Article Text

Download PDFPDF

Who stops selling? A systematic analysis of ex-tobacco retailers
  1. Eleonora Feletto1,
  2. Suzan Burton2,
  3. Kelly Williams1,
  4. Rae Fry1,
  5. Clare Sutton3,
  6. Lachlan Bagus1,
  7. Sam Egger1
  1. 1Cancer Council NSW, Woolloomooloo, New South Wales, Australia
  2. 2Western Sydney University, Penrith, New South Wales, Australia
  3. 3Cancer Council VIC, Melbourne, Victoria, Australia
  1. Correspondence to Dr Eleonora Feletto, Cancer Council NSW, P.O. Box 572 Kings Cross, Woolloomooloo, NSW 1340, Australia; eleonoraf{at}


Objective There is evidence that wide distribution of cigarettes contributes to smoking, and multiple commentators have called for a review of tobacco retailing. This study analyses retailers who stop selling cigarettes, why they do so, and discusses the implications for tobacco control.

Method An audit of tobacco retailers in the Australian state of NSW was used to identify retailers who had stopped selling tobacco, and they were then compared with current retailers to determine how many, and what types of outlets stop selling tobacco. Attempts were made to contact and interview all former tobacco retailers identified in three audited regions. In-depth interviews were conducted with 13 ex-tobacco retailers, or 31% of the subset of ex-tobacco retailers.

Results Low-volume outlet types were over-represented as a proportion of retailers exiting the market, and some had resumed selling within 18 months of the audit. Low profits were often cited as a contributor to stopping; however, in all but one case, the decision to stop selling was also influenced by a significant change in business circumstances—either legislative or other business changes.

Conclusions Few retailers stop selling tobacco while continuing in the same business, and those who stop disproportionately represent retailer types with low sales volume. The results suggest that legislative changes provide a window where retailers could be prompted to exit the market.

  • Economics
  • Prevention
  • Public policy
  • Taxation

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


There is increasing evidence that tobacco retailers contribute to smoking prevalence by making cigarettes more accessible and increasing environmental cues to smoke.1–6 This has led to calls by multiple commentators to review tobacco retailing and its regulation, with a focus on decreasing the number of retailers.3 ,7–11 A reduction in retailer numbers could be achieved through a number of mechanisms: a sharp increase in the price of an annual tobacco retailer license,12 a progressive reduction (or ‘sinking lid’) in the number of retailer licenses,13 restrictions on the number and location of retailers,7 incentives for retailers to stop selling tobacco8 and banning tobacco sales at pharmacies where such sales are permitted.3

Despite the theoretical attractiveness of the aforementioned strategies, there is limited evidence supporting their effectiveness,14 with a 2012 review identifying a lack of information about why retailers other than pharmacies voluntarily stop selling tobacco.15 Since that review, two small studies have identified regulatory pressure, declining tobacco sales and health or ethics-related concerns as the main motivators for retailers to stop selling tobacco.14 ,16 The reasons for stopping appear to vary between retailer types, with one study finding that independent US pharmacies were motivated only by the harmful effects of tobacco, in contrast with grocers, who listed more diverse reasons.16 While providing important preliminary information, existing studies have been hampered by the lack of a comprehensive tobacco retailer listing, and thus by an inability to assess how many retailers stop selling, and to what extent they are representative of all tobacco retailers in their legislative area.

A systematic audit of tobacco retailers conducted in New South Wales (NSW) (Australia's most populous state) identified, inter alia, retailers who were no longer selling tobacco,17 providing the opportunity to explore which retailers are most likely to stop selling tobacco, and why. Tobacco is widely available in Australia, and is typically sold in supermarkets, tobacconists, convenience stores, petrol stations, newsagents and licensed premises (ie, outlets licensed to sell alcohol). Australian pharmacies do not sell tobacco,18 unlike those in the USA.14 ,16 NSW retailers are required to notify the NSW Ministry of Health (MoH) if they sell tobacco products, after which they are listed on a register of tobacco retailers with no initial or recurring licensing costs. There is no requirement for ongoing communication with the MoH, unless there is a change in business ownership or address, in which case retailers are theoretically required to update their details. NSW legislation allows for a retailer to be banned from selling tobacco if they are convicted of multiple infringements of tobacco sales regulations,19 though there is no available evidence of this having occurred. The current study used data from an audit of NSW tobacco retailers listed with the MoH to compare current and ex-tobacco retailers by outlet type and, based on follow-up interviews with ex-tobacco retailers, to investigate the main drivers of retailers’ decisions to stop selling tobacco.


Part 1: state-wide audit

A copy of the MoH registry was obtained, and a systematic audit was conducted of a sample of 2279 retailers in 95 randomly selected postcodes from 8 geographic regions of NSW between November 2012 and February 2013. Full details on the methodology are reported elsewhere.17 The audit was conducted by trained volunteers and staff of Cancer Council NSW, using a specially developed audit form. After removing duplicate listings, 1793 retailer records remained in the randomly selected postcodes, and 1743 (97%) were visited. From those 1743 visits, 178 listed retailers (10.2%) were found to no longer be selling tobacco (including listings where the listed business had moved or changed).

Part 2: follow-up with former tobacco retailers

The follow-up study was designed to investigate why retailers who remained in comparable businesses stopped selling tobacco in three of the audit's eight geographic regions (Central and Southern Sydney, Greater Western Sydney and Hunter). Together, the three regions encompass approximately 57% of the state's population.20 Those regions were chosen because of their sizeable representation of the state's population, because they included both urban and outer regional areas, and because face-to-face interviews were feasible. Ethics approval for the study was obtained from the Cancer Council NSW Ethics Committee. The purpose of the study was explained to all participants, assurance of confidentiality given and written consent obtained before each interview. Retailers were not compensated for their participation.

Interviews were conducted from March to June 2014, that is, 13–17 months after the audit. Two researchers were present at all interviews to reduce interviewer bias.21 The interviewers used a semistructured interview guide based on a behavioural reasoning framework, which allowed for the exploration of the context-specific reasons that are critical in decision-making, intention formation and behaviour.22 Themes explored included the background of the retailer and their business, their motives for making the decision and the consequences (if any) of stopping tobacco sales.

The interviews were recorded and transcribed verbatim, with transcriptions verified against the interview recordings. The data were analysed using thematic analysis with NVivo software, to sort participants’ comments into categories derived from the themes of the framework.23 ,24 All interviews were independently reviewed by three researchers, who then met to discuss the various categories and reach a consensus on the emerging issues expressed by the participants. There was little discrepancy in opinion between the researchers, but there was extensive discussion on how to best represent participants’ viewpoints. Representative quotations are reported below using an alphabetic code to preserve interviewee confidentiality, followed by the outlet type of the interviewee.


Part 1: state-wide audit: how many, and who, stopped selling

The audit showed that 10.2% of NSW retailers listed on the register had stopped selling since the registry began in 2009, for a ratio of 8.8 current tobacco retailers to one ex-tobacco retailer. Over the same period, however, new retailers were added to the registry, so the results do not provide information on net change in the number of tobacco retailers, but indicate the number and outlet types of those who stop selling.

Table 1 shows the number and outlet type of current and ex-tobacco retailers, as well as the market share for the larger retailer types, based on other data.25 It shows that retailers who stop selling are disproportionately likely to represent small market share retailer types: licensed premises, which capture only a tiny percentage of the tobacco market share (despite comprising 34.4% of tobacco retailers), made up the majority (53.4%) of retailers who stopped selling. Conversely, supermarkets and tobacconists together capture 71.5% of the market share (although they comprise just 14.1% of the retailers audited) and represented only 3.9% of retailers identified as no longer selling tobacco.

Table 1

Current and former tobacco retailers from the NSW audit and market share in Australia by retailer outlet type

Part 2: why retailers stopped selling, and the limited effect on supply

The sampling base and sample for the follow-up study is outlined in table 2. Of the original 178 retailers identified by the audit, 114 were located in the study areas. One-third (n=38) were no longer operating at the listed premises, (ie, the business had either closed or relocated) and were excluded from the study. An attempt was made to contact the remaining 76 retailers by phone, and an additional retailer was excluded due to insufficient English-language skills. The remaining retailers were asked if the store was selling cigarettes, resulting in a further 16 retailers being excluded as they were selling tobacco (having apparently resumed). Another 14 retailers were excluded because they were under new management, so no one working at the business could discuss the reasons for stopping, and three because they claimed to have never sold cigarettes, leaving a total of 42 retailers eligible for interview.

Table 2

Sampling base and sample

Of the 42 eligible retailers, 17 refused, 1 was unable to be contacted and 11 licensed premises (ie, bars and/or liquor stores) were not pursued for participation to avoid oversampling that outlet type, due to its over-representation (44 out of 76, or 57.9%) among the former tobacco retailers still operating comparable businesses. The retailers who refused most commonly cited lack of time or lack of interest in the study as a reason for not participating. The final interview sample of 13 (or 31% of eligible retailers) comprised 7 licensed premises (53.8%), 4 newsagents (30.8%) and 2 convenience stores (15.4%).

Reasons for stopping selling

Lack of profitability: necessary but rarely sufficient

The interviews revealed a variety of reasons why retailers stopped selling tobacco, but a common theme emerged that tobacco sales had not been very profitable. Twelve of the 13 retailers explicitly or implicitly described the lack of profitability from tobacco sales; one retailer described tobacco sales as ‘not worth the hassle’. Several reported that, over time, their profit from tobacco had steadily fallen. Many commented on the costs of maintaining adequate stock, in some cases requiring a higher insurance premium to sell tobacco, and additional associated retail costs, such as, for licensed premises, paying financial sureties to tobacco vending machine owners. Interviewees repeatedly suggested that these costs meant that tobacco sales resulted in low or, as some suggested, no profit:We had like a big range of cigarettes but then it kind of got to the stage that it was costing us too much…so it wasn't a good profit for us. (Interview M, Newsagent, now combined with pharmacy)There was no money in it…the extra…I had to pay in insurance to insure the place against getting the cigarettes stolen,…it wasn't worth the amount of money that you were making out of cigarettes, and so it was just a pain in the arse…. (I, Convenience store)

One retailer (a convenience store) had stated at the initial phone contact that he no longer sold cigarettes, and at the interview said that ‘I would basically say I don't sell cigarettes now’, but also said that he sold cigarettes ‘very rarely’, when he got a good deal from a sales representative:You know, ‘you buy fifty packs, I'll give you ten packets free’. Well, then it's worth buying from them, yeah.(C, Licensed premises).

However, low profit appeared to be the sole reason for stopping selling tobacco for only one retailer, a convenience store, where the owner was clearly struggling financially.In the last six months, the prices went up, and it didn't make any sense selling because, yeah, it's very expensive to buy as well…and I am hardly covering—you know, including my rent as well… you need to keep a lot of varieties and there is quite a lot of funds involved…I worked out even if I (could) get three times more sales from cigarettes, I would still not be able to cover my rent because (I'm) spending so much money and getting peanuts back. (H, Convenience store)

The final straw

Although low profits from tobacco sales contributed to the decision to stop selling for 12 out of the 13 retailers, for all except the struggling convenience store discussed above, there was a precipitating event, or ‘straw that broke the camel's back’, as one participant called it, which led to the decision. For six retailers, individual business changes precipitated the decision to stop selling; three significantly changed their business mix (one to focusing more on café style offerings, one to co-locate with a pharmacy and one to a location which was judged as too insecure for tobacco storage). Three others had remodelled their premises, and had decided not to incorporate tobacco sales. For one bar, the decision was also encouraged by a tobacconist opening nearby, which was anticipated to decrease tobacco sales. Another bar stopped selling due to the need to renew the contract for a vending machine, and the business owners’ reluctance to provide financial surety for a new machine. Five other retailers (three licensed premises, a newsagent and a convenience store) mentioned changes to tobacco legislation (ie, the requirement to remove tobacco from display, and in bars to have vending machines operated by bar staff), coupled with low profits, as prompting the decision to stop selling. For example the convenience store owner believed that removing tobacco from display had decreased impulse sales, so selling tobacco was no longer worthwhile:The straw that broke the camel's back was no display…basically if (customers) can’t see them they don’t buy them…I've got to fork out five grand, ten grand [i.e. pay $AUD5,000- $AUD10,000 to buy stock] to make nothing. (C, Licensed premises)

One bar manager who mentioned new regulations as contributing to the decision to stop selling tobacco had also had a redesign of the premises. For that retailer, the redesign, legislative changes and low profits from tobacco were discussed as all jointly contributing to the decision not to sell tobacco. Only one ex-retailer did not discuss the low profit from tobacco sales. The retailer had taken over management of a news agency and decided not to continue selling. The decision seemed to have been in part influenced by seeing the graphic health warnings on tobacco packets during early training sessions:The first time we touched a cigarette box—and it's not something which we wanted to touch…,because the picture was, you know, really heart breaking to us…then I thought, it's not something we wanted to do. (D, Newsagent)

Alternative sources of supply

There was no evidence that this group of former tobacco retailers would have any material impact on tobacco supply. Of the 13 retailers, 11 identified at least one alternative tobacco outlet ‘just down the road’, ‘up the road’, ‘across the road’, ‘next door’, ‘on the corner’, or ‘20 metres from the premises’. One retailer did not discuss, and was not asked about, competitors, but was located on a retail strip with another tobacco retailer within 300 m. The most distant location of the closest alternative tobacco retailer was said to be ‘a good kilometre’ away, thus still within close proximity.

Consequences of stopping selling tobacco

Unsurprisingly, given that nearly every former tobacco retailer commented on the low profit from selling tobacco, none expressed major concerns about loss of profits, though some were concerned that they could lose incidental sales. For example, one bar manager speculated that he might be losing business if customers left to buy cigarettes:If a customer wants to buy a packet of cigarettes, then they have to leave the premises, and there's a good chance that they may not come back. I can't give you a figure on how much of a percentage or how much our business would have declined…So obviously there's a (lost) revenue there. (B, Licensed premises)

Several interviewees reported resistance from tobacco companies when they decided to stop selling tobacco. For example, the interviewee who stopped selling tobacco because she felt uncomfortable selling discussed being encouraged to sell by a sales representative:They would offer all these commissions …So they did try to convince us, but it was not something which we wanted to do. (D, Newsagent)


The audit data provides the first comprehensive evidence about how many and which retailers stop selling tobacco. It shows that the decrease in smoking rates (from 15% of NSW residents aged 18 or over in 2010 to 12.2% in 201326) has not been accompanied by an equivalent decrease in the number of tobacco retailers. The audit found that 10.2% of retailers listed on the registry had stopped selling, which might suggest a large decrease in the number of retailers, but about 14% (n=16) of those who were recontacted had resumed selling, perhaps reflecting the type of occasional sales made by the convenience store owner who claimed, despite not selling cigarettes, that he sold them ‘very rarely’. In addition, almost a third of the former tobacco retailers identified appeared to have stopped selling due to business turnover, with the business closing or changing its retail offering, rather than due to any decision to stop selling tobacco. These former tobacco retailers would be at least partly balanced by new retailers who started to sell over the same period, though the nature of the MoH registry meant that it was impossible to determine net change in retailer numbers. The results do show, however, that where inclusion on a tobacco register, in this case a one-time registration system, is at no cost and requires no ongoing action or recurring licensing costs from the retailer, few retailers stop selling and those who do will be at least partly offset by new retailers entering the market.

The audit data also show that former tobacco retailers disproportionately represent low-volume type retailers. Store types with the smallest market shares (licensed premises) were over-represented among retailers leaving the market, while those with the largest market share (supermarkets and tobacconists) were under-represented. While the market share data for an outlet type may not reflect sales at any particular outlet, the follow-up interviews were consistent with exiting retailers having low sales, because every retailer except one discussed low profits from tobacco and suggested that their tobacco sales were not high. When coupled with the reports from all but one retailer that there were other tobacco retailers located very close nearby, it seems unlikely that the decision by these retailers to stop tobacco sales would make any material difference to tobacco supply in their local market—regardless of any potential new tobacco retailers. Reducing the number of tobacco retailers, and thus reducing incidental exposure to tobacco outlets, should, however, reduce the temptation for smokers to buy from those outlets and reduce impulse sales, especially by attempting quitters.2

The follow-up interviews suggest that low profits from tobacco sales, while an important driver for retailers to stop selling tobacco, were almost never a sufficient reason for a business to stop selling. The interviews suggest that those who stop selling predominantly do so in response to low profits and significant changes in business conditions—either legislative changes or business-specific changes such as relocation, repositioning of the business or new ownership. In our sample, legislative changes appeared to have the largest impact on licensed premises due to the new requirement for manual activation of vending machines, coupled with a prohibition of point-of-sale display.

Policy implications

Reducing the number of tobacco retailers has repeatedly been advocated as one way of decreasing the harm due to smoking.7 ,9 ,10 Such a measure received wide public support, except from daily smokers, in one New Zealand study,27 although another study found that New Zealand retailers themselves believe that legislation would be required for them to stop selling tobacco.11 Previous authors have suggested that licensing tobacco retailers, and a moratorium on new licenses, could be used to reduce tobacco retailer density by attrition.8 However, this study suggests that attrition alone will have limited impact on tobacco availability, since only a small number of retailers stop selling tobacco—and restricting new licenses could make retailers more reluctant to stop selling, as holding one of a limited number of licenses could be seen as being more valuable for ongoing tobacco sales and/or for resale of the business.

Licensing of tobacco retailers facilitates spot checks of retailers and, potentially, revocation of the license to sell tobacco if infringements of sales regulations occur.28 However, a one-time licensing or listing scheme for tobacco retailers, as in NSW, provides no mechanism for reducing the number of tobacco retailers. Recurring licensing costs, such as a yearly license fee, would encourage retailers to reconsider whether they want to continue to sell tobacco, and may overcome some of the inertia that appeared to cause retailers in this study to continue to sell until other business changes encouraged them to re-evaluate selling. A study in another Australian state found that the number of tobacco retailers decreased by 23.7% over 2 years following a sharp rise in the license fee (from $A12.90 to $A200).12 Even in the absence of such a large rise, our results suggest that payment of any annual license fee could trigger tobacco retailers to reconsider the profitability of tobacco sales, and encourage low-volume retailers to stop selling. However, we noted other sources of supply close to every former retailer, which suggests that a significant number of retailers would have to leave the market to have any appreciable difference on supply. A fee-based disincentive for selling tobacco will inevitably be most effective in encouraging low volume retailers to leave the market, but closing even low-sales outlets would eliminate impulse sales of tobacco from those outlets, and may therefore be particularly effective at assisting attempting quitters resist an urge to smoke by decreasing the number of stores that sell tobacco.


The results show that despite falling smoking rates and a one-time registration system, if there are no recurring licensing costs imposed for selling tobacco, few retailers will stop selling tobacco, even though those who do cite low profits as a driver. However, legislative change (such as altered display regulations, or an annual license fee) or other business change can, for some tobacco retailers, overcome the apparent inertia which promotes ongoing tobacco sales, thus encouraging some retailers to leave the market.

What this paper adds

  • Previous small-scale studies identified various motivators for retailers voluntarily exiting the tobacco market, but there has not been a systematic analysis of this area.

  • Understanding how many tobacco retailers stop selling, and why, will help show if specific interventions are necessary to reduce the number of tobacco retailers.

  • Despite falling rates of smoking and a one-time registration system, this study shows that few retailers stop selling tobacco, and those who do stop tend to disproportionately represent low market share outlets.

  • Low profits from tobacco appear to be a necessary, but insufficient, reason for retailers to stop selling tobacco, with most stopping selling only after external changes (typically legislative changes) or other altered business conditions prompted them to re-evaluate tobacco sales.


Samantha Ghazi, Kathy Chapman, Scott Walsberger, Anita Tang, Andrew Penman.



  • Twitter Follow Eleonora Feletto at @ellefeletto

  • Contributors EF, SB, KW and RF contributed to the design of the study. EF, KW, CS and LB were involved in conducting the interviews. EF, SB, KW, RF, CS and LB interpreted the qualitative data and SB, KW, RF and SE interpreted the quantitative data. All authors reviewed, contributed to and approved the final manuscript.

  • Competing interests None declared.

  • Ethics approval Cancer Council NSW Ethics Committee.

  • Provenance and peer review Not commissioned; externally peer reviewed.