Friday, May 29, 2009

Could these be related?

Following up on yesterday’s post, have a look at these two graphs of the past fifteen years. Do you suppose these trends could be related?

Slide2 Slide1

click image for larger version

The Part 1 Crime graph depicts the number of offenses per 1,000 population. Part 1 crimes are Murder and non-negligent homicide, forcible rape, aggravated assault, robbery, burglary, auto theft, and larceny-theft.

10 comments:

Anonymous said...

If they are I am sure they can not be wed in the state of Nebraska. There ought to be a law.

Anonymous said...

It makes me think of "analysis" pieces, written by supposedly brilliant journalists, that simultaneously question rising rates of incarceration and the falling rates of crime, as if it never occurred to them that if a "frequent flier" is locked up, they aren't able to re-offend.

car54 said...

I think your right Chief, the relationship does exists. It's been my experience, for what it's worth, that if at ever opportunity you have contact with you frequent fliers committing minor crimes you “help”them from progressing to more serious crimes. If they know we are aware of them and pay attention to them we create a situation where they feel they have no opportunity to “misbehave”.

I have always thought that the greatest crime deterrent is swift and sure justice. Lacking the connection between committing a crime and being held accountable for that crime seems to encourage some people to continue their criminal behavior.

Anonymous said...

Are economic conditions related to increased criminal arrests or are there just more bad people doing less bad things? I always like your charts Chief.

Tom Casady said...

10:42-

Impossible to say, but that doesn't prevent me from rendering an opinion. I don't think it's either. I think the increase in arrests is attributable to greater number of police officers, and greater productivity by our officers. Sgt. Don Scheinost this morning brought up the idea of "arrest per officer." I'm going to do some work on that and post the results. My guess is that it has increased significantly, so the increased number of arrests is reflecting both more officers and more productive officers. As to the relationship between arrests and crimes, this is complex and open to many interpretations. Personally, I happen to agree with car54: when chronic offenders feel the heat (or are on probation, parole, or incarcerated for short or long periods of time), their offending inevitably slows down for at least a while.

Anonymous said...

I wonder if the citations go up and the rate comes down partly as a function of the increase in the ease of communicating with the police and the numbers of eyes out there in the community. Cell phones have allowed more people to communicate with police via dispatch during in progress calls, and security cameras abound, assisting officers in verifying a crime victim's account of an incident. Arrests that might have gone unmade for lack of evidence might now be more likely. Just a thought. . .

Tom Casady said...

12:29-

You're reading my mind. Or maybe you're reading my blog post from last August (especially the section "Factors nobody seems to talk about.")

This case was a good recent example of the difficulty criminals now encounter in getting away with their capers.

Mr. Wilson said...

Since one graph controls for population and the other doesn't, even this simple exercise of visual comparison isn't especially useful. How 'bout a version of the criminal arrests graph as "criminal arrests per 1,000 population"?

And remember, boys and girls: correlation does not equal causation!

Tom Casady said...

Mr. Wilson:

Okay. And yes, correlation does not equal causation, hence the question mark: could these be related?

Mr. Wilson said...

Sure, you phrased your post as a question. But it's an awfully leading question, don't you think? :-)

I certainly don't think you need to be reminded of the old "correlation vs. causation" warning. I repeat it often primarily because media "analyses" of data are typically so shallow that they have helped lead the general public to expect that if two trends "look similar" on a graph, they must be related. The eyeball test is great for starting discussions and generating research ideas -- hint hint, thesis and dissertation writers! -- but not much more than that. We all need a "be skeptical" reminder now and then.