Objective We aimed to assess change in the availability of illicit tobacco in small mixed business retail outlets following the December 2012 introduction of plain packaging in Australia.
Methods 303 small retail outlets were visited in June and September 2012 (baseline months), and in December 2012 and February, April and July 2013. Fieldworkers requested a particular low-cost brand of cigarettes and then pressed the retailer for an ‘even cheaper’ brand. The cheapest pack of cigarettes offered was purchased and later examined to assess any divergence from prescribed Australian packaging regulations. The price paid was compared with tax liability and recommended retail price for the particular brand and pack size. In a sub-set of 179 stores, fieldworkers then asked the retailer about availability of unbranded (chop-chop) tobacco.
Results Thirteen (2.2%) of 598 packs purchased pre-plain packaging were either non-compliant with Australian health warnings and/or suspiciously priced. Four packs (1.3%) of 297 met either or both criteria in the December implementation month, and five (0.6%) of 878 did so in the three collection months following implementation. Chop-chop was offered upon enquiry on 0.6% (n=2) of 338 occasions prior to implementation, 0.6% (n=1) of 170 occasions in the December 2012 implementation month, and 0.6% (n=3) of 514 occasions postimplementation. The likelihood of a ‘positive’ response (either an offer to sell or information about where unbranded tobacco may be purchased) did not differ across preimplementation, during-implementation and postimplementation waves.
Conclusions Overall, packs judged likely to be illicit were sold in response to requests for cheapest available packs on fewer than one percent of occasions. Offers to sell unbranded tobacco were rare. No change in availability of illicit tobacco was observed following implementation of plain packaging.
- Illegal tobacco products
- Advertising and Promotion
- Packaging and Labelling
Statistics from Altmetric.com
From 1 December 2012, all tobacco sold in Australia was required to be packaged in plain, drab, dark brown packs, with 75% front-of-pack graphic and 90% textual and graphic back-of-pack health warnings. Display of the brand name and variant is limited to a standardised font type and size.1 The introduction of Australia's plain packaging law was preceded by a vigorous campaign against the policy by the tobacco industry which argued, among other things, that it would lead to an influx of cheap illicit tobacco2 ,3 that could in turn lead to an increase in tobacco consumption, deprive the Government of revenue and benefit criminal gangs.2 ,4 ,5
Two categories of illicit tobacco have been identified in Australia. The first is unbranded tobacco sold either as loose tobacco or as prerolled cigarettes usually in plastic bags without any sort of labelling. This form of illicit tobacco is generally known in Australia as ‘chop-chop’.6 The second category is branded cigarettes (and much less commonly, branded roll-your-own tobacco) on which customs/excise duty should have been paid but has not; that is, contraband tobacco products.
Consumers can readily report whether or not their tobacco purchases come out of an unmarked plastic bag or box, and so, over the past 12 years, Australia's nationally representative National Drug Strategy Household Survey has included questions aimed at quantifying self-reported use of unbranded tobacco.7–10 The 2010 survey of over 25 000 respondents found that 4.9% of smokers currently smoke unbranded (chop-chop) tobacco, with just 1.5% of all smokers smoking it ‘half the time or more’.10 The surveys have not asked users from where they purchase unbranded tobacco; nor have they attempted to ask about use of branded contraband tobacco products which are much more difficult to define in self-report surveys.
In places such as the UK where very cheap cigarettes are common, many people understand terms such as ‘fag house’ and ‘cigarette man’, and know that contraband products are implied by the term ‘cheap cigarettes’.11 ,12 However, in jurisdictions such as Australia, where use of illicit tobacco is thought to be much less common, consumers lack a shared vocabulary. In addition to definitional ambiguity and the likelihood of double counting, industry-funded consumer surveys that have used terms such as ‘counterfeit’ and ‘contraband’ have been criticized13–15 for relying on highly biased samples.
Studies estimating prevalence of illicit cigarettes based on surveys of discarded packs are also problematic. Packs discarded in public rubbish receptacles or as litter are unlikely to be representative of all the packs discarded by smokers across the country, most of which would find their way into domestic and work-based rubbish systems. The places from which packs are surveyed are rarely fully described. Serious questions can be raised about assumptions underlying calculations of the percentage of foreign packs observed that are illicit as opposed to merely purchased overseas and carried or mailed in within duty-free limits or with duty paid.16
Little information is available in Australia about sources of purchase of branded contraband cigarettes. Some would be brought into Australia by returning travellers who exceed personal import allowance for tobacco products and fail to pay customs duty. Nothing is known at present about the extent of availability of contraband products from retail outlets in Australia. To date there are no reports in the international literature of observational or audit studies of sales of illicit tobacco products from either conventional stores or street sellers.17 In the context of the introduction of plain packaging, we established an audit of availability of very cheap tobacco in a representative sample of conventional retail outlets in major cities across the country.
Tobacco companies have claimed that use of illicit branded tobacco is likely to increase due to plain packs being easier to counterfeit.2 At least one tobacco company has also claimed that use of unbranded tobacco will also increase due to an erosion in customers’ emotional investment in brands.3 Clarke and Prentice have queried the likelihood of this prediction, pointing out that higher price competition between the major companies following introduction of the policy would be likely to reduce demand for illicit tobacco.18 To assess the industry's claims, we monitored the characteristics of cheap cigarettes available for sale, and the willingness to sell illicit tobacco in small mixed businesses across Australia before, during and for 8 months after the December 2012 introduction of plain packaging.
Store sample and selection
The store sample was derived from a larger sample (described elsewhere,19) established in May 2012 located in four Australian capital cities. Sixteen postcode-based areas in each of Melbourne and Sydney were sampled, and eight in each of Adelaide and Perth. Areas were stratified by socioeconomic status (SES) and randomly selected from each SES quartile.20 For efficiency, a secondary postcode in close geographic proximity and with similar SES characteristics to each primary postcode was also selected, but only sampled if store quotas were unable to be filled in the primary postcode of that area.
Stores were required to sell tobacco and have a visible price board listing at least three cigarette brands. Fieldworkers travelled to a predetermined starting point and, using a rigorous set of walking rules, sampled every potentially eligible store they encountered. Fieldworkers continued sampling until they had at least one of each of the required store types in the sample. If they did not meet store quotas in the first location, they repeated the procedure in the next and subsequent locations until the store quota was filled or all locations had been visited.
A minimum of six stores per area were required, including at least one of the following four store types: small general grocery stores: small businesses and family run stores, typically denoted as ‘milk bar’, ‘deli’, ‘corner store’, or ‘general store’; convenience stores: independent supermarkets and branded chain convenience stores (such as ‘7–11’) with extended opening hours and product range; newsagents/lottery outlets: selling newspapers, magazines and stationery as their primary business and/or lottery tickets; and petrol stations: selling petrol, including those branded as convenience stores or supermarket chains. Large supermarkets run by national chains (Coles and Woolworths/Safeway) and petrol stations were excluded from the unbranded chop-chop substudy sample as it was deemed highly unlikely that such stores would sell unbranded tobacco, since they are nearly all operated by high-profile publicly listed companies with procedures in place for ensuring compliance with tax and other government legislation and company policies. Supermarkets were also excluded the purchasing component of the study (see below) as one of the major supermarket chains does not allow its staff to identify the cheapest pack of cigarettes available, a key step in our data collection process. Further, to ensure credibility, the survey protocol required that the store not be busy with other customers at the time of the enquiry, a situation rarely encountered in these types of stores.
The resulting panel of stores for each substudy is described in table 1. Stores were visited during each of six periods—between 20th June and 2nd July 2012 and between 6th and 24th September 2012 (both prior to implementation of plain packaging), again from 3rd to 12th December 2012 (within the implementation period), and during the periods 14th to 28th February 2013, 11th to 24th April 2013, and 15th to 23rd July (well after the period of initial implementation). Stores that were temporarily or permanently closed were not replaced in subsequent visits, but were revisited in every subsequent data collection period to confirm their status.
Fieldworkers (n=23) ranged in age from their 20s to their 60s, with most aged in their 40s and 50s, and 61% were men, consistent with the profile of users of unbranded tobacco detected in government national surveys.21 Fieldworkers underwent thorough training using mocked-up price board examples, as well as a 60–90 min training session on all steps of the data collection, including the request for cheap cigarettes and chop-chop. Fieldworkers were able to practise these methods in unselected stores prior to the main data collection.
In each store, the fieldworker asked to buy a pack of low-cost cigarettes—a particular variant (for ease of memory) of a value brand in a small pack size so that the retailer knew that the purchaser was interested in cheap cigarettes. The time taken to retrieve this pack was unobtrusively recorded, and is reported elsewhere.19
After the retailer retrieved the requested pack, the fieldworker enquired whether a cheaper pack of cigarettes was available, asking ‘What's the cheapest pack you have?’ Fieldworkers were instructed to only accept a pack that was lower in purchase price, not lower in stick price. As the fieldworker paid for the proffered cheapest pack, they asked the retailer about the availability of chop-chop.
Fieldworkers used one of three phrases (rotating these across stores), always including the words ‘chop-chop’ in their request, such as:
Thanks. I was just wondering if you knew where I could get some really cheap loose tobacco, you know, chop-chop?
The fieldworker took note of the retailer's response as they completed the purchase of the cheap pack and then exited the store. If the fieldworker was directly offered unbranded tobacco, they declined to purchase it. If the retailer confused chop-chop with branded roll-your-own (RYO) tobacco, the fieldworker made one attempt to clarify that they wanted chop-chop and not RYO tobacco. Unless the retailer then gave a positive response (ie, information or offered to sell it), these responses were coded as ‘confused with RYO tobacco’.
The fieldworkers coded the retailers’ responses into one of eight options. These were (1) unbranded tobacco directly offered when fieldworker asked for the cheapest pack available; (2) the retailer said ‘no’, ‘don't know’, or similar; (3) the retailer confused chop-chop with RYO tobacco; (4) the retailer gave vague information about where chop-chop can be purchased; (5) the retailer gave specific information about where chop-chop can be purchased; (6) the retailer directly offered to sell it to the fieldworker; (7) the retailer became suspicious of the fieldworker (and gave no other information); and (8) other, where the fieldworker recorded the retailer's response verbatim to be back-coded later. ‘Vague information’ included non-specific directions, such as ‘Try the market’ or ‘I think you can get it at one of the tobacconists around here’. ‘Specific information’ included where more direct information was provided, such as ‘I've heard of people buying it from X store’.
Assessment of irregularity in health warnings and suspiciousness of prices
All packs purchased were examined by the study coordinator to determine whether the packaging was irregular or the price suspiciously low. To assess whether health warnings were in any way irregular, each pack was compared with the images contained in either the Australian health warning regulations applicable between 2006 and 2012,22 or the more recent Consumer Product Information Standard applicable from 2012.23
Price was analysed in a two-stage process. First, the tax liability for each pack purchased was calculated by multiplying the rate of excise/customs duty applicable in each month—0.34681 cents per stick in June/July 2012; 0.34889 cents per stick in August 2012 to January 2013; and 0.35447 cents per stick from February to July 2013)24—with the number of cigarettes in each pack purchased. An allowance for the Goods and Services Tax of 10% on top of this was also added.
Second, recommended retail prices (RRP) were sourced from the price lists in the applicable edition of The Australian Retail Tobacconist published by the NSW Retail Tobacco Traders’ Association25–27 or, in the case of Richland Express brands, from the price catalogue from the Richland Express website.28 ,29 In Australia, the price of cigarettes sold in supermarkets is generally substantially lower than that of those sold in other retail outlets.30 Twenty per cent is the maximum level of discount from RRPs we observed in packs sold in major Australian supermarkets in multiple enquiries between April and June 2012. Being large publicly listed companies subject to internal risk management procedures and external auditing, we believe such supermarkets would be unlikely to sell contraband tobacco. Thus, 20% represents a plausible limit below which prices are much less likely to be legitimate.
For the sake of this exercise, we deemed prices that were more than 20% cheaper than the RRP as ‘suspicious’ and packs cheaper than the tax liability for that pack ‘highly suspicious’.
Assessment of likelihood of tax not having been paid on cigarettes
All products that were irregularly packaged or suspiciously priced were examined further by the principal author and study coordinator, and their appearance and price considered in the context of a number of factors identified in the ongoing study of brand offerings and prices of which this audit formed a part. These included: introduction dates of those brand variants; proximity of the price to $A10, which was an important ‘price point’31 for small packs at the time of the study; and the typical prices observed for each of the brands and store types and individual stores in question. Each pack was then assigned a likelihood of tax not having been paid on it. ‘Likelihood’ is subjectively defined, meaning ‘as perceived by the investigators taking into account the objective criteria and the contextual information’. Percentages were calculated of packs purchased before and after plain packaging for which it was deemed at least somewhat unlikely that tax had been paid.
Data cleaning and analysis was undertaken using Stata/S.E. V.12.1.32
Due to the very low number of irregular packs of branded cigarettes, results are presented in a descriptive format, and statistical analysis was not undertaken.
The frequency of store exclusions and retailer responses to the request for unbranded chop-chop tobacco were examined between waves of preplain packaging (June and September 2012), during implementation (December 2012) and postplain packaging (February, April, and July 2013). Retailer responses were aggregated into three categories for multivariate analysis: (1) positive responses—direct offers to sell, and vague and specific information; (2) negative responses—no/don't know and confused with RYO; and (3) suspicious—the retailer behaved as if they were suspicious about the fieldworker's request. Logistic regression analyses were undertaken to examine differences between waves in the likelihood of encountering a negative, positive and suspicious response. All analyses controlled for store type, SES category33 and city. As no retailers who responded with suspicion were encountered in Adelaide or Perth, city was omitted as a covariate from those models. All models used robust SEs to control for each store having multiple observations.
Irregular branded cigarettes substudy
Table 2 shows that a small percentage of eligible stores were excluded due to store closure, data collection error, or retailer refusal to cooperate.
Numbers and pricing of packs with irregular health warnings
Twenty fully branded packs carrying outdated health warnings (warnings permissible only between March 2006 and November 2012) were sold to fieldworkers in December 2012, but the price of all 20 packs was in line with the RRP. These were all judged to be most likely tax paid, preplain packaged stock that retailers had not managed to sell before 1 December 2012. (News reports at the time indicated that companies were not prepared to provide small retailers with refunds for such stock,34 though such retailers were eligible to receive refunds for the tax component of the wholesale price.35)
Inspection of all 1773 packs purchased showed that just 0.5% were irregular (n=8), in that they did not comply with either the 2006 or 2012 Australian packaging regulations. These packs were all fully branded and lacked Australian pictorial health warnings, and some packs had non-English text. None of the stores that sold the fieldworker an irregular pack did so more than once.
Suspicious pricing of packs purchased that were compliant with health warning regulations
Of the 1765 packs purchased that were compliant with health warning regulations, 14 (0.8%) were more than 20% cheaper than the actual or imputed RRP for the brand in question.
Packs meeting one or both criteria
Overall, 22 packs, or 1.2% of all purchased packs, met one or both criteria.
The online appendix tabulates the percentages in each purchase month for all packs, with and without required health warnings, where prices were more than 20% cheaper than the RRP for those brands (or for comparable brands where RRPs were not available) as well as those that were cheaper than the tax liability for packs of the applicable size.
Thirteen (2.2%) of 598 packs purchased in preplain packaging met either or both criteria. Four packs (1.3%) of 297 met either or both criteria in the December implementation month, and five (0.6%) of 878 did so in the three collection months following plain packaging.
Assessment of likelihood of tax not having being paid
Each of the irregularly packaged and suspiciously priced packs was then considered in detail, and an assessment made about the likelihood of each being tax-not-paid, taking into account the contextual information collected in the survey of prices of which this audit formed a part. These assessments are described in detail in the online appendix.
Based on these assessments, online supplementary table S1 shows tax was judged to be at least somewhat unlikely to have been paid for a total of 10 (0.6%) of the 1773 packs purchased. The percentage of packs at least somewhat likely to be tax-not-paid was no higher after the introduction of plain packaging (2 (0.2%) of 878) than before (6 (1.0%) of 598).
Unbranded chop-chop tobacco substudy
A total of 179 stores were eligible for the enquiry about chop-chop tobacco after supermarkets and petrol stations were excluded. This panel of 179 stores were sampled in every month, and the fieldwork procedure was able to be completed on all six visits in 143 (79.9%) of these. Overall, the enquiry about unbranded tobacco was completed 1022 times over the 6 months, or in 95.2% of store visits. Table 3 shows that the most common reason for the procedure not to be completed was that the store was closed at the time of fieldwork (23 instances, 2.1%), or that the store no longer sold tobacco (3 instances, or 0.3%). Data for nine stores (0.8%) were discarded in June because of fieldworker error; the store was mistakenly excluded from the collection procedure in six instances (0.6%); and the fieldworker was unable to complete the preliminary collection task (purchasing a particular pack of cigarettes) on three occasions (0.3%). Finally, stores were deemed too busy to initiate this procedure eight times (0.7%), meaning the fieldworker felt there were too many other customers to plausibly complete the enquiry about illicit tobacco.
Availability of unbranded tobacco
Table 4 shows that in 88.5% of the total 1022 enquiries, retailers responded that they either did not know what chop-chop tobacco was, did not know where it could be purchased, or they confused unbranded tobacco with RYO tobacco. Fieldworkers were given vague information about where it could be purchased in 4.5% of enquiries (n=46), and provided more specific information about where it could be purchased in 13 instances (1.3%). Fieldworkers were directly offered unbranded tobacco in six enquiries (0.6%). During 53 enquiries (5.2%), the retailer appeared to become suspicious of the fieldworker and did not give a response.
Of the six direct offers to sell unbranded tobacco (0.6%), the first four instances occurred in different stores, while the same store offered unbranded tobacco in April and July. In February, the store that offered unbranded tobacco did so when the fieldworker asked to purchase a cheap pack of cigarettes.
Logistic regression analyses, first using the preplain packaging wave then the implementation wave as the reference category, found no significant differences across waves in the likelihood of encountering a positive response to the request for unbranded tobacco (preplain packaging: 7.4%; implementation: 7.1%; postplain packaging: 5.5%). A significant difference was found for negative responses. Negative responses were significantly more likely to occur in the postplain packaging wave (90.7%) compared to the implementation wave (82.9%; OR=2.08 (95% CI 1.22 to 3.54), p=0.007). There were no differences in negative responses between the preplain and postplain packaging waves.
Finally, retailers responding with suspicion toward the fieldworker were significantly less likely to be encountered in the preplain packaging wave (4.7%; OR=0.44 (95% CI 0.21 to 0.91), p=0.026) and the postplain packaging wave (3.9%; OR=0.36 (95% CI 0.17 to 0.73), p=0.005) compared to the implementation wave (10.0%). There were no differences in the frequency of retailers responding with suspicion between the preplain and postplain packaging waves.
All the stores that offered to sell illicit tobacco appeared in all six collection months of the survey. A sensitivity analyses was conducted by repeating the above analyses, including only those stores that appeared in all 6 months (n=143). The percentages in each category of response changed very slightly, and the pattern of results was the same.
An estimated one billion packs of cigarettes are purchased each year by Australian smokers.36 The packs purchased in this study are a sample not of this total ‘universe’ of purchases, but rather of purchases when retailers were specifically asked for the cheapest pack in the store. Only eight (0.45%) of the 1773 packs purchased in these circumstances appeared irregular (non-compliant with required health warnings), and only one of these was identified in the plain packaging postimplementation period, other than those branded packs purchased in December 2012 that were deemed to be likely surplus stock. This is perhaps not surprising given how much more visible fully branded irregular packs have become in the plain packaging environment, and given the well-publicised penalties for breaches. The fine for selling tobacco products non-compliant with the legislation—the subject of a Government information campaign prior to implementation—amounted to $A340 000 during the early months of 2013.1 Of the 14 packs that did bear required health warnings, but were priced cheaper than 20% below the estimated RRP, all but two were new on the market in Australia, and low introduction pricing rather than tax evasion seems a much more likely explanation for these observed low prices. Overall—taking into account criteria and the contextual factors collected as part of the wider study of which this study was a part—fewer than 1% (n=10) of the 1773 packs purchased were deemed at least somewhat unlikely to be tax paid, with only two of these appearing after the introduction of plain packaging.
Retailers in this study were no more or less likely after the introduction of plain packaging to directly offer to sell unbranded chop-chop tobacco, or provide information about its local availability. The study protocol worked well, with minimal indications of wariness about the legitimacy of the fieldworkers. The increase in retailers responding with suspicion during the December 2012 implementation month was temporary, dropping markedly in 2013, and so appeared most likely due to increased wariness among retailers about the possible presence of Government personnel employed to oversee compliance with the plain packaging legislation.
This study of willingness to sell illicit tobacco in retail outlets does not cover sales from informal sources, such as friends and families and contacts selling illicit tobacco from informal venues, such as market stalls and car parks promoted by word of mouth. Sales of such tobacco are not amenable to detection through audit studies, and would need to be assessed through consumer surveys.
The industry-sponsored reports by Deloitte consulting37 ,38 reported that specialist tobacconists were one of the most common sources from which users obtained illicit tobacco. While such stores could not be included in the current study, in a companion study, we tested willingness to sell unbranded tobacco among a group of 54 specialist tobacconists in Victoria in 2013.39 Similar to this study, there were very few instances of positive responses to the enquiry about unbranded tobacco, with tobacconists offering to sell unbranded tobacco on only 5 (3.1%) of 162 total purchase occasions, with specific information about where it could be purchased locally provided on a further five occasions (3.1%), and vague information on a further 10 (6.2%). Additionally, three fully branded packs (1.9% of 162 purchases) without required health warnings were offered for substantially less than RRP.
Audit studies, such as this one, are not without limitations. Fieldworkers were well trained, rotated across stores and demographically similar to known users of illicit tobacco. Nevertheless, it is possible that retailers would be more willing to sell to known regular customers. The study may not have captured what could be a small number of retail outlets known by users to sell large amounts of illicit tobacco, although if such stores did exist they would be inconvenient for smokers unless they lived and/or worked nearby. Our study, by contrast, does reflect the likely experience of ordinary consumers attempting to locate unbranded tobacco from everyday small mixed business tobacco retailers. Our study also has the advantage of a large sample, random selection of areas and a low rate of store loss from the sample. By contrast, no information is provided in the most recent industry-funded survey of discarded packs40 concerning areas of cities sampled to generate an estimate of the percentage of packs in Australia that are foreign in origin.16 By contrast with our study which finds very low levels of willingness to sell unbranded tobacco or cigarettes likely to have been illicit, that study has estimated that illicit tobacco comprises more than 13% of the total market in Australia in 2013.16
The additional contextual information collected in our study suggested that neither ‘lack of compliance with Australian health warning legislation’ nor ‘suspiciously low prices paid’ are failsafe indicators of illicit (tax-not-paid) status of cigarette packs sold in retail outlets. The simplistic application in Australia of indicators, such as those used in the European study of Joossens et al41 could overestimate the extent of illicit sales in Australian retail outlets. It should be noted that while the current study was able to identify packs on which it is highly unlikely that tax has been paid, we could not definitively determine the tax-paid status of any pack. It is conceivable that duty could be paid on packs non-compliant with health warning regulations. This might occur where small numbers of packs are being manufactured in Asia for export to Australia, and where orders of particular brands in a given month exceed supply of packs with correct health warnings and manufacturers (well intentionally though not legally) supply a number of packs manufactured for the local market to make up the shortfall. Packs on which taxes have been paid could be priced well below the RRP because they are close to the ‘use-by’ date, because the manufacturer wishes to promote a new or renamed or reformulated brand, or for some other reason. Conversely, other packs may be contraband but still be priced close to or above the RRP, providing very high profits to the illicit suppliers and/or retailers. Estimates derived from our study avoid much of the definitional ambiguity and problems of incomplete understanding and reticence among consumers described earlier. However, the results should not be interpreted in isolation as a definitive estimate of the extent of illicit trade in Australia. Rather, they complement data collected and intelligence from other sources.42 While it is impossible to be certain about the illicit status of any individual product, the results of this study suggest that the incidence of sales of illicit (tax-not-paid) cigarettes in small retail outlets is most likely very low. To reach an overall assessment of the likely scale of illicit trade in a country such as Australia, evidence from studies such as this one needs to be combined with results of audits of specialist tobacconists, surveys of consumers, intelligence from customs surveillance and interceptions, and analysis of patterns in revenue collection.
We conclude that availability of illicit tobacco from small mixed businesses in Australia has not increased in this early implementation period of plain packaging.
What this paper adds
Tobacco companies vigorously opposed the introduction of plain packaging in Australia.
A key argument in the tobacco industry's case against this and several other major regulatory initiatives is that such laws will increase the prevalence of illicit trade.
It is important for governments to quantify the extent of illicit trade and evaluate the impact of policies.
This study provides an early indication that the availability of illicit tobacco from retail outlets did not increase following implementation of plain packaging in Australia.
Contributors MS and MW codesigned the study and MB coordinated fieldwork and undertook data analysis. All authors contributed to manuscript writing and revisions and approved the final version of the paper.
Funding This study was supported by Quit Victoria, Cancer Council Australia, Cancer Council South Australia, Cancer Council Victoria and Action of Smoking and Health (UK) and its partner organisations Cancer Research UK, Smokefree Southwest, Fresh and Tobacco Free Futures.
Competing interests None.
Ethics approval Institutional Research Review Committee, Cancer Council Victoria.
Provenance and peer review Not commissioned; externally peer reviewed.
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.