Article Text
Abstract
Objective: To assess the influence of point-of-sale (POS) cigarette displays on unplanned purchases.
Methods: Intercept interviews were conducted with customers observed purchasing cigarettes from retail outlets featuring POS cigarette displays. Measures included intention to purchase cigarettes prior to entering the store, unprompted and prompted salience of POS tobacco displays, urge to buy cigarettes as a result of seeing the POS display, brand switching and support for a ban on POS cigarette displays.
Results: In total, 206 daily smokers aged 18–76 years (90 male, 116 female) were interviewed. Unplanned cigarette purchases were made by 22% of participants. POS displays influenced nearly four times as many unplanned purchases as planned purchases (47% vs 12%, p<0.01). Brand switching was reported among 5% of participants, half of whom were influenced by POS displays. Four times as many smokers were supportive of a ban on POS tobacco displays than unsupportive (49% vs 12%), and 28% agreed that such a ban would make it easier to quit.
Conclusions: POS tobacco displays act as a form of advertising even in the absence of advertising materials. They stimulate unplanned cigarette purchases, play an important role in brand selection and tempt smokers trying to quit. This justifies removing POS tobacco displays from line of sight—something that very few smokers in our sample would object to.
Statistics from Altmetric.com
In Western Australia, tobacco advertising bans have existed for television and radio since 1976 and for newspapers, magazines, billboards, posters and sponsorship since 1992. Limited point-of-sale (POS) advertising at checkout counters remained one of the last avenues for legitimate tobacco advertising in the State until this too was banned in 2006. The display of tobacco products is still permitted, but restricted to one example of each product and a total maximum display area of 1 m2. Similar bans are currently in force in other Australian jurisdictions in Victoria and the Northern Territory.
Although the current Western Australian legislation is restrictive, there is an argument to be made that any display of tobacco products is a form of advertising and hence current restrictions remain inadequate. Indeed, current regulations permit the display of 150 cigarette packets to form “power walls”.1 By definition, partial POS display restrictions such as those in force in Western Australia, Victoria and the Northern Territory are likely to only partially ameliorate possible effects on tobacco purchases. Several other jurisdictions around the world, such as Iceland, Thailand and most provinces in Canada, have gone a step further and ban any visible display of tobacco products on shop floors. In these jurisdictions tobacco products are required to be stored out of sight either within closed draws or under the counter. Similar bans are being planned in New Zealand, Ireland, England and Wales, and in the Australian jurisdictions of New South Wales, Tasmania and the Australian Capital Territory.
Advertising industry data from the US suggests POS displays increase tobacco sales by 12% to 28%.2 3 However, these data are not freely available and are cited from a secondary source,4 and hence are difficult to independently verify. Perhaps more convincing is that the tobacco industry demonstrably fears bans on POS displays and has mounted unsuccessful legal challenges to such restrictions in Iceland and Canada.5
The tobacco industry argues that POS displays merely encourage brand switching among current smokers and do not recruit new smokers, nor increase overall consumption.6 A 2006 telephone survey of Australian smokers suggested smokers very rarely make use of POS displays in deciding which brand to purchase, with only 3.8% claiming sometimes and 1.0% to always use POS displays to decide which brand of cigarettes to purchase.7 By contrast, a 2008 survey suggested 25% of Australian smokers at least sometimes decide to buy cigarettes on impulse as a result of seeing POS tobacco displays, and that 34% of recent quitters and 38% of smokers who had attempted to quit in the past 12 months had to fight the urge to purchase cigarettes when seeing a POS tobacco display.8 These data provide a compelling case that POS tobacco displays increase unplanned purchases among current smokers and increase temptation to those trying to quit. However, both surveys used telephone interviews to gather retrospective data from smokers over the past 12 months.7 8 A limitation of this method is that when smokers are asked to recall previous purchase experiences the limits of autobiographical memory reduce accuracy and introduce biases.9 Memory loss usually occurs very quickly unless an individual makes an explicit attempt to memorise an event. Unusual or vivid experiences can be recalled with much greater ease and accuracy than “everyday” events such as a smoker’s routine purchase of cigarettes.9 To minimise this memory loss, we chose to conduct POS interviews with individuals immediately after they made cigarette purchases. The possibility that smokers might still make inaccurate self-attributions about the influences on their purchasing behaviours meant ideally we would have compared the self-attributions of smokers who made tobacco purchases from retailers with versus without POS displays. However, we were unable to locate any tobacco retailers within Western Australia who voluntarily keep tobacco products out of sight. As such all intercept interviews were conducted outside of retailers with POS tobacco displays.
METHODS
Ethics approval was obtained for this study from the Curtin University Human Research Ethics Committee. A structured questionnaire was devised to assess the role of POS tobacco displays in purchase decisions. Open-ended questions were designed to elicit spontaneous mention of the POS tobacco displays before specific prompting about the displays. The questionnaire protocol was piloted via a convenience sample of 10 smokers observed purchasing cigarettes from a supermarket with POS tobacco display, resulting in minor changes.
To obtain a broad social representation of respondents, supermarkets in two major shopping centres were identified in suburbs from the highest and lowest quartiles of the socioeconomic strata of the metropolitan area of Perth, Western Australia, as determined by the 2006 Socioeconomic Index for Areas (SEIFA).10 Supermarkets were selected because they account for a greater proportion of tobacco sales within Australia (51%) than all other retailer types combined.11
Researchers visited each supermarket and observed individuals purchasing cigarettes at the tobacco counter. All customers observed purchasing tobacco products were immediately approached after leaving the supermarket and asked to participate in a brief survey on “health issues”. Of 344 individuals approached, 206 agreed to participate, representing a response rate of 60%. Each interview was of approximately 5-min duration.
RESULTS
Participants
The 206 participants were 90 men (44%) and 116 women (56%) recruited from supermarkets in both shopping centres in approximately equal numbers (n = 100 vs 106). Participants had an average age of 37 years (range 18–76 years). All were current daily smokers of cigarettes (consumption range 1–30 cigarettes per day). On the occasion of the interview they had purchased an average of 1.2 packets of cigarettes (range 1–7) each containing a median of 25 cigarettes (range 20–50). Participants aged less than 25 years (n = 51) purchased fewer cigarettes (x̅ = 25) than their older counterparts (x̅ = 30) (t(203) = 3.144 p<0.01).
Unplanned purchase
When participants were asked “Did you plan to purchase cigarettes before entering this store?”, 45 (22%) indicated that they had not. A total of 31 participants (15%) made unplanned purchases after being reminded of running out of cigarettes while in store, 13 (6%) said they purchased the cigarettes on a whim and 1 was unable to articulate why she had made the purchase.
A comparison of the demographics of unplanned versus planned purchasers suggested no significant differences by sex or SEIFA but a difference between younger and older adults that approached significance (p = 0.072) (see table 1).
All participants were then asked “What, if anything, did you see in the store that prompted you to purchase those cigarettes?”. In total, 16 participants (8%) spontaneously mentioned the POS display, 1 mentioned “the bottle shop”, another mentioned a sign for the supermarket and another was prompted by seeing another customer purchasing cigarettes. The remainder (n = 187, 91%) indicated that nothing in the store had prompted them.
Participants were then asked “Did the cigarette pack display encourage you in any way to purchase cigarettes in this instance?”. A total of 40 participants (19%), including the 16 that had previously spontaneously mentioned the POS display, responded “yes”. No significant differences were observed by sex, age or SEIFA but the proportion of participants influenced by POS displays was greater among unplanned purchasers compared to planned purchasers (47% vs 12%, χ2(1) = 27.323 p<0.001).
Brand switching
Participants were asked “Are the cigarettes you just purchased your usual brand?”. In all, 11 participants (5%) stated they had not purchased their usual brand. However for only one of these was the purchase unplanned. When asked “What prompted you to try this brand?”, five said they were trying a different brand on the recommendation of another person, including the single unplanned purchaser, and six said they wanted to “try something different” or “felt like a change”. None spontaneously mentioned the POS displays as a prompt to switch brands.
When asked “Did the cigarettes packets on display prompt you to purchase that particular brand?”, all six who wanted to “try something different” or “felt like a change” responded “yes”. When asked to elaborate, all six indicated that they studied the cigarette products on display before selecting the packaging most appealing to them (eg, “I knew I wanted something different so I looked at the display for different options” and “I looked at the display and picked the one I liked the look of”).
Support for ban on POS tobacco displays
Participants were asked “How supportive or unsupportive would you be of legislation requiring shops to remove cigarette packs from display?”. Over four times as many participants were supportive (n = 102, 49%) than unsupportive (n = 24, 12%), but a substantial proportion were ambivalent (n = 80, 39%). When asked, supportive participants were in favour because “whatever helps people quit is worthwhile” (n = 31), “it might help stop impulse buys” (n = 16), “it might stop first time buyers” (n = 29) and “out of sight is out of mind” (n = 14). Those unsupportive were so because they felt they “have a right to see what’s available” (n = 15) and that “it wouldn’t make any difference” (n = 14). No significant differences were observed by sex, age or SEIFA.
All participants were then asked “Do you agree or disagree that removing cigarette packs from view in stores would make it personally easier for you to quit smoking?”. A majority of participants (49%) disagreed but a not insubstantial proportion (28%) agreed. No significant differences were observed by sex, age or SEIFA, but those who had made an unplanned purchase were significantly more likely to agree compared to those who made a planned purchase (59% vs 29%; χ2(1) = 11.455 p<0.01).
DISCUSSION
Just over a fifth of participants reported making an unplanned tobacco purchase on the occasion of interview. Most (68%) did not make an impulse purchase per se, but rather were prompted that they were running low on cigarettes while in store. Although only around half of unplanned purchasers made self-attributions that they were influenced by the POS tobacco display, they suggested such nearly four times more frequently compared to planned purchasers. Hence, POS tobacco displays clearly play a significant role in increasing unplanned, and therefore total, purchases of cigarettes.
What this paper adds
Currently, there is limited data of the effect on adult smokers of point-of-sale tobacco product displays when advertising materials are not present. The best data we have to date is from Wakefield et al,8 who reported that of a telephone survey of Australian smokers, 25% had at least sometimes made an unplanned purchase of cigarettes in the past 12 months as a result of seeing point-of-sale tobacco product displays.
The results from our immediate postpurchase interviews provide the clearest evidence yet that point-of-sale tobacco product displays have a strong influence on unplanned purchases—even if unaccompanied by advertising materials. Policies to remove tobacco products from line of sight are highly likely to impact on unplanned purchases.
In total, 10% of our sample reported being influenced by POS tobacco displays to make unplanned cigarette purchases. This is lower than the 25% of telephone survey respondents who suggested they had been influenced “at least sometimes” in the last 12 months to make an unplanned purchase as a result of seeing a POS tobacco display.8 However this would be expected given our sample was reporting on a single occasion and the previous survey respondents were recalling any instance over the past year. If anything, our results suggest instances of POS tobacco displays influencing unplanned purchases may be more frequent than previously realised. Our 10% figure is also lower than the previously reported 12% to 28% increase in tobacco sales generated by POS tobacco displays in the (less regulated) US market.4 Bearing in mind the potential limitations of the 12% to 28% estimate, it should be stressed that our data were collected only in supermarkets, and in a jurisdiction where tobacco advertising at POS is banned and the size of tobacco displays is heavily restricted. In jurisdictions where POS tobacco advertising and tobacco display sizes are less regulated (such as in the US), the influence of POS displays on unplanned tobacco purchases could well be greater. This matter would be clarified by replicating the present study in jurisdictions with fewer or no POS tobacco display restrictions.
Unplanned purchasers were likely to be of either sex and from all socio-economic strata, but somewhat younger. This is consistent with the near-significant age difference (p = 0.068) reported in the 2008 retrospective telephone survey of the influence of POS displays on unplanned tobacco purchases.8 It is possible that POS tobacco displays have a somewhat greater effect on younger smokers because they are less established in their brand loyalties. Although not strictly evidence within the criteria of Bayesian inferential statistics, these data are at least consistent with previous experimental research highlighting that younger people are particularly impressionable with regards to POS tobacco displays.12 13
Our data confirm POS tobacco displays are an important factor with regards to brand switching. In this respect, POS tobacco displays still clearly act as a form of advertising in Western Australia; a jurisdiction where all forms of tobacco advertising are ostensibly banned. Moreover, a not insubstantial proportion of our smokers (28%) thought the removal of cigarette displays from stores would make it easier for them to quit, replicating the 31% of Australian smokers who have previously claimed the same.8
In addition to previous data that suggest POS tobacco displays increase brand awareness among children and youth and help “normalise” the product,12 13 our data provide further reasons why tobacco products should be removed from sight within stores. Although our sample is modest, the data also suggest that legislating such a ban is likely to result in minimal, if any, voter backlash, with 88% of our sample being supportive, or at least ambivalent, towards a complete ban on POS cigarette displays. The main objection of those few smokers opposed to removing POS tobacco displays is that it would infringe upon their “right” to choose between tobacco brands. As is the case in some Canadian jurisdictions, customers could be provided a catalogue of the available tobacco products upon request and their “right” would therefore be preserved.
Footnotes
Funding: This research was supported by a Curtin Business School Area of Research Excellence (AORE) Grant. The Centre for Behavioural Research in Cancer Control is supported by funding from the Cancer Council WA. OBJC is part-funded by the Health Promotion Foundation of WA (Healthway) through a Tobacco Control Research Fellowship.
Competing interests: None.
Ethics approval: Ethics approval was given by the Curtin University Ethics Committee.