January 15, 2009

The truth about statistical 'lies'


Richard Kirk’s recent column ("A case of statistical deceit," Jan. 13, 2009) makes some strong allegations concerning my research as well as my professionalism. Had he contacted me for comment ---- as countless other journalists did following the release of my study ---- before publishing his disparaging remarks, he might have abandoned his one-sided perspective.

Because of this negligence, I must set the record straight.

Relying heavily on a critique by Steven Leavitt (in his Freakonomics blog), Kirk suggested that I lied about statistical data on black homicide. He, like Leavitt, complained that I reported on the recent growth in the number of young black males who killed or were killed, rather than the population-adjusted rate per 100,000.

Had Kirk examined Leavitt’s blog and my report more closely, he would have noticed that the rates and associated trend graph published in Leavitt’s blog were lifted directly from my study. The truth is that my report provided both counts and rates.

More important, it is absolutely wrong to say that the increase in killings was all or even mostly a result of population expansion.

Here are the facts.

From 2002 to 2007, the number of 14- to 17-year-old black males who were murdered increased 56 percent (from 300 to 468) and the number who committed murder jumped 43 percent (from 798 to 1,142), yet the population in this demographic grew only 11.5 percent (1.31 million to 1.46 million).

Shifts over a seven-year period (from 2000 to 2007) were somewhat more modest (an increase of 29 percent in victims, 25 percent in offenders, and only 12.8 percent in population), yet worthy of attention and concern.

Tracking trends in the volume of crime is standard practice in FBI reporting of crime data. Indeed, it is important to reflect on the number of homicides (not just rates) because these represent human lives. Rates per 100,000 often obscure this important dimension. Many people have a difficult time grasping the human toll when hearing that the murder rate by young black males increased from 60.2 per 100,000 in 2002 to 80.6 per 100,000 in 2007. This represents about 350 more perpetrators per year.

Finally, as to my overall "spin," as Kirk characterized it, I emphasized in my report that, despite the increases, the problem is not out of control. I indicated explicitly, backed up with long-term tables and graphs, that the offending rate for black teenage males was considerably higher in 2007 than in 2002, but nowhere near the historic peak during the early 1990s.

Still, why wait until the problem worsens to reinvest in the programs and strategies that were so effective in reducing crime rates before we grew complacent and eased up on our efforts to prevent and control youth violence?

James Alan Fox is the Lipman Professor of Criminal Justice and Professor of Law, Policy and Society at Northeastern University. His recent report on the surge in homicide among young black males can be obtained from his Web site www.jamesalanfox.com