This is the principle of "student" edition software.
Has this been demonstrated empirically, in the case of music and video? I see this argued both ways, and I don't know which is true.
I doubt the number of youngsters who pay for music once they have the means is huge. Would you pay for the Spice Girls song you downloaded 4 or 5 years ago and havent listened to for the last 2 yrs. No, probably not - but you might have had 2-3yrs enjoyment out of it... where do you draw the line?
I dont see any justification for downloading music you cant afford when there are free sites to listen to it on.... it's not as if that lack of funds is depriving said person of the ability ot hear and enjoy the song. It perhaps does inhibit how and in what context they can listen to the song (i.e. no ipods etc.).
In any case, there is no cost to the copyright owner if someone who can't afford it pirates it. There is at least a potential benefit.