If you don't have the bullets, you cannot fill a room with bullets.
Limiting bullet purchases will reduce the supply of bullets. Therefore there will be fewer rooms filled with bullets.
Where does the logic fail?
Whilst I'm still open minded about the conclusion you draw here, I'm not sure one
necessarily follows directly from the other - at least not in the way stated.
I agree that if you don't have the bullets, you can't fill a room with bullets. That suggests to me that if one were to plot gun-related deaths vs bullet sales, one of the points would be at or near the origin (probably not exactly at the origin as there would still be some gun deaths from imported bullets).
You don't however know how the relationship varies until you get to the current data point i.e the relationship might now be relatively insensitive to changes in bullet sales. It may even be (I don't know) that the gun deaths actually rise with falling bullet sales, at least for small changes in sales.
What we can say with relative certainty is that if we decrease bullet sales by
enough, then we'll reduce gun deaths. The question, I suppose, is whether we can achieve that "enough" purely by limiting over the counter sales, or whether something altogether more drastic is required.
PE
™ Version 4.2
