Tobacco.org

• The effectiveness of display bans: the case of Iceland (PDF) 

Executive summary

In this brief report I summarise the findings of my empirical investigation of the expected impact of display bans on tobacco consumption.

Display bans are regulations that prohibit the visual display of tobacco products within the point of sale. They are the most restrictive of all point-of-sale regulations, which include limitations on height and visibility of displays, prohibition of self-service displays, and restrictions on logos, banners, and window posters.

Whether display bans have an impact on tobacco consumption is an empirical question.

Also, the likely magnitude of that impact can only be estimated using empirical techniques. This is why this paper considers the case of Iceland, the only country in Europe to have introduced display bans before 2009.

Iceland introduced display bans in August 2001.1 . . .

The results I have just described are robust. I re-estimated the regression model including additional control variables (in particular health expenditure and different measures for tobacco prices) and found that the results remained qualitatively unchanged. Similarly, I re-estimated the model using data for all European countries with publicly available smoking incidence data.9 This implied extending the set of benchmark countries. I continued to find that point of sale regulation had no statistically significant impact on Icelandic smoking prevalence.

In summary, my analysis of the data shows that certain tobacco control measures reduce smoking prevalence and clearly establishes the impact of tobacco prices on consumption. But the data does not support the claim that a display ban is likely to cause a reduction in smoking prevalence. To the extent that there is a relationship, it is very small and the evidence shows it may be purely due to random chance. My statistical analysis therefore confirms the conclusions suggested by a simple inspection of Figure 1 and Figure 2 above: there is no evidence that the display ban in Iceland caused a reduction in smoking prevalence. . . .

I have been asked by Philip Morris International (PMI) to analyse the expected impact of display bans on tobacco consumption. . . .

Conclusions

My empirical investigation of the impact of display bans on tobacco consumption in Iceland shows that the introduction of this regulatory measure had no statistically significant effect on smoking prevalence in that country. This is true for all age groups for which data was available. Therefore, I found no support for the claim that a display ban is likely to cause a reduction in smoking prevalence.

In contrast, tobacco price increases, driven mainly by increases in taxes, had a negative and statistically significant impact on smoking prevalence. Furthermore, other tobacco control measures, like bans on smoking in public areas and health warnings on cigarette packages were effective tobacco control measures, as they had a negative and statistically significant effect on smoking prevalence.

In other words, the experience in Iceland does not suggest that a display ban would reduce smoking prevalence, and instead shows that other measures may be more effective in controlling tobacco consumption.

Jump to full article »

I found no support for the claim that a display ban is likely to cause a reduction in smoking prevalence. In contrast, tobacco price increases, driven mainly by increases in taxes, had a negative and statistically significant impact on smoking prevalence. Furthermore, other tobacco control measures, like bans on smoking in public areas and health warnings on cigarette packages were effective tobacco control measures, as they had a negative and statistically significant effect on smoking prevalence.
Display bans must be tremendously threatening to Philip Morris if its own report establishes taxes, warning labels and smoking bans as effective tobacco control measures. It seems PM would prefer ANYTHING but a display ban.