June, 2014:

http://www.theguardian.com/society/datablog/2014/jun/06/is-smoking-increasing-in-australia

Figures show the long-term trend in smoking is downwards. So is the Australian’s claim to the contrary just smoke and mirrors?

The front page of the Australian claims Labor’s plain packaging laws have failed, as industry research shows a small increase in tobacco sales volume between 2012 and 2013.

So, do the figures back up the Australian’s claim?

Firstly, it’s worth examining the industry data presented in the Australian’s article.

The key number they rely on is an increase in tobacco sales from 21.015bn in 2012 to 21.074bn in 2013*. Labor’s plain packaging laws were fully implemented in December 2012.

While this is a small 0.28% increase in sales year on year, they have failed to account for the increase in the population between 2012 and 2013.

Adjusted for population, tobacco sales per person have actually continued to decrease from 920.4 in 2012 to 906.9 in 2013:

Further, buried at the end of the article is another figure from British American Tobacco Australia spokesman Scott McIntyre:

“From 2008 to 2012 smoking incidence, or the number of people smoking, was declining at an average rate of 3.3% a year,” he said, pointing to Roy Morgan data.

“Since plain packaging was introduced, that decline rate slowed to 1.4%.”

So in their own article they have pointed out that the actual rate of smoking is continuing to decline. That is, the number of people smoking has decreased from 2012 to 2013.

Combined with the industry sales figures, all this suggests is that less people are smoking, but smokers may be buying slightly more tobacco.

These figures are at odds with the latest information from the ABS.

As Stephen Koukoulas points out, the latest ABS figures on household expenditure show as of March 2014, consumption of tobacco is the lowest it has ever been since 1959 when it first began to be recorded. Here’s a graph of said household expenditure, helpfully provided by Ketan Joshi.

The ABS national health survey has not yet covered the plain packaging time period, but again, the long-term trend is downwards.

Unfortunately we are unable to compare the methods of the ABS to the research body, InfoView, that provided the tobacco sales numbers as they’ve declined to release the full dataset.

The Australian also cites figures from the NSW population health survey showing a rise in smoking between 2011 (14.7% of all adults) and 2013 (16.4% of all adults). However between 2012, when plain packaging was introduced, and 2013, the figures for smoking in NSW actually show a small decline from 17.1% down to 16.4%:

Statistically it’s likely that 2012 and 2013 are actually indistinguishable due to the sample size, and the jump from 2011 to 2012 is actually due to a change in methodology, as the NSW health department site says.

“In 2012 mobile phones were included in the survey methods for the first time and this increased the number of younger people and males in the survey sample. Both of these groups have relatively higher smoking rates, leading to a higher overall reported rate of current smoking. The rate for 2013 has stabilised.”

From 2002 to 2011, where the methods change, the long-term trend in smoking is downwards.

*Although the Australian’s graph showed 21.015m and 21.074m for 2012 and 2013 respectively, I’ve been told this should actually be in billions.