Plain cigarette packaging has been a great Australian success story. There's now strong evidence that a record decline in smoking rates occurred soon after plain packaging was introduced in Australia in 2012. In fact, these early impacts were greater than expected.
We know the policy worked, but what's less clear is why it worked so well.
Our latest research tested the idea that part of the reason plain packaging has been such a success is because it strips away smoker's sense of identification with fellow smokers of their brand.
We showed that reductions in brand identity following the introduction of plain packaging predicted lower smoking behaviour. These effects were robust even after we controlled for the increased salience of warning labels and smokers' prior addiction levels.
Cigarette brands have social meaning
Before the introduction of plain packaging, many experts predicted plain packs would have limited effects on established smokers, but would deter young people from starting. Others emphasised that any effects for established smokers would occur because plain packs would make smokers attend more to the health warning labels (made larger at the same time as plain packaging was introduced).
There is evidence for both of these explanations. But my colleagues and I argued that, for established smokers, branded packs don't just look pretty and distract from the ugly warning labels, they are also full of social meaning that helps smokers define themselves and their smoking behaviour in a positive light.
In making this claim, we drew on some key ideas from established social psychological theory, in particular the social identity approach.
The first idea is that people's sense of self is powerfully determined not just by their individual characteristics (like their personality) but also by the groups to which they belong ("I'm a Queenslander", "I'm a smoker", "I'm a feminist", "I'm a Holden man").
The second idea is that people usually seek to positively define themselves and often find creative ways to do so.
This is important because defining yourself positively as a smoker has become a hard sell in recent years. While smoking was once seen as a mark of sophistication, smokers are now often stereotyped as unhealthy and dirty, even among smokers themselves.
One way people might respond to this is to identify as a smoker of a particular brand. Doing so deflects the negative connotations of the category "smoker". Tobacco companies get this, and use brand identities to subvert the dirtiness of smoking by appealing to the minty-freshness of a slender, smiling woman; or the rugged outsider status of a cowboy.
You can guess where the story goes from here. We predicted that by stripping away branding, plain packs take away the established smokers' sense of positive brand identity that was helping to maintain their smoking behaviour.
It's important to note here that media advertising for tobacco has been banned in Australia since 1992. In effect this meant that branded packaging was the last avenue for signalling to smokers themselves and to others about what brands mean.
Smokers' changing sense of identity
We ran an online survey of 178 smokers before and after plain packaging was introduced. On both occasions we asked participants about their sense of social identification with fellow smokers of their brand (such as, "I identify with the group of Marlboro smokers"), the stereotypes they linked with smokers of their brand, and their smoking behaviour and quit intentions.
As we predicted, we found positive brand stereotypes, people's brand identities, and smoking behaviour all decreased after the introduction of plain packaging. But for the first time, we also demonstrated these last two things were related: reductions in brand identity predicted people smoking less, attempting to quit and intending to quit in the future.
We found these associations were robust even after we statistically controlled for the increased salience of warning labels, how heavily people smoked to begin with, and other characteristics like age, gender and socio-economic status.
Finally, we predicted and found evidence that people who identified most strongly with their brand before the policy change experienced the sharpest declines in brand identity, and this went on to predict lower smoking behaviour. This last point makes sense: if cigarette brands weren't particularly meaningful to a smoker's sense of self to begin with, then we wouldn't expect plain packaging to have much of an effect for them.
This evidence is an important contribution to understanding why plain packaging works. Our findings support the idea that plain packaging decreases smoking in established smokers because of a loss of brand identity.
Of course a limitation to the study is that, even though its longitudinal, it's not experimental. So it's possible there's something else we're not measuring that's causing the declines in identity and smoking behaviour.
However, by examining changes over time, and controlling for the salience of warning labels, people's age and levels of prior addiction, we took account of the most obvious alternative explanations for our results.
More broadly, our findings are a good example of how the social identity approach is proving to be a powerful tool for understanding health behaviours and for developing novel "social cures" that harness group processes to drive positive behavioural change.