Let’s do a controlled test of plain packaging

Eric Crampton blogs:

There isn’t any real-world evidence on the effects of cigarette plain packaging legislation, mostly because nobody’s really done it yet. What we have are a bunch of surveys of smokers and non-smokers on how cigarette packaging makes them feel, whether they think different designs are more or less likely to encourage them to smoke, and the like. In other words, a bunch of hypothetical musings in low consequence environments.

If we’re stuck having Tariana Turia’s proposed legislation, let’s do some good with it. Set it up as an experiment. Implement plain packaging in part of the country, but not elsewhere. Then see what happens. If it seems successful after a few years, implement it everywhere; if it doesn’t, abandon it. Either way, publish all the results so we have a better handle on what works. So plain packaging in Christchurch but not in Dunedin, in Wellington but not in Rotorua. I’m sure there are plenty of folks who specialize in designing randomised control trials of this sort who’d be able to run things.

This is an excellent idea and one that plain packaging advocates should support. If they are sure it will reduce smoking rates, this is their chance to prove it.

If we apply plain packaging to the whole country at once, we have no way of knowing whether the policy does anything. A careful randomised control trial could tell us something useful.

Some measures you can not test geographically, but plain packaging is one you can. The Government should be pro-science and agree to trial it.