Willful blindness at internal affairs and why Austin PD hasn't embraced community policing

Austin PD last week released a massive report by the Matrix consulting group (uploaded here) evaluating the department's staffing in terms of its focus on community policing. (See coverage from the Statesman.)  This morning, Grits took a moment to read through portions of the document. Here are some items which jumped out.

First, Grits found it striking that APD's Internal Affairs unit seemingly intentionally keeps its records in such a way that it cannot perform basic data analysis.

In 2014, 1,116 people contacted the [Office of Police Monitor] to file a complaint. Of those, 582 were actually filed as complaints. The rest were either supervisor referrals (305) or not deemed sufficient to investigate. These statistics are compiled by the OPM. Internal Affairs logs all investigations, but does not have the capability of data retrieval based on type of complaint. The OPM does keep records on complaints filed by racial/ethnic groups while APD Internal Affairs does not because the software being utilized does not allow retrieval of data based on specific queries. As a result, if data is required, a hand search of files must be accomplished.

The consultant framed this as a need to "evaluate the software" IAD uses, but that's the least of the problem. They should also evaluate managers and commanders over them who felt they could effectively oversee the division without basic data-retrieval systems. Sometimes information-gap failures occur because of inadequate software, and sometimes they occur because to generate data would enable uncomfortable analyses to be performed and questions to be asked. Reading between the lines, that's what's been going on at APD internal affairs.

In addition, since the 1990s APD has had a much-touted "early warning system" that's supposed to flag potentially problematic officers for review by Internal Affairs. But we learn from the consultant that it mainly focused on triggers "such as use of sick time, use of force, and internal affairs complaints." That's not an "early warning" system at all! That's post-hoc notification after something has gone wrong. "Early warning" implies that you're trying to identify problems before they arise.

Probably because it was apparent Matrix would make some recommendations on this, now the early warning trigger "is being changed to include the number of drunk arrests, assault on a police officer arrests, and resisting arrests." It would behoove the agency to be a lot more aggressive and sophisticated about their early-warning mechanism. The version used until now sounds simplistic and ham-handed, bordering on worthless if the goal is to reduce officer misconduct or improve community relations.

The main focus of the report was to evaluate community policing in the department. But a cynic might be forgiven for imagining it was created more to justify additional, unpopular and expensive increases in sworn staffing, despite the gaping holes in APD's non-sworn ranks which desperately need filling. (The scope of the consultant's work did not evaluate those functions or take APD's broader management responsibilities into account beyond patrol.)

As noted in the report's introduction, "Full implementation of [the] recommendations would require significant investment ... as a result of adding more officers." And the main recommendation was to hire far more additional street cops than the council's draft budget had envisioned. The consultant recommended adding more than 100 new sworn officers in the next budget. The city manager's version would add 12, plus 21 new civilian positions.

Beyond staffing, the consultant's top findings: "'Community Policing' is not the coherent philosophy and strategy in the Department that it needs to be," and "internal support for community policing in the Department" presently is not "consistent with these strategies" from the mission statement on down. (Breaion King could have told you that!) Isolated internal efforts toward a community policing philosophy are "not supported by strong policies, human resource and management strategies and leadership," they found.

That part about policies and management not supporting community policing is particularly evident in the department's use of force policy, which encourages escalation instead of deescalation. If policies don't reflect community policing philosophies then neither will training, which invariably will be designed to implement the policy. One of the recommendations notes that APD trainers and academy personnel need train-the-trainer training on community policing, so they seemingly have little-to-no in-house expertise on the topic at all. The consultant is describing a management failure, not failings by cops on the beat.

The consultant could see no way to address this situation but to hire more officers. Supposedly because of understaffing, they found that "Patrol resources have limited opportunities to be more proactive – proactivity levels are at an overall level of approximately 22%, which is less than the typical 35% ‐ 45% considered an effective level of patrol service." That number varies by district - downtown, for example, proactivity rates run closer to 70 percent. In many residential areas, by contrast, police are mostly responding to calls.

Not mentioned in the report is the fact that APD patrol spends about 12 percent of its time responding to residential burglar alarms - the most unproductive part of the officers' jobs besides bathroom breaks. The city could free up officer resources by eliminating those corporate subsidies instead of hiring more sworn staff, but that more efficient, less expensive possibility was never broached.

Buried deep in the 239-page report, though, where no one who only read the recommendations would ever see it, was an intriguing observation which deserves to be explored more fully:

Overall, at 116.5 minutes  of  workload  per  call  for  service,  the  average  time required to handle incidents is very high in comparison to other agencies. The average backup unit handling time of 57.7 minutes, in particular, is much higher than the typical norm. Combined with a backup rate of  about 1.1 additional responses per call, each call represents a significant amount of time that must be staffed, in addition to resource needs that must be met in order to achieve a targeted level of service. (p. 83)

Why is Austin's  per-call time spent "very high" compared to other departments? Why does APD assign backup on so many more calls than other departments? And why are backup units spending an hour per call when that's not the practice elsewhere? These questions remain unanswered. The consultant noted the issue in passing but failed to suggest strategies to reduce those times or even provide a framework for how to judge them.

If you need more officers for community policing, there are several ways to get them. E.g., the city council could eliminate superfluous officer duties, like responding to residential burglar alarms. The city could move mental health first response out of APD altogether - an idea being promoted by the Dallas-based Meadows Foundation that could save a lot of officer time, reduce the number of use of force incidents, and connect people with services in the community instead of jail.

APD could free up officer time by reducing the number of low-risk calls to which backup is sent, or the excessive time spent per backup call. They could seek out strategies to reduce the "very high" average time spent (nearly two hours) in response to typical service calls. Or they could reduce or eliminate certain types of arrests for which statutes give officers discretion simply to ticket, like for Class C misdemeanors, low-level marijuana possession, and driving without a license. Those strategies would free up many hours of officer time for community policing, but, giving credence to cynics' imaginations, the only solution the consultant could only come up with was 'hire more cops.'

Finally, I thought this recommendation was a clever idea: "Implement a requirement that trainees complete a neighborhood portfolio that analyzes a specific area of the city, which will not only create a useful database for Community Policing activities, but will establish the foundation for partnerships between the community and the Department." That would be an interesting project.

There's lots more in the document but these were a few things which jumped out at Grits upon first glance. Interested readers should check it out for yourselves and record any additional observations  in the comment section.