zlacker

In order for the decline in smoking to remain linear, you must convince people who are increasingly less likely to quit. Consistent .5% decline is (weak) evidence for the effectiveness of efforts to combat smoking, not the opposite.

I don't know if that impacts your larger point with regards to nicotine addiction in general, but I think it's worth noting.

>>Stupul+(OP)
It's an interesting point. I would tend to agree with that in the sense that it's a log graph of "difficulty" required.

However I'm not sure how that would be supported without assuming there is some base-rate that would smoke no matter what, as though smoking specifically is a natural inclination, with everyone above the base rate on some log distribution of "ability to convince to stop smoking."