Tuesday, October 07, 2003
Tobacco Advertising Bans
Do bans on advertising curtail smoking? It turns out that it is pretty hard to answer that question
by examining international data on changes in smoking prevalence, before and after bans have been
introduced. Some studies find that advertising bans do reduce smoking, some find that they do not,
and some find that it depends on the breadth of the ban. A recent contribution by Jon Nelson,
an economist at Penn State, tries to correct for some of the problems faced by previous studies.
In particular, one complicating factor is that advertising bans do not just pop up at random. Rather,
bans are likely to be adopted precisely at times when pro-smoking sentiment is in decline. Falls in
smoking after an ad ban, then, may not be the result of the ban; the ban and the decline might both
be products of increased anti-smoking sentiment.
Nelson tries to account for this "endogeneity" of advertising bans, using data from 20 OECD countries
for the years 1970-1995. He concludes that the results of previous studies finding that advertising
bans reduce cigarette consumption are not robust to the correction for ban endogeneity.
Of course, one wouldn't want to base tobacco policy on a single econometric study, as the next study
down the pike, employing different data or a different estimation procedure (or both) might come to
the opposite conclusion. One thing that is clear is that smoking prevalence has declined very
significantly in OECD countries in recent decades. In the US, Nelson reports, "male smoking prevalence
declined from about 44% of the population in 1970 to 32% in 1985 and 26% in 1995..." Per capita
consumption of cigarettes in the US was highest in 1963, the year before the renown Surgeon
Nelson's study includes in a footnote one amazing factoid. A complicating factor in using cigarette
sales to determine the extent of cigarette smoking is that some people roll their own cigarettes.
Surely roll-your-own cigarettes are but a small part of overall consumption, no? Well, it depends on
what country you are considering. In most OECD nations, less than 10% of cigarettes are
hand-rolled. But in 1995, in both the Netherlands and Norway, nearly half (46%) of the cigarettes
were hand rolled! No word if this finding applies to all countries with names starting with "N".