PPMs tend to show lower listening for most stations, but espceially those with a significant Af-Am audience.
I read an article about this problem after the results from the first Philadelphia PPM survey showed a dramatic decrease for the "urban" formats.
There is talk about gerrymandering...I mean developing a "compensating" mechanism to address this effect. They'll probably come up with something that would please the likes of Lani Guinier.
As I have stated before, I feel this system is flawed and contradicts the whole concept of a demographically diverse panel survey.
Case in point: Last night I met some friends at a restaurant here in my neighborhood. The location had WKTU playing over their system, during the 1 1/2 hours we were there we payed no attention to the radio, I couldn't recall a single song played, let alone commercials.
However, if one of us had carried a PPM, they would be "credited" with listening to a station we had no interest in.
In fairness, this system is probably "slightly" more accurate then the diaries that the average schmo wouldn't bother to fill-out 'till the last day, and, in anticipation of this new system stations are implementing background-friendly formats. I suspect that this is why we got some form of oldies back.
Lino