Monday, 16 June 2014

Plain packaging's battle of statistics

There have been claims and counterclaims about the effect of plain packaging on cigarette sales in Australia. Or rather there been lots of claims that it has coincided with more cigarettes being sold and one counterclaim that it hasn't. The counterclaim is based on this graph, which only supports plain packaging if you pretend that the policy was implemented in December 2013 instead of December 2012.

The dispute revolves around the Australian Bureau of Statistics' sales figures (table 8). The industry says sales rose by 0.3 per cent, or by 59 million cigarettes. The ABS's sales figures do not tell us how many cigarettes were sold, nor do they tell us how many people were smoking, but they do show us the trend in (legal) tobacco sales. As the graph below shows, there was a long term decline (which goes back decades) which appears to have slowed, at best, in the first year of plain packaging.

The rearguard defence of plain packaging, led by The Kouk, relies on the focusing on the first quarter's sales figures for 2014 which show a dip in chain volume to $3,405 million. This is almost certainly the result of a major tax hike in December 2013 of 12.5 per cent, but The Kouk and others have deluded themselves into thinking that it somehow represents the delayed effect of plain packaging.

This is a pretty desperate excuse. The debate is about what effect plain packaging had in its first year, not what effect a price hike had over a year later. The treasury made much more than it expected from tobacco taxes in 2012/13 and expects to make even more in the years ahead (even taking account of further large rises in tobacco duty that are in the pipeline). And, as we saw last month, there is no evidence of a decline in smoking rates. Indeed, Aussie politicians are complaining that smoking rates are on the rise.

All this is nicely explained by the economist Judith Sloan in The Australian. It's a must read if you want to separate fact from fiction in this debate.

So are there any reasons to doubt claims by The Australian’s Christian Kerr that tobacco sales volumes have increased by 59 million sticks since December 2012 (an increase of 0.3 per cent) and consumption of the cheapest cigarettes has risen significantly more? The short answer is: Kerr 1, the Kouk 0.

There is additional evidence to back up the claim that plain packaging is failing to have an impact. Take the chart on imports of cigarettes, which shows a significant rise since December 2012. And then we have the information from the budget papers which shows actual and projected excise revenue from cigarettes. (Note that excise is levied on a per-stick basis.) The data is not quite complete because then treasurer, Wayne Swan, had something of a meltdown in his last budget and refused to release any figures because of supposed taxpayer confidentiality.

The messages from this chart are twofold: actual excise has exceeded projected excise in every year. And the year in which there is the largest gap between actual and projected excise is 2012-13, which includes six months during which plain packaging was in force. Moreover, the government is expecting to rake in large increases in future excise revenue.

Sensing that the game might be slipping away from them, Nanny Xenophon is calling for a floor price on cigarettes. Gosh, that would be a good idea — loading another ill-conceived intervention on one that doesn’t seem to be working.

But the final point should go to Professor Sinclair Davidson of RMIT University, who has interrogated the data in a systematic way. “I have no doubt that the consumption of cigarettes has risen since plain packaging was introduced; we just can’t be sure whether it is by existing smokers or new smokers.”

Do read it all.


Jonathan Bagley said...

Chris, how would you answer the argument I've heard that the population is increasing and so per capita consumption has if anything gone up slightly?

Jonathan Bagley said...

Sorry, that should be "gone down slightly".

Christopher Snowdon said...

That's true as far as it goes, but we've also got a secular decline going back fifty years so we'd expect to see a year-on-year drop anyway. In other words, the data are noisy.

Presumably the population didn't suddenly expand enormously in 2013? We're looking a quite a short space of time with these quarterly figures.

James In Footscray said...

See Australia's Media Watch yesterday on this very issue:

Do you think the ABC might support public health initiatives?!

Christopher Snowdon said...

Ha! The Kouk recommended I watch it. Gloriously one-sided!