SAARBRÜCKEN, Germany--(BUSINESS WIRE)--Yesterday, two researchers from the IPE Institute for Policy Evaluation Saarland & Department of Economics at Saarland University and from the Department of Economics at the University of Zurich have released a paper entitled “The (Possible) Effect of Plain Packaging on Smoking Prevalence in Australia: A Trend Analysis” which was commissioned by Philip Morris International.
The experts conducted a statistical trend analysis of smoking prevalence among Australians aged 14+ between January 2001 and December 2013, with the objective of determining whether there was evidence for a plain packaging effect on smoking prevalence at any time during the 13 months from December 2012 through December 2013.
The experts found no evidence for a plain packaging effect on smoking prevalence using standard techniques for statistical analysis, in particular requiring a statistical significance level of 5%, which is the standard in applied research. Only when the experts structured their analysis in a way that favoured finding an effect, in particular, by requiring a statistical significance level of 10% only, could they detect “evidence for a very short-lived plain packaging effect on smoking prevalence, namely in December 2012 only (after which smoking prevalence is statistically indistinguishable from its pre-existing trend).”
In conducting their analysis, the experts relied on data collected by and available for purchase from Roy Morgan Research, a well-known Australian research firm. In the past, this same data set has been relied on by public health experts and other researchers.
As explained by Prof. Dr. Ashok Kaul, the lead author of the report:
Using standard analytic techniques that are easy for other researchers to replicate, we found no solid evidence for a plain packaging effect in any month.
Only when using statistical techniques biased in favour of finding a plain packaging effect could we detect weak evidence for a one-time effect on smoking prevalence in December 2012 itself, after which smoking prevalence is statistically indistinguishable from the pre-existing trend.
Based on our analysis, one could, at most, claim an effect on smoking prevalence among the total Australian population in December 2012 only, that is, an effect that lasted no more than one month. From January 2013 on, even very powerful statistical techniques no longer can pick up any change from the pre-existing trend.
The full paper has been published in the University of Zurich Department of Economics Working Paper Series, here: http://www.econ.uzh.ch/static/workingpapers.php.
