Article Text

Download PDFPDF

Estimating the effect of a potential policy to restrict tobacco retail availability in New Zealand
  1. Lindsay Robertson,
  2. Louise Marsh
  1. Cancer Society Social and Behavioural Research Unit, Department of Preventive and Social Medicine, University of Otago, Dunedin, New Zealand
  1. Correspondence to Dr Lindsay Robertson, Department of Preventive and Social Medicine, University of Otago, Dunedin 9054, New Zealand; l.robertson{at}otago.ac.nz

Abstract

Introduction Reducing tobacco retailer density is suggested to be a key component of tobacco endgames. Where tobacco retail reduction policies exist, ‘grandfathering’ is typically used, where existing retailers are exempt from restrictions aimed at reducing outlet density. We examined the possible impact on tobacco outlet density in New Zealand of a policy preventing new retail outlets from selling tobacco, but allowing existing retail outlets to continue selling tobacco until they ceased trading or relocated.

Methods We obtained data on numbers of tobacco outlets and the number of outlet closures recorded annually from 2006 to 2016. We calculated the mean and the lowest and highest 3-year rolling average closure rates for each type of tobacco outlet. We projected decreases in the number of tobacco outlets that would hypothetically be permitted to sell tobacco from 2020 onwards and estimated when the combined number of tobacco outlets would decline by 50% and 95%.

Results Based on mean annual closure rates, the total number of tobacco outlets would decrease by 50% by 2032 and a 95% reduction could be achieved by 2072. By 2025, the number of tobacco outlets would decrease by 27%; by 2050, this reduction would reach 84%.

Conclusion A tobacco retail reduction policy that prevented new retail outlets from selling tobacco and grandfathered existing retailers would be unlikely to achieve New Zealand’s target of a 95% reduction in tobacco outlet density within several decades of being enacted. Nonetheless, this policy could achieve a 50% reduction in tobacco retail availability in the first decade of implementation.

  • public policy
  • end game
  • advertising and promotion

Statistics from Altmetric.com

Introduction

In New Zealand (NZ), tobacco availability is ubiquitous. Around 8000 outlets including supermarkets, service stations and convenience stores sell cigarettes alongside everyday products such as milk and bread and in settings frequently visited by children.1 Given the importance of distribution or ‘place’ as a marketing tool, the widespread availability of tobacco is a major form of tobacco promotion.2 Higher tobacco retailer density promotes youth smoking,3 increases consumption among established smokers4 and reduces the odds of smoking cessation.3 5 6 Limiting tobacco availability is viewed as an important part of the tobacco endgame,2 7 8 and as more countries adopt endgame goals,9–12 policymakers will increasingly need to consider strategies to substantially reduce tobacco outlet density.

NZ’s government has committed to an endgame goal that aims to reduce ‘smoking prevalence and tobacco availability to minimal levels’ by 2025,9 but has not initiated action to reduce tobacco availability since the goal was established in 2011. In the absence of a government strategy, NZ’s tobacco control sector developed an action plan to achieve the 2025 goal, in which reducing the number of tobacco outlets by 95% is identified as one of the three key priorities.13

Very few jurisdictions have implemented policies to reduce tobacco retail availability. In 2013, Hungary reduced tobacco outlet density by 83% by only allowing tobacco sales at 7000 new government-owned stores,14 though the manner in which tobacco retail licences were removed and redistributed appears to have been corrupt and resulted in legal action by retailers.15 In California, several jurisdictions have adopted a less radical approach, prohibiting new retailers from obtaining a tobacco retail permit if they are located in residential areas (eg, Huntington Park), or within a specified distance from schools or another tobacco retailer (eg, Santa Clara County, Richmond).8 16 17 San Francisco City passed a law imposing a limit of 45 tobacco retailing permits for each of the 11 districts, estimated to reduce tobacco outlet density by 50% over 10–15 years.18 In each of the Californian examples, authorities have used ‘grandfathering’, where existing retailers are exempt from restrictions. While grandfathering is more acceptable to retailers, the trade-off is a gradual decline in tobacco outlet density, achieved only through natural attrition of outlets from the market. Nevertheless, the Californian models of restricting tobacco availability are likely to be more politically feasible in NZ.19

We examine the likely impact on tobacco outlet density of implementing a policy in NZ, similar to Californian precedents, where any new retail outlets are prohibited from selling tobacco but existing retail outlets are permitted to continue selling tobacco, until they cease trading or relocate their business.

Methods

Outlet data

As there is no official register or database of all tobacco retailers in NZ, we obtained Statistics NZ Business Demography data20 on numbers of grocery store outlets (which includes supermarkets and convenience stores) and service station and liquor store outlets, recorded annually from 2006 to 2016. Only a negligible number of these outlets have voluntarily stopped selling tobacco; these categories thus represent around three-quarters of the tobacco outlets in NZ (the remaining outlets are mostly alcohol on-licensed premises; no pharmacies sell tobacco in NZ).1 An outlet is defined as the location where a business activity is operated and is distinct from an enterprise, which is a legal entity that owns the business (and which may own and operate multiple outlets).

We also obtained data on the number of outlet closures within the three tobacco outlet categories annually from 2006 to 2016. Closures of outlets are determined by Statistics NZ on the basis of changes in land, capital investment and employees associated with a business. An outlet closure occurs where an enterprise ceases trading at a particular location, or closes an outlet at one location and opens a completely new outlet at a different location (the outlet at the first location is classed as closure). If, on the other hand, an enterprise moves an outlet at one location to a new premises in close proximity to this location, while retaining all the same capital investment and employees, this change is not classified as a closure, but rather a continuation of the outlet. Changes in business ownership do not necessarily result in outlet closures, under Statistics NZ definitions.

We calculated the mean annual closure rate for outlets and 3-year rolling average closure rates, within each outlet category. We took the lowest and highest 3-year rolling average closure rates to model ‘lowest-rate’ and ‘highest-rate’ scenarios, respectively.

We used the closure rate estimates to project decreases in the number of tobacco outlets that would hypothetically be permitted to sell tobacco from 2020 onwards and to estimate when the combined number of tobacco outlets would fall by 50% and 95%. We chose 2020 as the starting date, as we considered this to be a date from which the NZ government could feasibly implement a tobacco retail reduction policy.

Assumptions

The assumptions for the hypothetical policy were: (1) from 2020, the government would prohibit any new outlets from selling tobacco, (2) once an existing outlet ceases trading or relocates it may no longer sell tobacco, (3) existing tobacco outlets are grandfathered, (4) the number of tobacco outlets is unchanged between 2016 and 2020 and (5) outlet closure rates remain constant each year.

Results

In 2016 (and hence in 2020 under our assumptions), there were 6057 tobacco outlets in NZ, comprising 3777 grocery stores, 1212 service stations and 1068 liquor stores. Between 2006 and 2016, the mean annual closure rate for grocery outlets was 6.7% (highest rate=8.9%; lowest rate=6.1%). For service stations, the mean closure rate was 4% (highest rate=4.9%; lowest rate rate=2.3%) and for liquor store outlets, the mean closure rate was 6.0% (highest rate=6.7%; lowest rate=5.4%).

Based on the mean annual closure rates, the combined total number of tobacco outlets decreases by 50% by the year 2032 (see figure 1). According to the highest-rate scenario, this degree of attrition is achieved by the year 2029, and the lowest-rate scenario suggests the number of outlets reduces by 50% by the year 2034. Based on the mean closure rate, a 95% reduction in the number of tobacco outlets is achieved by 2072; this level of attrition is realised by 2061 according to the highest-rate scenario, or by 2091 according to the lowest-rate scenario.

Based on the mean closure rates, by the year 2025 (NZ’s endgame target), the number of tobacco outlets decreases by 27%; by 2040, the number of outlets decreases by approximately 71% and by 2050 this reduction reaches 84% (figure 1).

Figure 1

Projected decreases in the number of outlets permitted to sell tobacco, by mean, highest-rate and lowest-rate business closure scenarios. Note: the horizontal lines correspond to a 50% and a 95% reduction in tobacco outlet density; the drop-down lines represent the year(s) in which those reductions would be achieved, based on the highest rate, mean and lowest rate closure estimates.

Discussion

Our analysis suggests a tobacco retail reduction policy in NZ, which prevented new retail outlets from selling tobacco and grandfathered existing outlets, could achieve a 50% reduction in tobacco outlets within 9–14 years of implementation and 70% reduction by the year 2040. This approach would not achieve NZ’s endgame goal of reducing tobacco availability to ‘minimal levels’ by 20259 or the sector’s target of a 95% reduction in outlet density by 2022,13 but would nonetheless result in a significant advancement from the status quo.20 Alternative regulatory models based on direct regulation would result in faster declines in outlet density21 22; however, this approach is yet to be implemented anywhere other than Hungary. An alternative would be to prohibit tobacco sales at certain types of outlet yet amortising retailers for a certain period of time (eg, 5 years), during which they may continue selling tobacco, but after which they would be required to cease.21 This would bring about faster declines in tobacco availability than a grandfathering approach and would provide retailers with sufficient time to adapt and transition. Another possibility is a progressive ‘sinking lid’ on the amount of commercial tobacco released for sale, with an auction system to allocate licences for tobacco retail suppliers.23

Reducing tobacco retail availability will likely have the greatest impact on smoking behaviour when the density of tobacco retailers is very low (eg, at <10% of current density),24 and modelling studies suggest achieving this reduction would help lower smoking prevalence and health inequities.25 Yet, previous research highlights the trade-off between policy effectiveness and acceptability.19 Gradual declines in tobacco outlet density are likely to be more politically acceptable than rapid changes.19 Policies that remove the ability to sell tobacco from existing outlets in a discriminatory way, such as by only targeting convenience stores or outlets located close to schools, are less acceptable for retailers,26 thus less politically feasible. Retailers are less likely to oppose regulation that either ensures all existing retailers are exempt from legislative change or—if existing retailers are affected—that the policy is not anticompetitive.26 Thus, a policy using grandfathering may be the most feasible way of progressing a reduction in tobacco outlet density.

Our study has several limitations. First, we were only able to estimate the impact of the policy on three categories of retailers, rather than all tobacco outlets, although the three categories represent three-quarters of tobacco outlets in NZ.1 Second, closure rates of outlets may change over time, particularly if a tobacco retail reduction policy was implemented. A licensing system for existing tobacco retailers would be required to operationalise this policy, which could lower the rate of retailer closure; our model did not account for those possible changes, and our estimates could therefore be optimistic. Yet, there are also ways in which the reduction in tobacco outlet density could be accelerated. For instance, our analysis assumed changes in business ownership did not necessarily result in an outlet closure, but in reality, the government could equate changes in ownership with a ‘closure’ (ie, loss of tobacco retail licence), which would increase the rate of reduction. Similarly, additional restrictions could be added to the policy we examined (eg, requiring all sales staff to be aged over 18 years, requiring staff training on tobacco sales legislation),8 with repeated non-compliance resulting in loss of a tobacco retail licence. A further reason why our estimates may be optimistic is that the number of tobacco outlets may, in reality, be higher in 2020 than in 2016, given some growth in tobacco outlet numbers between 2011 and 2016 (though there was no such growth between 2006 and 2011).20 Lastly, policymakers must consider legal barriers that constrain tobacco retail reduction policies21; it is possible that in some jurisdictions, grandfathering existing retailers is the only legal option.

Notwithstanding the limitations, our estimates are based on the best available data on outlet closure rates and provide the first estimate of the impact of a potential policy on tobacco outlet density in NZ. The policy we have examined would not disadvantage existing retailers and is therefore likely to be a politically acceptable means of reducing tobacco retailer density in NZ and in other jurisdictions.

What this paper adds

  • Decreasing the number of outlets that stock tobacco would help reduce youth smoking initiation and enable smokers to quit more easily. As an increasing number of countries adopt tobacco endgame goals, identifying feasible policies to reduce tobacco outlet density is pertinent.

  • Very few jurisdictions have enacted policies to reduce tobacco outlet density. Most policy precedents come from Californian jurisdictions, where restrictions on selling tobacco have been placed on new retail outlets, and where existing outlets can continue selling tobacco until they close their business.

  • The potential effect of this approach on tobacco outlet density in other jurisdictions has not been evaluated.

  • In New Zealand, a policy that prevented new retail outlets from selling tobacco and which allowed existing retailers to continue selling tobacco until they closed the outlet, could decrease the total number of tobacco outlets by 50% by 2032. The policy would be unlikely to achieve New Zealand’s target of a 95% reduction in tobacco outlet density within several decades of being enacted. Nonetheless, it would result in a significant advancement from the status quo.

Acknowledgments

We would like to thank Emeritus Professor Phil Gendall for helpful comments on an earlier draft of this manuscript.

References

Footnotes

  • Contributors LR conceptualised the project, obtained the data, conducted the analysis and prepared the first draft of the manuscript. LM provided feedback on the concept and contributed to the manuscript. Both authors have seen and approved the final version.

  • Funding LR and LM were supported by a grant from the New Zealand Cancer Society.

  • Competing interests Although we do not consider it a competing interest, for the sake of full transparency we note that authors have undertaken work for health-sector agencies working in tobacco control.

  • Patient consent Not required.

  • Ethics approval Ethical approval was not deemed necessary for this analysis.

  • 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.