Police reform and "brutally honest" data: A chat with Seattle PD's new tech chief
Last March, the Seattle Police Department made waves when it hired Amazon.com Vice President Greg Russell as its new chief information officer. The city's CIO must be more than just an IT expert: Whoever holds the position will have a hand in some of the department’s most pressing issues, particularly as it implements federally mandated police reforms. These include creating a system to recognize patterns of officer misconduct, balancing privacy issues with the collection of data to predict and fight crime, and implementing a body camera program for police in the city.
At the time of his hire, Geekwire reported that the position would be a “major cultural shift” for Russell, who had spent his career in the corporate trenches. This proved prescient: Russell abruptly quit only five months later, for reasons he would not disclose. In reaction, the department veered from hiring a cultural disruptor to hiring one of the most reliable tech experts in the local public sector.
Bill Schrier is a 30-year veteran of public service – serving as Seattle’s Chief Technology Officer from 2003 to 2012 – and one of the state’s top minds in civic tech. About a week into his tenure as the Seattle Police Department’s new CIO, he took time to discuss the biggest issues on the road ahead. The interview has been slightly abridged for length and clarity.
So, what happened with Greg Russell? An unnamed source told the Stranger that the city “took away his workforce.” Any merit in that?
I really don’t know what happened to Greg. Certainly he was used to being able to move pretty fast from Amazon, in terms of hiring people and allocating budgets to projects, but that’s anecdotal from talking to folks at Seattle Police. He wanted to do something similar (at SPD), and of course we have elected officials and budget offices who need to weigh in on decisions. That was probably a frustration.
You’ve said you’ll be implementing the program that Russell developed. Can you offer some of the broad strokes?
The top priority is something we call the Decision Analytics Platform, or DAP. It has several different facets, but the first to be implemented will be essentially an officer database.
We have a number of databases at the department that relate to officers – what their training has been, their personnel files, times they’ve been involved in use of force. DAP will try to pull all those pieces together to give an overall picture of officers for a variety of purposes. One is potential early intervention, to see if there’s a pattern of behavior that needs to be addressed, to make sure officers comply with their orders and take proper training.
Is the department currently unable to get red flags on problematic officers, such as those predisposed to use force? That capability doesn’t really exist right now?
It exists, but it’s manual process with several different databases. This will speed the ability to help.
I've read you're also focused on improving real-time crime analytics in the field. Can you discuss the department’s predictive analytics work and how that informs enforcement strategies going forward?
That’s the second priority, something called the Real-Time Crime Center (as seen in other cities, like St. Louis, Miami, and New York). The idea there is to rapidly look at patterns of crime that are occurring in neighborhoods and geographies, look at historic trends, and shift deployment resources in terms of officers in a relatively short order, within minutes, to address crime trends.
Is it your understanding that the city is full steam ahead on police body cameras? Murray allocated $1.8 million for them in the budget he recently released, and has said there "isn't a conflict” over their implementation.
There are a couple of things that have to occur before they’re deployed. Policy issues need to be resolved, and the citizens police commission and others have raised concerns. Then we have an RFP (request for proposal) to look at potential (body camera) products on the market.
As you noted, the Community Police Commission, as well as the American Civil Liberties Union, have raised privacy concerns about the cameras. They believe new laws regulating their use and data collection should be passed before they’re implemented. Looking at the issue yourself from inside the department, do you believe this data can be collected without the privacy violations they’re worried about?
It’s going to be hard, because of Washington’s public disclosure act, which says that all video needs to be turned over unless it’s the subject of an active criminal investigation. It’s good that there’s a wide-ranging public disclosure law like that, but you have to think about who’s on the video. Washington state law specifically prohibits revealing the identity of juveniles, either through audio or video, for example. We also want to protect victims in complaints who might be caught in video. That means redaction.
How are we going to handle redaction? You think of 200 or so patrol officers every day, that’s a lot of video footage that has to be turned over on demand and redacted first. It’s a bit of a technological challenge.
It takes a ton of time to redact video – three hours for each hour of footage, from what I’ve read. Does this obstacle need to be overcome before this program can be implemented?
I’d say that’s the most significant one. There are not good automated tools.
Your predecessor Greg Russell said, “Transparency to me means you’re being brutally honest with the information (the department has gathered). You’re not trying to sway it one way or the other. The data is the data. You’re just making it available.” How can Seattle police be more transparent and brutally honest about their data?
The city is committed to the White House Police Data initiative. The cities involved with that – I think there’s 26 – are committed to opening up a variety of data to the public, such as use of force complaints. We’re working on a database on when force has been used, more than just complaints. We’re actively working on figuring out how to make that sort of data transparent while protecting privacy of the people involved.
For another example, the city has said... it was exploring ways to be more transparent about the license plate tracking data it gathers on city drivers, and how individual driving patterns are being retained (my Seattle Weekly article on the issue is here). Can information on public surveillance programs like that be more accessible and transparent?
The short answer is probably yes. That’s not an issue I’ve looked into in any detail since I came on to SPD, how that data is collected and stored.
I bring that up to get at the sorts of data the city can get more transparent about. Use of force is one. What are some others?
Certainly another thing we need to consider is citizen complaints against police officers. In most cases, 80 percent or more, the police officer is exonerated in the complaint. How you make that sort of data public is another potential consideration.
What are your thoughts regarding the SPD’s wireless mesh network, which was turned off in 2013 over concerns about its potential surveillance implications?
I don’t know the department’s position on that. I will say that one way to keep the public safe is to have more video cameras in public places, where there’s no expectation of privacy. You can’t have enough police officers to put one on every corner, and we wouldn’t want that anyway. To have video cameras so when a crime occurs, you can rapidly identify and apprehend the perpetrators, seems like a logical step to me. I don’t know what the department’s official position is.
Do you see a value in wireless mesh networks, or do you think cameras are the way to go? Do you think the mesh networks will get turned back on?
I read that article in the Stranger, where they said these mesh networks might be able to track cell phones. That is not something the police should be doing. If that capability exists, that’s a pretty broad capability that could snare citizens as well as criminals. A network should be used for communications within the department, like when an officer does a report in a car, and wants to upload that report rapidly to a database. Or communications to dispatch, or 911 calls. Those are the sort of things the wireless network should be used for.
The U.S. Department of Justice said in a December 2013 report that data gathering from SPD’s IT department, which you now run, has been “error-ridden and inadequate,” and that it was “alarming” that the department couldn’t “manage the risk of unconstitutional conduct” among officers or provide relevant data to investigators and decision makers. How have things improved since then?
I do not know the answer to that question. I know we’ve done a couple of interesting initiatives. One is a crisis intervention team, with specially trained officers who can help the homeless and mentally ill. Rather than arresting them, unless they’re a danger to themselves or others, we’re getting them the help they need in terms of getting them to a homeless shelter. We do collect data on that. SPD officers have had 2,500 interventions in three months. Attorney General (Loretta) Lynch mentioned it when she was in town. We’ve collected data on interventions we’ve had, and how we can better help those people on the street get to services.
But in terms of that report, I don’t know if things have improved or not.
Let’s end broadly. How can your office help the police better protect and serve, and alleviate some of the issues that led the feds to mandate reforms of the department?
We need to put better tech in the hands of individual police officers. Officers on the street need to have department-issued tablets and smartphones to better do their job, to more rapidly access info and file reports. We have a system now that’s very user-un-friendly. We need to redevelop that. Officers should only be spending 20 to 30 minutes on reports, not worrying about clunky old technology.
Another thing we should do with crime incident data is get it out to officers quicker. But we also need to get it to block watch captains and those who run blogs in communities, like Capitol Hill Blog or West Seattle Blog. Because they oftentimes are the first people on the scene, and we’re rapidly exchanging info with bloggers, the more they can participate in keeping the public safe the better. All those things, while knowing we have to respect privacy.