Numerical distinctions, differences, doubt


The regular reporting of crime statistics by law enforcement agencies is commonplace all across the United States. Every state has laws that mandate some sort of reporting. Here in Arkansas, for instance, all law enforcement agencies are legally obliged to report their monthly criminal incident numbers to the state repository, the Arkansas Crime Information Center. ACIC in turn compiles all the agency reports and relays them to the Federal Bureau of Investigation’s Uniform Crime Report program.

The process and standards governing these reports have changed over the years, but the basic premise has remained constant: The people have a right to know certain things about the overall picture of crime in their communities and states. As such, the leadership of various local agencies issue regular reports detailing how many of this crime or that crime occurred over some span of time.

All of this is a good thing. We should be regularly updated by law enforcement leaders. Unfortunately, many law enforcement agencies don’t know how to report their periodic statistics in a way that is meaningful; and we in turn aren’t very sophisticated consumers.

Take for instance the recent Jefferson County Sheriff’s Office report relayed in the Commercial. According to the JCSO 2012 year end summary, Jefferson County sheriff’s deputies responded to 231 more incidents that required a report in 2012 than they did in 2011.

On one level we know what the numbers mean. Supposing we know how to count, JCSO filed 231 more reports in 2012 than in 2011. So what? What do those 231 additional reports mean?

They could mean a lot of things. They might mean that crime has risen a lot. If the population stayed constant, that might be one interpretation, but it didn’t. It went down by approximately 690 people. So, if you look at the incidents as rates instead of raw counts, you get a more nuanced view. Using U.S. Census data along with the JCSO report, we see that the rate of reports in 2011 was 2,188 per 100,000 residents. Whereas, the 2012 rate was 2,487 per 100,000 residents. From this perspective it seems that reports rose around 300 per 100,000 population. So, maybe crime increased even more than we thought.

Here again, is that good or bad? What does it portend for the general climate of safety in Jefferson County? Is the difference meaningful?

To answer this we would have to compare our rates to the rates of some other place. Maybe that place is the U.S. national average. Maybe a better comparison is neighboring counties or the Arkansas state averages.

The point is this: until we compare our “apple” to somebody else’s “apple” we don’t really know what those numbers mean.

Of course — as this kind of thing rarely is — it’s not that simple. You’ll note that the JCSO statistics reflect incident reports. This means (in most cases) somebody called and a deputy came out, investigated and filed a report. Maybe a crime was committed. Maybe it wasn’t. Maybe an arrest was made. Maybe it wasn’t.

Maybe this means that the true picture of crime stayed the same, but Jefferson County residents were just more willing to call in their complaint. The thing is, we just can’t tell.

What we’re left with is a vague, albeit guiding set of numbers that seem to imply there was a little more crime in 2012 than 2011.

By extension this shows us why the monthly reports that police chiefs and sheriffs are typically obliged to give are largely meaningless, disconnected snapshots. Without some kind of broader historical context, the numbers from any given month have little probative value. It’s like being sent a postcard photo of people you don’t know. Maybe you can tell they went to the beach or the mountains, but beyond that the card creates more questions than answers.

This is exactly why we as consumers of these reports need to ask our law enforcement leaders for more nested, more connected and contextualized reports. They need to know that our central questions are much more qualitative: Is it safe where I live? Will I be robbed if I go out into public? While the numbers help us draw conclusions, they are in themselves no meaningful quantity. Even so, we’d like to see them.