Article Text

Download PDFPDF

The impact of an increase in excise tax on the retail price of tobacco in New Zealand
  1. Louise Marsh1,
  2. Claire Cameron2,
  3. Robin Quigg1,
  4. Janet Hoek3,
  5. Crile Doscher4,
  6. Rob McGee1,
  7. Trudy Sullivan2
  1. 1Cancer Society Social and Behavioural Research Unit, University of Otago, Dunedin, New Zealand
  2. 2Department of Preventive and Social Medicine, University of Otago, Dunedin, New Zealand
  3. 3Department of Marketing, University of Otago, Dunedin, New Zealand
  4. 4Faculty of Environment, Society and Design, Lincoln University, Canterbury, New Zealand
  1. Correspondence to Dr Louise Marsh, Cancer Society of New Zealand Social and Behavioural Research Unit, Department of Preventive and Social Medicine, University of Otago, P.O. Box 913, Dunedin 9054, New Zealand; Louise.marsh{at}otago.ac.nz

Abstract

Background In 2010, the New Zealand (NZ) government introduced an annual 10% tobacco excise tax increase. We examined retailers' adherence to recommended retail prices (RRP), and whether the RRP included the full tax increase.

Methods We collected price data on three British American Tobacco (BAT) factory-made cigarette brands, (premium, mainstream, and budget), and one roll-your-own tobacco brand before and after the 2014 tax increase from a sample of tobacco retailers. We examined price increases in each tobacco brand and compared these with the RRP. The extent to which the excise tax increases had been included in the RRP since 2010 was estimated using data sourced from the Ministry of Health and NZ Customs.

Findings The median increase in price from before to after the tax change was only 3% for the budget brand (461 retailers). This contrasted with the median of 8% for the premium brand (448 retailers), and 11% for both mainstream and roll-your-own brands (471 and 464 retailers, respectively). While many retail outlets made changes according to the RRP set by BAT, several did not comply. Our analyses suggest BAT may be undershifting excise tax on the budget brand, and overshifting tax on brands in other price partitions.

Conclusions Tobacco companies do not appear to be increasing the RRPs of budget brands in line with tobacco excise tax increases. The increasing price differential between budget brands, and mainstream and premium brands may undermine cessation and impede realisation of New Zealand's Smokefree 2025 goal.

  • Price
  • Public policy
  • Taxation
  • Tobacco industry
View Full Text

Statistics from Altmetric.com

Introduction

The World Bank has stated that “increasing tobacco taxes is the single most effective measure that governments can take to reduce health burdens.”1 Evidence for the effectiveness of tax and price policies is overwhelming;2 ,3 increased costs of tobacco to consumers are strongly associated with declining smoking prevalence and reduced tobacco consumption.3 ,4 A recent review estimated an average demand elasticity for tobacco to be −0.4 for high income countries (ie, a price increase of 10% would reduce consumption by 4%).3 Young people tend to be more price sensitive compared with adults,2 ,5–7 with a price elasticity of demand for teenagers between −0.5 and −1.2.3 Increases in the price of tobacco also reduce smoking initiation and uptake among young people,2 ,8 as well as progression from experimental to regular use.2 ,8

There is strong evidence that low-income groups within high income countries show greater price sensitivity.2 ,3 ,5 ,6 However, concerns have arisen that a regressive tax may inequitably burden smokers from lower socioeconomic groups who continue to purchase tobacco.9 In New Zealand (NZ), lower socioeconomic groups and Ma¯ori (indigenous population of NZ) have disproportionately higher smoking rates,10 and have shown less rapid declines in smoking prevalence11 than other population groups. Tax increases that stimulate greater cessation among these groups could help reduce health inequalities and enable quitters to make financial savings in addition to obtaining health gains.5

Industry documents reveal tobacco companies understand the effects of large tax increases on smoking prevalence, particularly among young people.12 Given tobacco companies’ on-going need to find new, young smokers, they could ameliorate the impact of tax increases on groups with higher consumption rates by absorbing these taxes or by applying these taxes differentially. For example, tobacco companies could minimise increases on budget brands, while applying these fully to premium brands. This approach would minimise the impact (and hence stimulus to quit) for poorer smokers who predominantly smoke budget brands and potentially mitigate or nullify the price increase for smokers willing to change from a premium to a lower cost brand.

UK data provide some evidence that tobacco companies have not applied tax increases to ‘budget’ brands, but have instead differentially shifted price increases between brands in different price partitions.13 This strategy creates a greater price gap between the premium and budget brands.13 Evidence that the price of ‘ultra low price’ cigarettes has remained constant while the market share of this partition has increased suggests widespread use of this strategy, which helps retain price-sensitive smokers, such as young people,12 ,13 and offers alternatives to cessation when price increases occur.

In NZ, tobacco excise duty is levied on all domestically manufactured goods, while excise-equivalent duty is levied on imported tobacco. The excise tax rate is weight based; for manufactured cigarettes this weight is per 1000 cigarettes, while for the roll-your-own (RYO) tobacco it is based on the per kilo tobacco content. Excise tax is reviewed annually to allow for changes in government taxation policy and also for movement in the consumer price index (CPI), a measure of household inflation. Tobacco is also subject to Goods and Services Tax (GST) at a rate of 15% from October 2010, and 12.5% prior to this date. In April 2010, the NZ government announced a series of tobacco excise tax increases. The first of these was a surprise tax increase which added 10% to factory-made cigarettes, and 25% to RYO tobacco; the second was a 10% annual excise tax increase on manufactured and RYO tobacco for 2011 and 2012. In 2012, the government extended this provision to apply until 2016.14

The unannounced April 2010 tax increase saw the proportion of smokers making a quit attempt increase from 30% in 2009 to 39% in 2010,15 while calls to the Quitline (the national smoking cessation service) increased significantly.16 Callers were more than three times as likely to cite cost as the reason for quitting in 2010 relative to 2009.15 Yet, while tobacco price increases stimulate cessation attempts,2 ,17 their impact may not be maintained. In NZ, successive tax increases have not elicited the same response; for example, inbound Quitline calls have declined since 2010.18 Reduced call volume may indicate acclimation as smokers anticipate and accommodate tax increases by switching to a cheaper alternative of tobacco, such as RYO19 tobacco, reducing and regulating consumption, and displacing other purchases.20

While smokers may adjust their behaviour to accommodate price increases, tobacco companies may also attempt to reduce the impact these increases have on smokers. Mitigating the effect of tobacco excise tax increases would present a serious challenge to NZ for becoming a smoke-free nation by 2025.21

Given that only tax increases which directly alter retail prices will affect smoking prevalence, it is surprising that few studies in NZ have examined the relationship between tobacco tax increases and the retail price of tobacco. We therefore examined (1) whether, on average, tobacco retailers charge British American Tobacco's (BAT's) recommended retail price (RRP) for different brand segments, and (2) whether BAT applied the annual 10% tobacco excise tax increase differentially to budget, mainstream, premium and RYO brands over the period 2010–2014.

Methods

Sampling

We collected tobacco price data from a sample of the 5059 known NZ tobacco retailers, identified using a national spatial database of tobacco retailers from 19 of the 20 District Health Boards (DHB),22 and by including retailers from the sole DHB not represented in the database. We drew a sample of 2663 retailers from 10 of the 20 DHBs who agreed to participate (figure 1).

Figure 1

Sampling strategy (DHB, District Health Boards).

We focused on four main retailer groups: convenience stores (dairy, corner store, mini mart, and small supermarket), supermarkets, service stations and liquor stores; these outlets constituted 72% of the retailers in the database.22 Other retailer groups were omitted for practical reasons (eg, on-license premises were omitted as their operating hours varied greatly from other outlets); less than 10% of smokers purchased their current cigarette pack or tobacco pouch from retailers in the ‘other’ category.23

These four retailer groups comprised 1530 of the 2663 retailers. We initially sampled half of these 1530 (765 retailers) retailers using a proportional random sample based on area deprivation (as defined by NZDep2006;24 retailers were classified into low (1–3), medium (4–7) and high (8–10) deprivation locations). The sample chosen for each DHB/deprivation combination was directly proportional to the number of retailers in each area. However, time and resource constraints meant one DHB could provide only limited assistance; thus the sample was recalculated and the final sample comprised 711 retailers.

Data collection

Price data were collected by Smokefree Enforcement Officers (SFEO) between November and December 2013 (stage 1), and again in January and February 2014 (stage 2). Where an SFEO was unavailable to undertake all data collection, staff from the University of Otago and the Cancer Society of NZ also collected data.

We followed Burton et al's25 methods and provided each data collector with complete instructions as well as information that could be offered to retailers, if required. Data collectors used a survey form to record where a retailer no longer existed, no longer sold tobacco, was unable to be visited, declined to take part, or did not stock a brand. If the retailer sold tobacco, the SFEO recorded the retail price (ie, purchase price consumers paid) of the four BAT tobacco brands described below; these prices included the 15% GST applied to all grocery products. Completed survey forms were returned to the researchers for data entry.

Measures

BAT brands were studied as BAT has the largest market share in NZ (72%),26 thus increasing the likelihood that most retailers sampled would stock BAT brands. Tobacco companies typically provide stores with tobacco storage units and influence the brands stocked. We collected data on BAT brands as we expected to elicit more complete data for these than for competing companies’ brands. We reviewed three BAT brands from each of the main manufactured cigarette market partitions27 (premium, mainstream and budget). We defined these groups using Australian research28 and BAT's 2013 price of a single cigarette (premium >0.86c, mainstream 0.80–0.85 cents, and budget <0.79 cents),29 and selected the market leading brand (defined as highest volume sales) in each partition: Rothmans King Size 25 (premium), Holiday Red 20 (mainstream) and Pall Mall Red Baseline 20 (budget). We also reviewed the leading BAT RYO brand, Port Royal original 30 g. The retailer group (convenience stores, service stations, supermarkets and liquor stores) was also recorded using the researchers’ retailer database.22

Tobacco manufacturers and importers must file annual tobacco returns, including the RRP of all their products, with the NZ Ministry of Health. We compared the RRP of the four BAT tobacco brands for 2013 and 2014 against the retail price of those tobacco brands, as recorded by SFEO.29 ,30 To estimate the impact of annual excise tax increases on the RRP for the years 2010–2014,31–33 the RRP was sourced from the tobacco returns data, and tobacco excise tax rates were obtained from NZ Customs.

Data analysis

The retailers were divided into two groups: those who had changed their tobacco prices in line with the RRP of BAT and those who had not. Descriptive analyses examined compliance with BAT's RRP by tobacco brand and identified characteristics of non-compliers. The RRP, GST and excise tax of the four brands were also examined to determine the trend in price over time, and the extent of possible undershifting or overshifting of the excise tax. When the GST exclusive RRP was greater than expected, we assumed the retailer had passed on the full amount of the excise tax (overshifting); when the GST exclusive RRP was less than expected, we assumed the retailer had not passed on the full excise tax increase (undershifting).

Results

Of the 711 retailers examined, price data were available for 492 retailers in stage 1 and 481 retailers in stage 2 (table 1). Table 1 profiles the sample, which reflected the distribution of outlet types in the full database,22 and sets out the main reasons why data were not available.

Table 1

Stage 1 and 2 retailer characteristics

We collected retail price data from 461 retail outlets for the budget brand at both time points: 471 retail outlets for the mainstream brand, 448 for the premium brand and 464 retail outlets for the RYO brand. Table 2 shows the RRP set by BAT for 2013 and 2014, and the per cent change in prices over this period; BAT recommended tobacco retailers to increase the price of the budget brand by 3%, the mainstream brand by 11%, premium brand by 8% and RYO brand by 11%. Changes in the observed data reflect these recommendations as the median per cent change for each brand matches the recommended change. However, table 2 also reveals considerable variation in the prices charged by tobacco retailers; while some have set prices below BAT's RRP, others have set prices that are much higher than BAT's RRP.

Table 2

Percentage change in retail price after the excise tax increase

Many retailers made price adjustments in line with BAT's RRP, and their overall compliance with BAT's individual brand recommendations was 82% (table 3). Retailers were more likely to adjust prices for the mainstream and RYO brand in line with BAT's RRP (87% compliance for each), while compliance dropped for the budget and premium brand (77% for each).

Table 3

Compliance with recommended retail price

Retailers pricing outside BAT's RRP guidelines

Overall, 331 retailers (18%) had not changed prices according to BAT's recommendations. Of retailers selling the budget, mainstream and RYO brands who did not follow BAT's RRP guidelines, around half increased the price of their cigarettes more than the recommendation, though within 20% of BAT's advice. Larger increases were rare, except for the premium brands where prices of 17% of the brands had increased by more than 20%.

Among the price changes below BAT's recommended increase, a decrease or no change in price was most common for the budget brand (38%), whereas for the other price partitions and RYO tobacco, a smaller than recommended increase was most common. A small number of retailers reduced the price of their cigarettes, including 6% of those selling the budget segment brand and 3% of those offering the premium segment brand.

RRP and price changes 2010–2014

To examine tobacco pricing further, we analysed changes in the RRP and excise tax component of tobacco prices from 2010 to 2014 (table 4). For the budget brand, which was introduced in 2011, the RRP (GST excluded) increased by 15% between 2011 and 2012, by 13% between 2012 and 2013, but only by 3% between 2013 and 2014. For each year examined, the RRP was less than expected, taking into account the 10% increase in excise tax and CPI adjustment. This finding suggests retailers did not pass on the full annual increase in excise tax to consumers for this brand. In contrast, the mainstream, premium and RYO brands’ RRP (GST excluded) increased between 9% and 15% each year between 2010 and 2014, suggesting the excise tax increase was passed on in full each year except in April 2010. Although retailers are not required to pass on the full excise tax increase, our analyses show tax has been shifted between the four brands examined, with the budget brand experiencing the smallest price increase.

Table 4

RRP and expected RRP of four BAT brands, 2010–2014

Discussion

The NZ government has legislated annual excise tax increases of 10% plus CPI each year from 2010 until 2016 to reduce smoking prevalence and support their goal of becoming a smoke-free nation by 2025. However, while tobacco taxes are an effective tool in reducing smoking initiation and prompting cessation,34 these taxes will only affect smoking prevalence if the price of tobacco actually increases. Our study suggests that the dominant tobacco company in NZ may have manipulated tobacco prices to ameliorate the effects excise tax increases have on smokers. Specifically, our findings show the full excise tax increase was not applied to the retail price of BAT's budget brand (undershifting), while brands in other price partitions showed higher than expected price increases (overshifting). The RRP of BAT's leading budget brand increased by only 3% following the 2014 tax increase, an outcome that increased the price differential between low-cost and other brands, especially the leading mainstream brand (where BAT's recommended RRP increase was 11%).

Eight in 10 retailers adopted BAT's RRP recommendations, showing high compliance with the leading tobacco company's pricing strategy. Strategies to minimise the effect of excise tax increases on ‘budget’ brands undermine policymakers’ intention by providing smokers with opportunities to downtrade to cheaper brands rather than quit. Since smoking prevalence is concentrated among lower socioeconomic status groups, mitigating the effect of price increases is important to tobacco companies as their largest market is also their most price-sensitive market. At the same time, while maintaining smoking among lower income groups, tobacco companies maximise their profits by increasing the price of the brands that appeal to higher income groups, whose members are less price sensitive and more brand loyal.35

UK research also reported minimal changes in the price of ‘ultra low price’ cigarettes and corresponding increases in the size of this price partition.13 This pattern mirrors NZ data, where the market share of budget brands increased from 8.5% in 2009 to 41.2% in 2013, while mid-price range brands decreased from 68% to 39%, and premium brands remained relatively unchanged.26 These data suggest that during a period of high excise tax activity, NZ smokers have responded to the introduction of budget brands by moving from mid-priced brands to new, lower priced brands. Tobacco industry documents also reveal strategies to maintain very low price brands that appeal to price-sensitive smokers, such as young people,12 minimise the impact of tax increases on low-income smokers and provide brand switching options for smokers who might otherwise quit in response to price increases.36

Furthermore, the long time frame between the announcement of the tax increase and the date when that the tax comes into effect gives the tobacco industry advance warning, and enables them to develop strategies to ‘dilute the effectiveness of the tax increase’.37 These strategies include stockpiling product, building the tax increase into their cost structure, developing new ultra low cost products, and delaying the price increases.

Our findings should generate concern among NZ policymakers, especially given recent modelling that suggests continued annual 10% tax increases will not realise the 2025 smoke-free national goal.21 Even with annual 10% tax increases, the effect of regular tax increases appears to have declined over time17 ,18 and suggests on-going excise tax adjustments may benefit from modification. First, larger increases (20% or greater, rather than 10%) could make the current cross-subsidisation between brands more difficult to sustain. Second, surprise one-off large tax increases (such as 40%) would not allow tobacco companies to plan ahead as they currently may. We recommend policymakers consider adopting both measures. Doing so may also help reduce health inequalities, given recent evidence that smokers from Ma¯ori and Pacific Island communities are more responsive to tax increases than smokers of other ethnicities.38 While large, unexpected, price increases may cause stress to heavily addicted smokers, hypothecation of the taxes collected could greatly expand the cessation support currently available to smokers and would reflect current evidence that excise tax increases should be combined with additional smoking cessation support for low-income smokers.39

Additional policy measures that could limit tobacco companies’ ability to circumvent tax increases include flat rate minimum price laws or mark-up minimum price laws.37 The former measures may prevent introduction of new ultra low cost products while the latter could ensure the full tax increase is reflected in the retail prices. Similarly, price caps could limit the tobacco companies’ profitability (and their capacity to cross-subsidise the tax increase), increase government revenue, and inhibit switching to cheaper tobacco brands.40

This is the first study we are aware of that examines changes in the price of tobacco following an excise tax increase using tobacco price data collected from retail outlets. The findings provide actual environmental data and augment findings from earlier studies.13 While we were limited by the lack of a national tobacco retailer register, a database compiled earlier represented almost every region in NZ and allowed us to ensure that the retailers sampled represented this wider group. We examined brands from only one tobacco company (although this company has 72% market share) and did not review ‘ultra low price’ tobacco products. Future research could assess the feasibility of examining other tobacco companies’ brands, and could also ensure a higher proportion of the sample drawn is visited. Notwithstanding these limitations, we were able to compare a major tobacco company's recommended pricing strategy with retailers’ actual response. Our access to tobacco returns data supplied to the Ministry of Health and actual retail prices provides a robust context for the analyses presented.

Conclusion

The current excise tax and pricing structure allows tobacco companies to differentially shift tax increases between price partitions within the tobacco market and keep prices of budget brands low, thus facilitating brand substitution and impeding cessation. Policymakers need to consider using larger, unsignalled price increases to reduce tobacco companies’ ability to anticipate and ameliorate excise tax increases. They could also consider enacting other legislation to limit the responses currently available to tobacco companies. Excise taxes are highly effective and should play a pivotal role in supporting endgame goals. At present, however, their risk being undermined by tobacco industry strategies designed to sustain smokers’ behaviour for as long as possible.

What this paper adds

  • Excise tax increases are the most effective way of reducing smoking prevalence, but tax increases do not invariably lead to retail price increases.

  • In the first study to use retail price data to examine how excise tax increases affect tobacco retail prices, we found clear evidence that a major tobacco company's recommended retail prices did not reflect an even application of excise tax increases.

  • Regular and predictable excise tax increases may need to be augmented by additional policy measures that increase the probability tax increases will be reflected in retail prices; less predictable and larger increases in prices will leave less room for manipulation.

Acknowledgments

The authors wish to thank the Smokefree Enforcement Officers from the participating District Health Boards who collected data for this study, as well as Delia Teesdale for her data collection. They also wish to thank Professor Richard Edwards for helpful comments on the manuscript.

References

View Abstract

Footnotes

  • Contributors LM, JH and RQ designed the research. CD developed the original retailer database. LM and RQ oversaw data collection. CC analysed the data and wrote the results. LM, RQ, RM and JH developed the manuscript. TS contributed to the revised manuscript. All the authors have reviewed and approved the final version of the submitted manuscript.

  • Funding This work was funded by an Internally Funded Research Grant from the Department of Preventive and Social Medicine at the University of Otago, New Zealand. The Cancer Society Social and Behavioural research unit is supported by the Cancer Society of New Zealand and the University of Otago.

  • Competing interests None declared.

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

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.