News from FBI is good but incomplete

By Abigail Goldman

Tuesday, Jan. 20, 2009 | 2 a.m.

Chipper is not a good word to describe most FBI press releases. Last week, however, the feds sent out an e-mail with a subject line almost giddy by bureau standards: “SOME GOOD NEWS: Crime Is Declining.”

And why not swoon? Bureau statistics show that violent crime dropped 3.5 percent nationwide in the first six months of 2008 compared with the same period in 2007, while property crime fell 2.5 percent. The FBI’s recently released Preliminary Semiannual Uniform Crime Report — the reason for that punchy e-mail — also revealed a drop in Clark County’s crime. Combine the stats from Metro, Henderson and North Las Vegas police departments and you get a 3.3 percent reduction in violent crime and a dazzling 11.4 percent drop in property crime.

So it’s kind of heartbreaking that a reporter has to twist that news release punch line and deliver the BAD NEWS: That nationwide crime drop comes with caveats.

Since 1929, the FBI has been building a databank of Uniform Crime Reports — basically a tally of what are known as “index crimes” — homicide, rape, robbery, aggravated assault, burglary, larceny-theft, car theft and arson — culled from police agencies and added up for tracking purposes. This is the databank from which the bureau drew to determine the crime drop, and academics like UNLV criminology professor Timothy Hart, the type of guy who seemingly enjoys curling up with a big book of crime statistics, will tell you that the UCR reports are a reliable tally — of what the feds are counting.

There’s that caveat: They’re not counting every crime. Not even close. The UCR is a database of incidents that have been reported to police. So that news release probably should have said “SOME NEWS: Police Not Getting As Many Reports of Crime.”

Of course this could mean there was less crime. To be fair, it probably does. But it could also mean that victims were too scared to file police reports, or worried it would get them deported, now that Metro screens for citizenship in the county jail. Or they figured it wouldn’t do them any good, so why waste time on filing a report about that stolen car stereo?

Ultimately, using the UCR to demonstrate that crime is definitively declining is like using the Dow Jones as a sole indicator of the economy. It’s a barometer, but the weather is too complicated to measure with one instrument.

Police agencies often make this point when UCR reports are used against them. When Morgan Quinto Press releases its annual listing of America’s “Most Dangerous” cities, which the company calculates using UCR data, officials from “dangerous” police departments that rank high — like Metro, which was awarded the No. 4 spot in 2008 — are quick to note that statistics are easily manipulated.

Even the FBI recognizes the danger in drawing too many conclusions from its UCR data, and actually has a disclaimer that reads, in part, “These rough rankings provide no insight into the numerous variables that mold crime in a particular town, city, county, state, or region. Consequently, they lead to simplistic and/or incomplete analyses that often create misleading perceptions adversely affecting communities and their residents.”

The Justice Department introduced a second crime measure in 1973 to get around problems with the UCR. This second system, the National Crime Victimization Survey, is a census used by academics in tandem with the UCR to get a sharper focus on crime trends — for the survey, more than 135,000 Americans self-report their own victimizations. Yes, there’s a whole new set of problems with this methodology too, but the idea is that combined, two studies are — like bad news — better than none at all.