In December 2008, the health secretary for England, Alan Johnson, announced that the government would proceed with legislation banning the display of tobacco products in shops. According to Johnson, the logic was impeccable, as young people 'see the point of display and as a result of seeing it, it encourages them to take up smoking'. Since then, the legislation, which is incorporated in the bill providing a constitution for the National Health Service, has gradually made its way through the House of Lords.
To the jaded, this latest piece of anti-tobacco legislation might appear to be just another instance of a quarter-century of bad tobacco policy, in which successive governments have attempted to address a particularly serious and complex public health problem through a series of gimmicky soundbite measures that do nothing to address the underlying reasons why young people smoke.
But this legislation is, in fact, something much worse. It is the first time that a government has legislated on tobacco not only without any compelling evidence that what it is proposing will work, but actually in opposition to what its own experts have demonstrated to be the case.
According to the advocates of a tobacco display ban, seeing cigarettes displayed in shops is one of the major reasons why children in the UK smoke. The trouble is that, as the legislation has progressed through the Lords, it has become increasingly apparent that this is not the case.
A major piece of evidence used by the government to support the claims behind the legislation is a study by a team of Australian researchers led by Melanie Wakefield. Wakefield exposed several hundred school children to one of three digitally manipulated photos of convenience shop points of sale. One photo showed cigarette advertising and tobacco packs, one photo showed only tobacco packs, and one photo showed no tobacco packs.
The results are striking, as they refute each of the key claims about the supposed effect of tobacco displays on children's intentions to smoke.
First, there were no statistically significant differences in the estimates of how many of their peers smoke between the children who saw the convenience shop with no cigarettes and those who saw the shop with cigarettes.
Second, there were no statistically significant differences in approval of smoking between those who saw the shop with no cigarettes and those who saw the shop with cigarettes.
Third, there were no significant differences in assigning favourable attributes to children who smoke between those who saw the convenience shops with no cigarettes and those who saw the shop with cigarettes.
Fourth, there were no statistically significant differences in their perceptions of the dangers of smoking between children viewing the shop with no cigarettes and those seeing the shop with cigarettes.
Finally, and most importantly, there were no statistically significant differences in the children's intention to smoke between those who saw the shop with no cigarettes and those who saw the shop with cigarettes.
This research is not produced by the tobacco industry. Wakefield works at Cancer Council Victoria. This is the government's best evidence supporting its claim that seeing tobacco displays in shops leads children to smoke.
But matters get even worse when one looks at the other evidence presented by the government to the Lords. For example, according to the government, the Department of Health, in its consultation on the future of tobacco control conducted last year, found that one of the reasons that tobacco displays lead to children smoking was that they were effectively tobacco advertising.
Here, again, the government's own evidence suggests that this is not the case.
First, in the government's own expert study, Wakefield, as well as others, distinguishes between tobacco advertising and marketing and tobacco product displays. If tobacco displays are not tobacco advertising, using claims about tobacco advertising to support banning tobacco displays fails to support the case for tobacco displays. Second, even if tobacco displays were tobacco advertising, the government's own expert evidence suggests that such advertising is not the cause of young people smoking.
For instance, the Department of Health in its consultation on the future of tobacco control lists four predictors of 'regular smoking among young people': age and gender; home environment; drug use and drinking alcohol; and truancy and exclusion from school. Tobacco advertising is not included.
Again, two major studies commissioned by the Department of Health itself contradict the claim that tobacco advertising, let alone tobacco displays, lead young people to start smoking.
The Goddard Report, a longitudinal study of the causes of youth smoking and one of the largest studies of adolescent smoking conducted in the UK, found that becoming a smoker was associated with seven risk factors: being a girl, having brothers or sisters who smoke, having parents who smoke, living with a single parent, having relatively less negative views about smoking, not intending to stay on in full-time education after 16, and thinking that one might be a smoker in the future.
Becoming a smoker was not statistically associated with being aware of tobacco advertising or being able to name more brands of cigarettes. Indeed, Goddard found that children overwhelmingly did not like tobacco advertisements.
The second study, by Clive Smee, carried out an extensive examination of advertising and adolescent smoking in the UK from 1960-1987. It found that advertising in any form did not have a statistically significant effect.
Both of these studies were done when tobacco advertising was virtually unrestricted in the UK. If full-flown tobacco advertising did not lead to adolescent smoking, it is unlikely that seeing cigarette packages in the shops will do so.
In a further effort to bolster the quickly unravelling case for display bans, the government has circulated a study by Gerard Hastings and colleagues from the Centre for Tobacco Control Research at the University of Stirling. Hastings claims, most recently in a letter to The Times (London) this week, that an increased awareness of tobacco brands, supposedly from tobacco displays, increases young people's susceptibility to becoming a smoker.
Yet Hastings claims are refuted, first by the fact that, according to a US Department of Health and Human Services study, interest in smoking and intention to smoke drives brand awareness rather than the other way around. Young people interested in smoking are interested in tobacco brands. Interest in tobacco brands does not lead to an interest in smoking. Clearly, Professor Hastings has confused the sequence.
Second, Hastings' claims are further refuted by the British experience in which there has been a significant decline in awareness of tobacco brands. Yet, according to the most recent statistics for England, there has been no decline in youth smoking. Indeed, there has been an increase in smoking among adolescent girls. If Hastings' claim about awareness of tobacco brands driving youth smoking were true, then one should expect a sharp decline in awareness to bring about a corresponding decline in smoking.
In a final attempt to make its case that display bans work, the government has argued in the Lords that banning tobacco displays in other countries has resulted in fewer young people smoking. But, once again, the government's own evidence shows that this is not the case. Studies from both Canada and Iceland fail to show that banning tobacco displays has had a statistically significant effect on youth smoking.
In Canada, for instance, there is no obvious difference in youth prevalence between provinces with and without tobacco display bans. Moreover, according to the UK government's own data, the highest prevalence is found in Saskatchewan, the first province to ban tobacco displays.
In Iceland, in 2001 when the display ban came into effect, smoking prevalence among adolescents rose by 3.1 per cent. During 2002, the first year after the display ban, adolescent smoking prevalence was the highest it had been for five years.
The evidence presented by the government in favour of banning shop displays of tobacco, far from showing why such a measure is justified, actually shows why this legislation will not work. The bottom line is that a tobacco display ban will not prevent a single British adolescent from beginning to smoke.