Lies and statistics

Math is not my strongest skill, so when anything more complicated than a balanced checkbook comes along, I reach for a calculator and reach fast.

However, reviewing statistics about law enforcement participation in a gang activity survey conducted by an arm of the Justice Department required no cheat sheet. More than 20,000 agencies were asked to respond. From information the local agencies provided, the Justice Department calculated gang membership at 700,000. OK, a big number and a large problem that needs management, but maybe not all that surprising.

What shocked and awed was that membership number was derived from a low response rate. Out of 20,000 surveyed, 455 replied. For those that like specific numbers, I cite my calculator: more than 97 percent of those asked did not reply. Assuming they did so for fear of unsettling their community – this implies gang activity was kept on the down low. Or the opposite case might be the culprit; there may have been nothing to report. But really, if no news is good news, why did the abstainers not send the survey back with a long column of goose eggs?

The lack of compliance suggests an issue between these agencies and the Justice Department. If 97 percent plus of law enforcement agencies are comfortable not providing data to the Justice Department, is it possible that this sentiment runs in the opposite direction, and that the Justice Department withholds information requested by local agencies?

Whatever it means, based on that turnout rate, I believe the number published is conservative. Maybe the Justice Department will sanction a do-over.