The robust public-health case for tobacco harm reduction
Kulik and Glantz based their conclusion, in part, on an analysis of public survey data from the Tobacco Use Supplements of the Current Population Survey. They had information on the percentage of smokers (prevalence); the percentage of smokers who made a quit attempt in the past 12 months; the proportion of former smokers among ever smokers (also called the quit ratio); and daily cigarette consumption (cigarettes per day, or CPD) for each state and for several survey years from 1992 to 2011. Using linear regression, they found that a 1 percent decline in smoking prevalence is associated with a 0.6 percent increase in quit attempts, a 1.1 percent increase in the quit ratio, and a reduction in consumption of 0.3 cigarettes per day.
The analysis was seriously flawed, as the authors failed to consider other factors that may significantly affect smoking. For example, Kulik and Glantz should have considered data on the percentage of smokers who faced workplace or home smoking bans – information that was available in the survey datasets. The effect of state cigarette excise taxes should have been weighed. Additional factors, such as differences in smoking norms and anti-smoking sentiments in the various states, are commonly analyzed through the use of a standard fixed effects variable. They did none of this.
My research group has recreated the Kulik and Glantz analysis and taken into account the missing variables. Our results have now been published in the journal Addiction. We found that the “results are not robust…The inclusion of state fixed effects and state-level policies led to a large drop in the coefficients…and became statistically non-significant…the omitted state-level characteristics are most likely responsible for…[Kulik and Glantz]’s results.”
We also note:
One further point needs to be made. [Kulik and Glantz] claim that their study bears on the question of whether smoke-free tobacco products have a contribution to make to tobacco control. They claim that much of the argument for smokeless tobacco and e-cigarettes is dependent on the assumption that the smoking population is hardening. This is not the case. The argument for these considerably safer products e-cigarettes is simply that they may be short-term aids to cessation or permanent substitutes for tobacco cigarettes. In a population such as that in the US, that argument is relevant as long as there are substantial numbers of smokers who will use them.
For the 39 million smokers in the United States, there is no public-health basis to withhold either safer cigarette substitutes or the potentially life-saving facts about such products.