St. Louis: No. 1 in Bad Crime Coverage

We've got good news. We've got bad news.
And we've got some really bad news coverage.
First, the good news: In 1998, the city of St. Louis enjoyed a dramatic drop in the incidence of murder and rape, according to FBI figures released last month. There were 26 percent fewer murders and 25 percent fewer rapes than in 1997.

Now, the bad news: The overall crime rate rose last year in St. Louis despite a 7 percent decline nationally. The main culprit: a stunning 15 percent increase in larceny thefts, a category that has seen decreases elsewhere.

And finally, the really bad news coverage: The Post-Dispatch thoroughly misanalyzed, distorted and otherwise fumbled the FBI numbers to achieve a false and damaging, top-of-the-front-page conclusion: that St. Louis has the highest crime rate in America.

Here's what the Post said: "City Ranks No. 1 in Crime Per Capita, FBI Reports."

Here's what the FBI said: "We don't rank or compare cities."
Here's the reality: Citing a "report" that the FBI didn't make, the Post applied weird science to raw crime statistics and landed as far out of context as a real news item on the "Imagine St. Louis" page.

In taking FBI crime totals and lazily dividing them by population numbers to compute a comparison with other cities, the Post ignored a few tiny little details:

* The city's long-frozen borders render it incomparable to almost all other cities in America -- demographically speaking -- because so many others have city limits that include the equivalent of places like University City, Clayton, Richmond Heights and perhaps even farther-outlying suburbs. Suburbs generally have lower crime rates than urban core areas, but they aren't included in "St. Louis" FBI numbers.

* Dividing crime totals by residential population (and then comparing cities) is only valid if one presumes that the cities have the same number of transient visitors. If one city has tens of thousands more people passing through than another -- as tourists, daytime workers, late-night bar-hoppers or whatever -- it's likely to have more incidences of crime. Dividing by the number of residents ignores this factor.

* There's no way to know -- in any direction -- whether police departments are consistent with one another in the reporting of crime data to the FBI. The FBI collects data passively and apparently has no ability or desire to sanction communities that underreport crime.

Not surprisingly, reaction to the Post story was anguished.
"It's not a fair comparison to look at it on a per-capita basis," Police Chief Ron Henderson said. "It's not an accurate representation of what has happened with the St. Louis crime rate for the last several years. Crime has been going down here."

A bit more surprising is that Henderson's reaction wasn't after the fact: It was high in the Post story. It wasn't refuted there at all, nor were the obvious arguments about demographic differences, transient populations and crime-reporting inconsistencies.

Yes, they were all there to read in the Post. In fact, anyone who bothered to read the whole thing probably came away convinced that the headline about ranking No. 1 in crime was a crock.

But anyone who just glanced at the headline without reading the story (which, of course, is the large majority of a daily newspaper's audience) -- or, for that matter, anyone who just happened by a news rack -- would have been left with the distinct impression that St. Louis is America's most crime-ridden city.

And that's no small problem.
"It does affect us in a big way," Bob Bedell, CEO and president of the St. Louis Convention and Visitors Commission, told me Tuesday. "It affects our staff and all the people who help us bring conventions and trade shows to St. Louis. We will spend a lot of time doing damage control."

Added Henderson: "We tried to talk the Post out of writing it this way. I've already had calls from people asking why it's so bad in the city. A lot of folks don't dig into the story past the headlines and see what we're saying about the city getting safer."

Criminologist Scott Decker, from the University of Missouri-St. Louis, didn't criticize the story, but he did point out that the St. Louis metropolitan area ranks "somewhere between 28th and 30th nationally" in its overall crime rate. Decker said he thinks the city did a better job than many others in accurately reporting crimes, especially relatively petty offenses like the theft of auto-license tags, which Henderson claims has inflated the city's larceny-theft numbers.

The FBI didn't exactly rush to the Post's defense.
"I think (the Post story) is inaccurate, because we did not report any ranking of cities," John Gulley, FBI special agent in St. Louis, told me. "There are too many shortcomings involved in comparing one city to another. We don't do it."

Gulley went on to list the same factors -- the demographics, the reporting differences and so on -- that should have been obvious to the Post.

So what does the Post have to say for itself?
Managing editor Arnie Robbins pointed out that the rankings were indeed what the Post found from "FBI data that we had our computer-assisted specialist review," but he added this: "As we point out, there are clearly questions about the accuracy of the data. There are a lot of apples and oranges out there."