Whatever You Do, Don't Think Of The Children!

It seems that Simon Chapman is a trifle irritated by my pointing out that, despite plain packaging, a long term decline in smoking amongst 12 to 17 year olds in Australia has been replaced by a massive increase of 36%.

As I mentioned the other day, this would equate to the addition of around (corr) 15,000 teen smokers since the previous survey in 2010. Not a great result for a policy designed to stop kids from smoking, I think you'll agree.

So irritated is he that he has taken the unusual step of linking to my little site and trying to pretend this kind of information is irrelevant.

However, a tiny ray of hope remained. A tobacco-loving English blogger noticed that in the 12-17 year age group (the principal target of plain packaging legislation) the percentage of daily smokers actually rose from 2.5 per cent to 3.4 per cent. 

The jubilant blogger took the trouble to construct a bold graph that emphasised this massive uplift. But he failed to tell his readers that for five of 10 data cells that made up the figures, the standard error was more than 50 per cent ("too unreliable for general use") and another two cells with lower standard errors "should be used with caution").

Well, it wasn't me who produced the 'bold graph', but we'll set that aside and put it down to his usual senile frailties in not understanding the concept of blogging very well, shall we? But let's look at his main point, that the increase should be ignored.

It is true that the Australian Institute of Health and Welfare (AIHW) express reservations about many of the figures they have presented, as they explain in the preamble.

Estimates that have relative standard errors (RSE) greater than 50% are marked with ** and those with RSEs of between 25% and 50% are marked with *. Results subject to RSEs of between 25% and 50% should be considered with caution and those with RSE greater than 50% should be considered as unreliable for most practical purposes. Only estimates with RSEs of less than 25% are considered sufficiently reliable for most purposes.

There are a couple of examples of this in the screen grab I took last week.

You'll notice that the 2013 entries for occasional and ex-smokers are qualified with an asterisk, which refers to a note about the dubious RSEs. It is also true that many of the data sets in the background were considered to be less than perfect, but the AIHW’s data tables showed no asterisk next to its data point on daily smoking prevalence for youths between 12 and 17. The institute’s table saying it was at 3.4% with no qualifiers - along with everywhere else in the excel tables where it was similarly cited absent of any asterisk - indicates they were comfortable that the percentage was reliable and, as such, statistically valid.

Or, in the AIHW's own words, "considered sufficiently reliable for most purposes".

It is understandable that the tobacco control industry is finding this so unsettling, because every campaign in every country has focussed solely on glitzy packs and how they apparently attract children. Yet last week's dramatic claims of heroic achievement were conspicuously absent of any discussion of this pretty vital piece of information. Do you think they may have been hoping that no-one would notice?

In reality, this is the only statistic which is relevant in the plain packs debate when it comes to prevalence. Has plain packaging been successful in stopping kids from smoking, or not? It's that simple.

Sadly for Chapman and his fellow tax-spongers, the AIHW survey shows that there has most certainly not been a reduction, and that it's more than arguable that plain packaging has actually made things worse.

For once, it seems that tobacco controllers everywhere are desperately hoping you'll not think of the children. Fancy that!

Nothing to see here, look away ... look anywhere but at the statistics on youth smoking prevalence. Got that? Good.