IG report on ATFE violent crime initiative
A few years ago, BATFE got $20 million for a violent crime initiative, and thereafter announced its success. Justice Dept's Inspector General thinks otherwise. (pdf file).
A quick read suggests (1) it was a pretty good plan -- zero in on small violent crime hotspots, and focus on the worst perps in them; (2) headquarters, however, designated the targets with little input from the field agents, who might actually know the local conditions; (3) followthru on the plan details was lacking (sounds as if focusing on the worst of the worst wasn't yielding many arrests, so the offices gradually expanded it to include bad guys in general, not a bad thing but not in accord with the plan) and (3) the success consisted of showing declines in gun homicides city-wide, not just in the targetted areas, and in cases where it was limited to the targetted areas, the number was so small that large percentage variations could be expected (one target area had a 50% drop -- but we're talking about going from four homicides to two, which could be sheer chance).
I suspect it'd be impossible to show success or failure by the test being used (six months before and after, in a small area). If you did put the ten worst perps in jail for ten years, the benefits will be spread over ten years -- and only begin with their arrest, which won't happen the day the program starts.
I agree it's a good idea to zero in on crime hotspots (if such exist--not everything is always guaranteed to be localized, or guaranteed to stay stationary even if it's attracting increased attention.) However, I am a bit mystified as to why it's anything any federal agency should be involved in. The more localized a hot-spot is, the less likely it will be to cross state lines, right?
Posted by: Kirk Parker at September 28, 2006 07:05 PM
So the IG caught the BATFE lying with statistics?
Prepare to Ho-Hum, Ho-Hum.
I grew up in DeeCee, and as a child, remember my dad coming home one evening from the bookstore with a copy of "How to lie with Statistics" by Huff and Geis.
I read it, but being a young teenager, didn't get much out of it except the mantra that in a bureaucracy where statistics are the measure of success, one must EXPECT the math to be messed with.
That was almost 50 years ago, and I doubt if anything has changed in DeeCee yet.
Posted by: Rivrdog at September 29, 2006 09:20 AM