott said that her research team doesn’t have plans to substantively change the survey’s framing; doing so, she said, would prevent SPD and other city leadership from comparing public responses from year to year. But as the city’s approach to public safety shifts away from the police-based model on which the survey was built, its relevance to decision-makers seems likely to decline.
So I read the survey, and the reporting on the survey, and I’m really kind of disappointed in the coverage of what the survey says but also what it doesn’t say (and how the survey works). I brought it up with a friend of mine who has a Master’s Degree in Psychology (specifically Human Factors, so bad survey design is kind of their thing), and they raised some basic objections to the survey. I feel like this is valid and important because I think there are important questions and caveats about not only the survey itself, but how the data is reported on and used.
To start, the topline scores in the survey (police legitimacy, social cohesion, informal social control, social disorganization, and fear of crime) are the average of people’s responses to several different questions. Those questions are supposed to get at different aspects of the general concept. Then, the survey team neither reports the full responses to the individual questions nor says anything about the focus group portion of the survey, which should be following up on people’s reasons for their survey answers.
Question content matters. For example, the legitimacy score could have been artificially (but not intentionally) inflated through the phrasing of some of the individual questions. Three of the the fourteen items on the “legitimacy” scale are variations on “you should do what the police tell you”. Apart from people who are actually willing to fight the police, I’m not sure I would expect to see many low answers to those questions, because “I will follow orders because I trust them” and “I will follow orders because I think the police will kill me” would both result in high scores, dragging up the overall average. Remember, even the ACLU tells you to do what the police say, even if you believe the order is unlawful or unconstitutional, so I would expect a high legitimacy value on a question like that regardless of motivation.
By the way, that hypothesis would fit with the observation that Black respondents rated the police as “more legitimate” than white respondents. I would expect Black folks (and POC more generally) to be more afraid of the consequences of defying the police than white people.
Next, the format of the questions and the possible distributions the resulting answers might have aren’t really exposed to the public or published in the survey data. Remember, these questions are *not* ‘yes or no’ approval rating-style. Instead, these are the averages of many individual scores on scales from 0-100 where 0 is “completely disagree” and 100 is “completely agree”. Yet, the survey seems to reduce this to find that 50something% of Seattlites have a net positive opinion of the police. If it’s not just the mean and there are actually a lot of *individual responses* to survey items near 50, that would be consistent with most, of the folks taking the survey having an overall opinion of something like “I *kind of* trust the police,” or “I *sometimes* trust the police”.
If that’s true, it indicates widespread belief that the police are not irredeemable but are obviously flawed. On the other hand, if the average reflects a small number of 0s averaged together with an overwhelming majority of 70s, then most people would be content with the police, and anti-police sentiment really is just AnTiFa TrOuBlEmAkErS. Again, the data probably exists to answer that question; it’s just not made public and/or I don’t have access to that… but I can see why that data might be *really* important (and I hope you can too).
And on that note, the way this is framed here! Even if we ignore everything I’ve said and assume, for the sake of argument, that those are approval rating scores, then this still implies that upwards of 40% of the population disapproves of SPD, which is maybe not good news? Widespread distrust of SPD that doesn’t get all the way to “violent outrage” shouldn’t exactly be a great situation from anyone’s POV, in terms of things like “will people cooperate with investigations?” or “will advocates for (or against) police reform get reelected?”. People are setting themselves up for surprise failure.
Finally, addressing the idea that popular distrust of the police hasn’t changed much, I want to point out that the report collected data from October 15, 2019 to November 20, 2020. George Floyd died precisely in the middle of the data collection range. Now, I would argue that the people who collected data would know when the survey responses came in, but they don’t mention a single thing about George Floyd. I find it nigh-on-impossible to believe that the people collecting the data for this survey couldn’t separate 2020 into two separate years (and still present this aggregate data). If there’s a legitimacy gain in the first half and a decrease in the second half, it supports the idea that
Also as an aside, the report has a really poor choice of colors for their lines in some of their charts, but also apparently people like the police in Seattle more than other cities, which is both amazing and depressing to me.
If you don’t like the survey how about Publicola doing their own? I’d be interested to see how your survey questions are framed.
Steve, the survey is self-selected respondents, so it’s not an accurate picture of how people in Seattle feel. It’s an accurate picture of how people who took the survey feel.
Paul: Thanks for noting how the survey “primed” respondents. That question about disrpecting elders really threw me when I took it. That did not belong in the survey IMHO.
I do wish you would have highlighted the self-selection issue in the post as it really bears on interpretation. Honestly disturbed that it was presented to city council with credulity.
It was probably always self- selected respondents, which means it can still be used to track changing policing opinions over time. Perhaps the problem here is that you just don’t like the conclusions.
Paul: In other words, black people are not as scared of the police as white people are trying to convince them they should be.
Also, that nasty “police-based model” will decline because the warm and fuzzy approach is more successful with violent criminals. If we could only hand them enough free stuff and convinced them they are loved…problem solved. Lets try that Paul and see if it works. I will make my prediction right here and now: I will be LMAO at you dipshits.