Each month, when the Pine Bluff Police Department's public information officer du jour trots out local crime statistics, I give a little chuckle. It's not because I am inured or indifferent to the terrible human toll these grim numbers represent. Rather, it's because the agency issuing them and the administration it represents routinely display a complete and utter ignorance about their own statistical records and what their numbers say about crime in our community.
Each month, when the Pine Bluff Police Department’s public information officer du jour trots out local crime statistics, I give a little chuckle. It’s not because I am inured or indifferent to the terrible human toll these grim numbers represent. Rather, it’s because the agency issuing them and the administration it represents routinely display a complete and utter ignorance about their own statistical records and what their numbers say about crime in our community.
It’s tragically amusing in that the monthly report is reminiscent of a scene from the Will Ferrell comedy, Anchorman. Paul Rudd, as weatherman Brian Fantana, attempts to tell Ferrell’s anchorman, Ron Burgundy, why his pungent cologne — Sex Panther — is so alluring, “They’ve done studies, you know. Sixty percent of the time, it works every time.”
Simply put, the agency administration apparently does not know the proper — or at least most comparable — way to report their numbers. As such, their monthly press release represents junk calculations masquerading as meaningful information.
First and foremost, they invariably report raw counts. While they are right to trumpet raw count dips in crime, these numbers have little value as a tool for comparison to other places and times — which is after all the whole point of the exercise.
The better measure would be a rate of crime per unit of population (i.e. homicides per 100,000 population). Unfortunately, the math to produce rates is a little more complicated than merely adding column totals. To begin, you have to know how many people live here, taking into account the average annual departure of approximately 600 people (2000-2010).
Secondly, comparing this year to last year or this month to last month is largely a meaningless enterprise. It sounds like something it just isn’t. Here’s why: If one were to compare the rate of a given crime from July of this year to July of last year, chances are the difference may be within the statistical margin of error. This is to say, any observed fluctuation (good or bad) may be attributable to random chance or measurement error. Measurement in rates can also turn an apparent gain into an actual loss. Unless you do the math, you just can’t tell. None of this even touches analysis for statistically significant trends.
To the extent that meaningful differences exist, the police are to be acknowledged. Of course, there’s a bugaboo here, too.
When a crime rate falls, the police want to take credit for it. Unfortunately, causality and correlation aren’t the same thing. One thing can cause another (i.e. smoking causes lung cancer), or it can merely correlate with another (i.e. smoking is correlated with alcoholism). Just because two things occur together does not mean that one caused the other, even if we wish it were so.
Accordingly, when the police make statements about a particular tactic or strategy “causing” a reduction in crime — more often than not — they can’t actually prove it. This isn’t a Pine Bluff phenomenon. This is a ubiquitous policing (and political) problem.
There are many reasons for this, but the limits of space do not permit a full accounting of them. Suffices to say, the police may be right. They may be wrong. Either way, there’s no rigorously empirical way to tell, especially if you don’t have a handle on your baseline numbers.
Then there’s the angry elephant in the room: public perception. The police do not now and likely have never engaged in a systematic appraisal of public sentiment about crime, safety and the agency itself. Everything is hipshot and anecdote. “People tell me… what I hear on the streets…” is all well and good, but of little probative value.
One thing we do know: The public has engaged in a sustained pattern of voting with its feet. That’s a sure sign that the current regime and its police administration are a consistent failure.
So, when departmental spokesman-of-the-moment, Capt. Greg Shapiro, says “There is no question on these numbers. The perception of crime and the frequency of incidents in the media and the — it leads someone to believe that there is more crime, but statistically speaking, I’m telling you — this time last year we had handled 28,449 incidents and right now today that’s down compared to last year — 27,807 incidents,” he’s right.
One number is smaller than the other. It just may not mean what they want you believe it does. Sad thing is, the police administration doesn’t seem to know the difference… and that’s no laughing matter.
• • •
Matthew Pate, a Pine Bluff native who holds a doctoral degree in criminal justice, is a senior research fellow with the Violence Research Group at the University at Albany. He may be contacted via email@example.com