After the Webinar: Race, Ethnicity and Police Deployment. Q&A with Dr. Alexander Weiss

Webinar presenter Dr. Alexander Weiss answered a number of your questions after his presentation, Race, Ethnicity and Police Deployment. Here are just a few of his responses.



Audience Question: Cleveland has an 8-hour day shift and 2 10-hour night and second shift and a third shift. It’s an unusual situation, could you explain a little bit more how the different hours of a shift can affect the difference? 

Dr. Alexander Weiss: The thing about the 10-hour shift is, we started 40 years ago or so. The basic idea was, much of it came from Southern California. It’s expensive to live in Los Angeles, Officers sometimes had to buy a house in Barstow which is a hundred miles away. It would be nice to come to work 4 days a week instead of 5. And they said, if we can have 4 10-hour days a week, that’s still 40 hours, it’s the same. Why don’t we do that? So they tried it. The thing they did not recognize about 10-hour schedules is you have X number of people that we’re providing service for 24 hours a day. Now, you have 3 10-hour shifts in a typical configuration, now you’re providing service for 30 hours a day with the same number you use to provide service for 24. Something had to give right? And almost immediately people will say, wait a minute, there are fewer people out here today than there were yesterday. When you think about it for a minute, this officer’s been working 5 days a week before, now they’re on 4. What you’ve described has a slightly different configuration. When you have 10-hour schedules I think you have to ask yourself 2 questions. And one of the advantages for this is the overlap. There’s a period of time when 2 squads are working together. The question you have to ask yourself is “Does that overlap meet the workload?” My experience generally is that it does not. You have increased double staffing for example between 9 PM and midnight right? When the peak demand time is at 4 o’clock afternoon right? So the way we think about it is “what are we getting for this 10-hour schedule?” and “What’s the loss?” right? There’s always going to be an opportunity cost. That’s the way to think about it. If it matches up with your workload, it’s fine but if you ask many of the officers, they just like it. I think it’s more important that you staff based on what you need rather than what people want.



Audience Question: So what’s your view on the one versus 2 officer scenario if you have actually beat foot patrol teams and they don’t have vehicles. Have you had any research on that? 

Dr. Alexander Weiss: I think its dependent  in part on what you expect them to do. If they’re walking around trying to engage the community and make them feel safer in public places probably 1 might be adequate. If they’re working in high crime areas where communication is a challenge or back up is a challenge, then I think you have to judge that  on its merits. But there are organizations that just kind of default to 2. I worked last year with a Puerto Rican police bureau where no one ever works alone and every single job there’s called the buddy system and there are many positions for which it was unnecessary. But I think foot patrols follow the example where you have to kind of look at the environment and see what are risk factors and kind of make a risk assessment. I think the places that use bicycles have  the same kind of decision. I often see motorcycles working in pairs. Why do they need to do that?



Audience Question: Do all of the agencies you’ve worked with utilize a beat-based system or are there other alternatives that happen successfully? 

Dr. Alexander Weiss: Typically the agencies I’ve worked with use a workload based system. They based the staffing on what the workload looks like per shift. More typically, for example, you have more folks assigned to the afternoon shift than the nightshift because there’s more work. Now they do that in the context of some kind of geographic distribution. Ok well, we have the afternoon shift we have 20 officers assigned, there are 6 posts. Then they assign people to those post. That’s a different approach than saying we have the 6 posts so we need to assign 6 officers.



Audience Question: For priority 0 calls did you take a look at the type of unit dispatch, whether they vary by district and what she was wondering is if we’re calling on tactical teams and rapid response team might be an indicator of understaffing? 

Dr. Alexander Weiss: My understanding is what we captured was the dispatch of the beat car. Not the tactical unit. It might have been that in some of those cases the tactical units agreed to go but the first unit dispatch would always be the beat cars, in my understanding.


Audience Question: I think her point was if the beat cars are not available, they might be calling those tactical units to cover for that beat car and is that an indicator of understaffing.

Dr. Alexander Weiss: Yes. If the beat car in that for high priority call is not available, typically the call will be assigned to the next beat over, that’s how the cad system will do it. I guess what I’m trying to acknowledge is maybe if the case, the beat  cars were pretty busy, it might be the case that the tactical unit simply volunteered to take the call. I saw very little evidence that tactical units were being assigned to these calls as the primary response.



Audience Question: Michael is wondering if you can share the timing, when did you do your staffing for Chicago Police?

Dr. Alexander Weiss: The first one was in 2010  the most recent one was based on data from 2016.



Audience Question: How do you recommend we handle deployment when there is a major incident and officers have to be called from other posts to cover?

Dr. Alexander Weiss: That happens. Depending on the community, it happens more or less frequently. I think one of the things that agencies have to do is think what happens if there’s kind of a catastrophic event. What’s the capacity, how quickly can we get people here? One of the things NYPD does which is quite good is they have these kinds of tactical deployment plans where if something happens in the city, each of the precincts has to kind of offer up a certain number of people to go right away. So that’s kind of in place. They do an amazing job of getting job of getting 500 officers on the scene, seems almost instantaneously. I think there is a risk though of spending too much concern about these unusual events. I’ve worked with agencies who say, “last year, we had this thing and took up all of our resources, the fact is it hadn’t happened again and it hadn’t happen before that for a long time”. So we can’t really staff for these catastrophic events. I think you have to think about it. Think about your partnerships. One of the things about the police community is when something bad happens people step up and the job gets done. I think that speaks well for the industry.



Audience Question: When you start working with an agency, do you end up having to spend much time on establishing common vocabulary such as defining the call for service?

Dr. Alexander Weiss: First thing I do is I have a briefing session for the command staff where we talk about this methodology. I want them to understand very clearly, what we are looking at, how we define it. If there’s a labor organization, I usually have the same talk with them, right afterward because I want them to understand it as well. I think one of the ways you can be successful in this work is to define the methodology upfront. That way at the end, people can say we didn’t like the results but they can’t say you did it differently than when you say you are going to do it. So, I think everybody needs to understand the common vocabulary and the methodology.



Audience Question: Regarding the call for service-based assignment model, do you recommend agencies further consider the distribution of call types and the relative amount of time spent on particular call types in determining appropriate staffing? 

Dr. Alexander Weiss: Yes that’s certainly useful. Calls take varying types. I’ve found that the approach I use, we use kind of average calls, this case, average time by the district is pretty good. One the things that happened after first the analysis in Chicago when I presented my results, the superintendent at the time said, “I’d feel better if we had more data”. So they actually took my years’ worth of data and using the same methodology went back into 3 years’ worth of data. They looked at Calls for Service, they looked at Crime, they looked at Homicide, they looked at Violent Crime and results turn out the same. Turns out this data is fairly stable over time. But the more detail you know about when things are happening and what is happening, that’s good for analysis but also kind of informs your deployment scheme when you really need people.


Click Here to Watch a Recording of Race, Ethnicity and Police Deployment. 


Additional Resources
3 years ago
Reducing Traffic Fatalities in Urban and Rural Areas: Notes from NIJ’s LEADS Program
When people refer to car crashes as accidents, it gives a tone that there isn’t much anyone can do […]
3 years ago
Police-Led Field Experiments: Lessons from the NIJ’s LEADS Scholars Program
When it comes to policing initiatives, it really helps not having to start from scratch. Evidence-ba […]
4 years ago
LEADS Scholars and Agencies: an Interview with Gary Cordner
Evidence-Based Practices have become a significant part of conversation throughout the justice commu […]
4 years ago
Scottsdale’s Restorative Justice Intervention Program Case Study: An Interview with Cassie Johnson
Diversionary programs to keep kids out of the court system are important - but they're only as succe […]