Thanks for the link. CDC's reference 11 seems particularly relevant[1], although the $170 billion measure is not totally attributable to the government: from Table 3, more like $108.4 bil between Medicare ($45.0 bil), Medicaid ($39.6 bil), and other federal programs ($23.8 bil; e.g. Tricare, VA health benefits, etc.). The paper's data set and method of extrapolating portions of healthcare expenditures attributable to smoking does raise an eyebrow at first glance, but I've admittedly not dived deep enough.

EDIT: > A major way (the only way?) to curb tobacco usage is to simply increase taxes on tobacco usage...

If by curb you mean reducing the number of new adopters proportional to overall population, I agree and would support such a measure. As a former smoker and current smokeless tobacco user myself, I'd say those of us who fell for the trap are fucked six ways from Sunday; as much as I'd like to quit, my mind and body are way too dependent on nicotine to function effectively on a day-to-day basis. I'd need roughly 3 months of leave and a lot of mindless distractions to drop the habit. A significant increase in per-unit cost for no apparent reason (like 2-3x overnight) would probably work too...like a proverbial slap to the face that would incite just enough outrage on the merits to serve as an effective distraction, hopefully for long enough to allow physiological dependencies to permanently fade.

[1] https://sci-hub.cc/10.1016/j.amepre.2014.10.012