A CASE OF STATISTICAL DECEIT

A popular aphorism identifies three types of lies—lies, darn lies, and statistics. A variation on that theme declares that figures don’t lie, but liars figure.

Both observations apply to the widely publicized study by Northeastern criminologist James Fox that touted an “alarming” increase in murders by and of black teenagers and attributed the spike to (what else) cuts in federal programs during the Bush years.

Fox’s spin on his numbers was released in late December and was reproduced without critical scrutiny by newspapers throughout the country. (See Dec. 29, North County Times: “Study: Murders among black youths on rise”)

A few weeks earlier Southern Californians had the rare privilege of hearing Mayors Jerry Sanders and Antonio Villaraigosa announce that crime in their cities had been reduced to levels not seen since the 1960s. Then came Fox’s black cloud standing ominously within those silver linings.

According to news accounts of Fox’s study the “rate” of black teenage murders had increased precipitously (40%) from 2000 to 2007. This “rate” of increase, however, was illustrated by providing absolute numbers, not ratios. As anyone of modest intellectual ability knows, official crime statistics are always translated into ratios (usually per 100,000) in order to take into account population shifts. (A numerical increase of 20% in a population that increases 20% produces a statistical straight line, not a 20% increase.)

I found the omission of ratios odd and was pleased to see that “Freakonomics” author and University of Chicago economist, Steven Leavitt, felt the same way. Indeed, Leavitt isolated for analysis the same ratio chart within Fox’s data that I found revealing of a publicity ruse.

That chart, taking demographic changes into account (a 15% increase in the 14-17 year old black population), showed only a modest increase in black teenage murders during the last seven years—a virtual straight line that remained near the bottom of the precipitous statistical slide that began in the mid-1990s. As Leavitt notes, this statistical “blip…doesn’t seem so frightening.” The ratio chart, of course, was absent from Fox’s publicity spin or any newspaper accounts that I saw.

While no increases in murder figures are welcome news, it would be nice to see some skepticism on the part of journalists vis-à-vis academic reports that are funded by government grants and whose bottom line focuses on the need to devote more public money to their area of concern.

The problem of gang violence, which ticked up in San Diego during the first 9 months of 2008 (but declined in L.A.) is a continuing source of concern. But according to County Supervisor Pam Slater-Price, the feds last year targeted $1.3 million for the San Diego region to help address this problem.

What’s sure to boost crime rates among blacks, whites, teenagers, and adults is a policy of early release of criminals that’s been proposed by California lawmakers and may be required by a panel of jurists intent on capping the state’s prison population. That story line, however, doesn’t warm the cockles of most journalistic hearts.