Cigarettes and Plain Packs: The Ad Campaign

Blogging here has been light for a little while, and probably will be for a little while longer because of Stuff and Things – but something caught my eye in Sunday’s Indy* that struck me as worth comment.  It was a full-page advert placed by JTI, which describes itself in the small print as “a leading international tobacco company” (and it is).

Anyhoo… the main bit of the ad is a copy of an email, obtained by an FoI request, apparently from the UK Department of Health to the Australian Department of Health and Ageing; it reads as follows (with the highlighting copied from the ad here as closely as I can manage):

Dear [redacted]

I work on the UK government’s tobacco policy team, with [redacted] and you will be aware that the UK government is considering the introduction of plain packaging of tobacco products.

As I’m sure you’re aware, one of the difficulties regarding this is that nobody has done this and therefore, there isn’t any hard evidence to show that it works.   Therefore, I am wondering whether the Australian government drafted any type of impact assessment or cost analysis in which the likely benefits and costs are measured, and if so, whether you would be willing to hare the information with us.

Many thanks,

And then in bold underneath, JTI’s copywriters have added:


Now, the email is dated the 10th of May 2011.  Australia’s plain-packaging law came into effect on the 1st December 2012; as far as I’m aware, Australia was the first country to pass such a law.  So, at the time of writing, there could not have been any hard data about its impact.  There could not possibly be any hard evidence to show that plain packaging works to reduce smoking rates.

The argument of the advert – actually, no: the subtext, since it’s not really an argument – is that the UK government should not introduce plain packaging because of the lack of hard data.  This seems to amount to a claim that governments should not introduce policies without hard data concerning their efficacy.  And there’s something correct about that for the most part.  But hard data are only available in the wake of the introduction of a policy.  If the policy is novel – as it is – then JTI would seem to be committed to the claim that no government should be the first to introduce a policy; which is as much as to say that no government should introduce it at all.

It’s understandable that that’s what they think.  But why not just say so?  Isn’t this ad just rather disingenuously dressing up opposition as something else?

Or have I missed something?


*Saturday’s Indy, with its column-inches devoted to Andrew Wakefield, is worth rather a lot of comment: more than I can offer at the moment.  I’ll point you in the direction of Martin Robbins in the Staggers instead.

  • JTI’s argument (or “subtext”) is of course entirely without merit, and by even engaging with it you are giving them far too much credit. As a friend once advised me in an academic context, if you bear a grudge against Stockport County, everyone will think you are Lincoln City.

    But there is an interesting issue here about evidence and evidence-based policy. What kind of evidence should “count”, and especially what sort of evidence should “count” when one is considering something novel, and therefore such evidence as may exist may not be abundant, and may not be directly relevant?

    My view is that we should insist quite strictly that policies are evidence-based, but should take a broad view about what counts as policy.

    For example, in the plain tobacco packaging example, there may be evidence from marketing studies generally about how consumers in general respond to brightly coloured packaging or other branding strategies from which one might have to extrapolate to form a view about the likely impact on cigarette consumption in particular. It might be possible to undertake simulations of cigarette consumptions by experiments with smokers in a behavioural lab.

    And that of course is where policies often go wrong. Cigarettes may be special, due to their addictive properties or for other reasons. Cigarette companies may find alternative ways of getting consumers to continue to smoke their brands, and that may have an effect on overall consumption levels (which I assume is the policy issue at stake). And the rigour with which much behavioural research is conducted is (despite the best intentions) often not matched in the process of generalisation to real world policy settings.

    That of course is just the nature of social science research. Even in areas where evidence is abundant, it rarely speaks with a single voice. Basing policy on evidence is no guarantee against bad policy. But neither is anything else.