About the only thing that people agree about when it comes to crime statistics is that they’re a lousy way to measure crime. Case in point: Yesterday’s release of a study that is supposed to rank the country’s most dangerous cities, as opposed to the most crime ridden.

The report, the 14th annual "City Crime Rankings: Crime in Metropolitan America" was based on the FBI’s Sept. 24 crime statistics report. The most dangerous city was Detroit; Dallas was ranked 34th. It was 22nd in the same study two years ago and 34th last year.

Sign up for our newsletter

* indicates required

As soon as the report came out, the FBI issued a statement that said: "These rough rankings provide no insight into the numerous variables that mold crime in a particular town, city, county, state or region." The American Society of Criminology said the rankings were "an irresponsible misuse" of crime data. Even the study’s publisher says it is "considered by some in the law enforcement community as controversial."

The problem, like everything else in the cop business, is money. The best crime reporting system would combine the FBI’s Uniform Crime Reporting system and two other measurements which are more accurate but not widely used, several people who study the subject told me when I did a crime story for the magazine a couple of years ago. But no one has yet figured out a way to pay to do that.

The UCR, which the most dangerous cities study uses, has a number of flaws. It covers only about 90 percent of the country and doesn’t include Chicago. It’s inconsistent in reporting standards; what’s a crime in one city may not be one in another. And, by its guidelines, small, rural Florida towns on I-95 with a couple of murders a year, courtesy of someone driving through town on the interstate, would be the most dangerous places in the country.