I  found  very  similar  results when  I  used  the  same  econometric  model  to  estimate  the
impact of display bans on smoking prevalence of 15 to 79 year olds.
Average  estimated  impact  of  tobacco  control  measures  on
smoking  prevalence  for  this  wider  age  group.  This  model  also  performs  very  well
explaining more than 90% of the variation in smoking rates between 1987 and 2007.
I again found that display bans were not an effective tobacco control measure in Iceland.
The estimated impact is negative but very small (less than one percentage point) and not
statistically  significant. In  contrast,  these  results  support  the
effectiveness of the  introduction of 30/40 health warning labels. As
mentioned previously, the estimated coefficient of the 30/40 health warning is likely to be
also capturing the impact on smoking prevalence of the introduction of health warning in
the mid 1970’s and early 1980’s in Norway and Iceland and in 1993 in Sweden.33
According  to  my  estimates,  tobacco  price  increases  have  a  negative  and  statistically
significant relationship with smoking prevalence. In particular, I find
that increases  in  the  relative  price  of  tobacco had a negative and significant impact on
smoking  rates.  I  estimate  that  a  10  percentage  point increase  in  the  relative  price  of
tobacco decreased average smoking rates by 1.7 percentage points. That is, assuming a
smoking rate of 20%, an increase in the relative tobacco price index from 110 to 120 is
estimated to reduce smoking rates to 18.3%In summary,  the  empirical evidence indicates that  display bans did  not reduce smoking
prevalence of 15-24 year olds in Iceland. The same is true for 15-79 year olds.