Last week, I pointed to Callison and Kaestner's recent piece suggesting aggregate tobacco price elasticity might be lower than we'd otherwise thought: among adults, smoking wasn't particularly price sensitive.
Youths seem more price sensitive. Gabriel Rossman points to work by Huang and Chaloupka finding a participation elasticity ranging from -0.44 to -0.6 by exploiting a neat quirk in some youth survey data. The Monitoring the Future survey of youth annually surveyed kids, but didn't survey all schools at the same time. And a very large federal tax increase came mid-2009 that resulted in average price increases on the order of 22% for cigarettes. Because the tax increase hadn't hit when some schools had been surveyed, but had when other schools had been surveyed, Huang and Chaloupka are able to get a 2008 baseline over all schools and a 2009 year effect separate from the tax effect by comparing those schools surveyed post-tax with those surveyed pre-tax. So they can compare the 2008-2009 within-school changes for those schools whose students were surveyed before the tax change with those surveyed after the tax change while also adjusting for any differences across the two sets of schools that might have existed in 2008.
It'll be interesting to see what happens with New Zealand's fairly large recent and projected tax increases.