Partners: A Second Look

While reading the comments regarding a recent post on patrol partners, K-9 constable and blogger extraordinaire Sandra Glendinning reminded me that my discussion on two-officer units was not as relevant outside the US--as partners in policing are the norm in many other places.

She also said this:
...Sometimes I relished the opportunity to work alone, if only for a brief change of scenery...

Selecting a partner is serious business - in my career I've only had four, which may seem like a lot, but each partnership ended when one of us was transferred to another section. Each of my partners was like an extension of myself - in one case, the two of us shared a form telepathy (not kidding...). And because I spent more time with my partner than I did my spouse, it was very important to ensure we got along...
Her discussion on partners made me think: If given the choice, would US patrol officers prefer to ride solo or be assigned to a partner?

My personal thoughts were that officers did prefer to work alone, but I started with the research literature and found several studies that examined this issue. In sum, officers assigned to solo units were determined to respond to calls faster, and have the same risk of assault as two officer units.

Also, studies like one of the San Diego Police Department in 1977 (Boydstun et. al), found that one-officer units made more arrests and were, in general, more productive than two-officer cars. As a result of such studies and the cost savings involved, one-officer units are more commonly used in the US.

Two studies addressed officer perceptions of unit assignment. The same San Diego study determined that officer's felt safer and more productive in two-officer units in contrast to the research's actual findings.

Another study took a different approach rather than evaluate the question in terms of efficiency (response time, arrests, etc.). In 2003, Alejandro de Carmen and Lori Guevara studied a small sample of officers in a metropolitan police department in Texas. They measured the perceptions of officers as to one and two officer units. In contrast to the San Diego study of 1977, they found that officers believed, in general, that they would perform the same no matter the assignment. Also, the officers agreed that two-person units were necessary in some instances, but largely preferred being assigned to a solo unit.

And, what would a blog positional post be without a completely unscientific poll? I conducted such a poll over at and was pleased that 66 respondents took time to vote. Thirty-seven respondents favored riding solo, 24 wanted to have a partner, and 5 did not have a preference. I was surprised that the results were not more in favor of one-officer units, but I think the representatives from larger departments, where partners are common, were well represented in the results.

So, that is convincing, right? I have some limited academic studies in combination with an unscientific poll that supports my assertion that US officers prefer being assigned to solo units. In contrast, to the literature, I think the officer arguments in favor of solo units are primarily unrelated to police work (productivity and safety), and have more to do with personal preference. I believe that officers would prefer the freedom that riding alone provides, rather than having to entertain someone else for a shift on a regular basis.

I think the time difference in the San Diego and Texas studies supports this thinking. In 1977, partner policing was commonly used and shifting to solo units would have been a drastic change to the norm. In contrast, Texas officers in 2003 were used to riding solo and having a partner represented, for the most part, the drastic change.

With all of my efforts in exploring officer preferences for assignment, I think I have simplified reinforced one long-standing reality in policing: that officers are resistant to change (Argh...) and that such planned alterations should be backed with evidence and given time before any of the expected results can be obtained.


Sandra G. said...

Very good post, Slamdunk. Thank you for mentioning me (again).

I read the stats you presented with a great deal of interest. One officer units being more productive than two member units? Taken from a statistical point of view, maybe - in our department, the one officer cars are the 'report takers', producing a huge amount of actual files. Why? Because the calls they go to are not in progress, the suspect is long gone, and the report needs to be taken.

The more complicated calls, such as any in-progress call (domestics, assaults, robberies) or complicated investigative calls (sexual assaults, frauds) are usually assigned to two officer units. As a result, two officer units spend more time on one call than single unit officers spend on a multitude of 'report only' calls.

The reports and the studies you found are very interesting - I'll be taking a closer look.

As for how the VPD does business? COMSTAT is a huge driving force behind our style of policing...just google it and you will see what I mean.



Slamdunk said...

Thanks for the feedback Sandra, and unfortunately (for you) our discussion the other day about local or national police recruiting may feature your thoughts in another post. Hopefully, you will not begin charging for your comments.

You offer a good point on the one vs. two officer studies. Researchers would need to be very careful in simply comparing productivity of the two types of staffing since agencies that use both would send two officer units to more hot calls. As a result, simply looking at aggregate data would make it look like two officer cars make more arrests. One of the most well-known studies on the topic using an experimental approach is the Kansas City Policing Experiment.

The annual FBI stats on US officers injured includes info on two officer units vs. one officer units, but I think they have the same problem in predicting which is actually safer--two officer cars in urban areas are usually staffed like that for a reason.

If you are interested, I have most if not all of the articles in PDF format and can send them to you--just let me know.

In the interest of potentially using "Slamdunk is an idiot" emails for potential future blog posts, I will be posting an email address to the "About Me" section of the blog tonight.

Have a good weekend.