0 Authors Queries 60

Journal: Nicotine & Tobacco Research

Paper: 324053
Title: In sight, in mind: Retailer compliance with legislation on limiting retail tobacco displays
5 65
Dear Author
During the preparation of your manuscript for publication, the questions listed below have arisen. Please
attend to these matters and return this form with your proof. Many thanks for your assistance

10 70
Query Query Remarks
Reference

1 A declaration of interest statement report-

15 ing no conflict of interest has been inserted. 75
Please confirm the statement is accurate.

2 Can you please suppy the running head for

this article?

20 80

25 85

30 90

35 95

40 100

45 105

50 110

55 115

Nicotine & Tobacco Research ntr157423.3d 27/6/08 20:19:03

The Charlesworth Group, Wakefield +44(0)1924 369598 - Rev 7.51n/W (Jan 20 2003) 324053
 Nicotine & Tobacco Research Volume 10, Number 8 (January 2008) 1–8

0 60

In sight, in mind: Retailer compliance with

5
legislation on limiting retail tobacco displays 65

Megan Quedley, Brendan Ng, Nikhil Sapre, Matthew Blakiston, Alexandra

10 Crawford, Rosanne Devadas, Hamish McLaren, Shruti Anand, Mahu Tipu, 70
Viswas Dayal, Singithi Chandrasiri, George Thomson, Richard Edwards

Received 8 July 2007; accepted 17 November 2007

15 75

We investigated retailer compliance with point-of-sale display legislation, using a New Zealand region as a case
study. An observational survey was conducted of nonspecialist tobacco retailers in the lower North Island of New
Zealand during 2006. Compliance was assessed in relation to store type (dairies, convenience stores, supermarkets,
and service stations) and by characteristics of the population of the census area unit in which the store was situated.
20 80
These characteristics include the level of socioeconomic deprivation and proportions of Maori (indigenous New
Zealanders), Pacific Islanders, and children aged less than 19 years. Out of the 288 stores surveyed, 185 (64%) had
at least one breach of the point-of-sale regulations. The most common breaches were a failure to display a
‘‘Smoking Kills’’ sign, visibility of tobacco from outside the premises, and displaying tobacco less than 1m from
children’s products. Compliance was significantly worse in dairies (small local general stores) and convenience
25 stores. Stores situated in areas in the top quartile for the proportion of children were much more likely to have high 85
levels of noncompliance (>3 breaches) and to display tobacco products close to children’s products. This study is
one of very few to systematically investigate retailer compliance with point-of-sale display regulations for tobacco
products. The results suggest that the implementation of legislation to partly limit retail displays of tobacco
products can be difficult. A ban on retail displays of tobacco products is likely to be a more effective and
enforceable policy.
30 90

Introduction where tobacco marketing is largely or wholly

35 As in many other jurisdictions (Apte & Viswanathan,
2002; Bloom, 2001; Carter, 2003; Dewhirst, 2004;
Feighery, Ribisl, Clark, & Haladjian, 2003; J
Gottheil Marketing Communications Inc., 2005),
tobacco product retailing practices in New Zealand
40 have changed in response to increasing restrictions
on advertising, by increasing the role of retail
displays as a marketing method for tobacco pro-
ducts. Tobacco industry documents identify the
increasing importance of point-of-sale advertising
45 internationally (Lavack & Toth, 2006). In countries
Megan Quedley, Brendan Ng, Nikhil Sapre, Matthew Blakiston,
Alexandra Crawford, Rosanne Devadas, Hamish McLaren, Shruti
Anand, Mahu Tipu, Viswas Dayal, and Singithi Chandrasiri (medical
students), George Thomson, Ph.D., Richard Edwards, M.D.,
50
University of Otago, Wellington, New Zealand.
Correspondence: George Thomson, Department of Public Health,
University of Otago, Box 7343 Wellington South, New Zealand. Tel:
+1 (64)-4-385-5541 ext. 6054; Fax: +1 (64)-4-389-5319; E-mail:
george.thomson@otago.ac.nz
55
ISSN 1462-2203 print/ISSN 1469-994X online # 2008 Society for Research on Nicotine and Tobacco
DOI: 10.1080/14622200802238860
banned, point-of-sale displays (i.e., highly attractive
95
product displays at the point where the customer
pays for purchases) may now be the most important
communication channel between smokers, or poten-
tial smokers, and the tobacco industry.
The high visibility and ubiquitous accessibility of
100
tobacco products—in the most commonly visited retail
environments—helps reinforce smoker loyalty, sensi-
tizes youth to brands, and prompts impulse buying
(Lavack & Toth, 2006; Wakefield, Germain, Durkin,
& Henriksen, 2006). Some evidence indicates that
105
point-of-sale displays may specifically target children
(Feighery, Ribisl, Schleicher, Lee, & Halvorson, 2001;
Wakefield et al., 2006). Most cigarette smokers start
before the age of 16. Therefore, this age group is the
key potential market from which the tobacco industry
110
gets its next generation of consumers.
In many countries, governments have responded
to the increasing importance of point-of-sale displays
by introducing codes of practice or legislation
115

Nicotine & Tobacco Research ntr157423.3d 27/6/08 20:19:30

The Charlesworth Group, Wakefield +44(0)1924 369598 - Rev 7.51n/W (Jan 20 2003) 324053
 2 <
0 restricting the type, amount, and location of such
displays.
Evidence indicates that payments are made to New
Zealand retailers to ensure tobacco products are
displayed prominently (Forbes, 2005; Laugesen,
5 1999; Masters, 1999). The tobacco retail sector in
New Zealand includes supermarkets, dairies (small
general stores selling mainly tobacco, confectionary,
newspapers, magazines, and basic grocery items),
and convenience stores (small- to medium-sized
10 general stores with a similar range of products to
dairies but a wider selection of grocery items), service
stations (gas stations with a similar range of products
to convenience stores, mainly to supplement gasoline
sales and automotive products), pubs, tobacconists,
15 and some liquor outlets.
Smoking is a major public health problem in New
Zealand, with health equity problems due to the
much higher prevalence associated with lower socio-
economic status and Pacific Islander (ethnic groups
20 from Pacific Island states such as Samoa) and Maori
ethnicity. Maori are the indigenous people of New
Zealand, and they make up 15.3% of the New
Zealand population (Statistics New Zealand, 2007).
To help achieve the government’s aim ‘‘to reduce the
25
social approval of tobacco use, particularly among
young people’’, the 2003 New Zealand Smoke-free
Environments Amendment Act (SEAA) imposed
additional controls on the marketing, advertising, or
promotion of tobacco products, including requiring
30
tobacco retailers to comply with a series of point-of-
sale regulations (Table 1). In addition, Section 30(6) of
the SEAA requires tobacco retailers to have a notice
stating that the sale of tobacco products to those under
18 years is prohibited (New Zealand Ministry of
35
Health, 1990, 2003). Following passage of the SEAA
in 2003, the Ministry of Health sent one extensive
mailing to retailers, describing the new regulations.
Since then, little information has been provided to
40 retailers or the public about the point-of-sale display
regulations. Active enforcement of these regulations
has been minimal given that, due to resource
constraints, enforcement officers have been limited to 60
responding to complaints from the public. Fines of up
to NZ: 10,000 are possible for breaches of the regula-
tions, but we found no evidence of any prosecution
under these rules.
The international literature assessing retailer com- 65
pliance with point-of-sale display regulations is limited
(Letcher, Cameron, Inglis, & Wakefield, 2002), and we
were unable to find any studies conducted overseas or
in New Zealand that assessed retailer compliance with
national point-of-sale laws. We therefore decided to 70
assess compliance with the current regulations in New
Zealand and to examine whether compliance varied
with socioeconomic status, population density, ethni-
city, and the proportion of children in the local
population served by individual stores. 75
Our hypotheses were that there would be a higher
degree of noncompliance with point-of-sale regula-
tions among tobacco retailers in areas with (a) higher
socioeconomic deprivation, (b) a higher proportion
of Maori, (c) a higher proportion of Pacific Islanders, 80
or (d) a higher proportion of children.
Method
Store selection 85
We aimed to survey all tobacco retailers in the area
of the Greater Wellington Regional Council in the
lower North Island of New Zealand, excluding
specialist tobacconists and premises licensed to sell
90
alcohol. We omitted the latter two retailer types
because the regulations for the display of tobacco
products in tobacconists are different, and licensed
premises tend to sell tobacco only via vending
machines. These two types of premises sell less than
95
10% of the tobacco sold in New Zealand.
The Wellington Regional Council area has a
population of approximately 450,000, about 10% of
the New Zealand population (Statistics New Zealand,
2007). The area was chosen because it includes a range 100
of urban, suburban, small town, and rural areas and
was logistically practical for the survey.

Table 1. Point-of-sale regulations in New Zealand.

45 N The display of tobacco products at each ‘point of sale’ is limited to a maximum of 100 packages and 40 cartons, unless the retailer’s 105
place of business is a specialist tobacconist.
N The display must not be visible from outside the shop.
N Maximum of two packages of the same kind at the same point of sale (no block displays).
N The packaging and sale of tobacco with other products at a single price or at a reduced price is prohibited.
N Tobacco products may not be made available free of charge or with some kind of inducement or award or at a reduced rate other
50 than a normal trade discount. 110
N Tobacco products may not be displayed on the countertop or similar surfaces whether at point of sale or not.
N Tobacco products may not be displayed within 1 metre of ‘‘children’s products’’ such as confectionary and ice cream, soft drinks, and
products that are marketed primarily to children.
N If tobacco products are displayed within 2 metres of point of sale, a ‘‘Smoking Kills’’ sign must be displayed in clear view of the
customer at the point of sale and the sign needs to be at least 100 cm2.
N The maximum face size of any displayed tobacco product may not exceed the following dimension: package 66 cm2, pouch pack
55 105 cm2, carton 266 cm2. 115
N The legislation defines ‘‘point of sale’’ as ‘‘a checkout where tobacco products may be bought’’ or a cash register, even if it is not part
of a checkout.

Nicotine & Tobacco Research ntr157423.3d 27/6/08 20:19:31

The Charlesworth Group, Wakefield +44(0)1924 369598 - Rev 7.51n/W (Jan 20 2003) 324053

0 We used the online New Zealand Yellow Pages to
identify nonspecialist retailers, because this was the
most comprehensive available database of retail
outlets. We identified 300 retailers who were likely
to sell tobacco. Four were excluded because they
5 were no longer trading. Stores were allocated,
according to the judgment of the observers, to the
four categories of dairy, convenience store, service
station, or supermarket.
10
Data collection
The survey was conducted using a prepiloted check-
list, which included nine different categories match-
ing the point-of-sale regulations in the SEAA. The
15
checklist was developed after consultation with
enforcement officers responsible for investigating
complaints about violations of the point-of-sale
regulations within District Health Boards. We
randomly allocated each of 10 observers to a location
20
in which to survey all of the stores over a 1-week
period during August–September 2006. The obser-
vers visited stores unannounced and, while in the
store, judged whether the store was compliant or
noncompliant with each of the nine categories on the
25
checklist.
We assessed the tobacco product display at the store
point of sale (as defined in Table 1). For stores with
multiple points of sale (mainly supermarkets) where
there was variable noncompliance, we analyzed
30
compliance at the point of sale where noncompliance
was highest. For visibility of tobacco products from
outside a service station, we assessed this from outside
the forecourt, as this is the interpretation of the SEAA
legislation advised by the Ministry of Health. To
35
preserve the anonymity of the observers, the proximity
of tobacco products to children’s products, and of the
size and proximity of the ‘‘Smoking Kills’’ sign to a
point of sale, were visually estimated.
40 To test the interobserver reliability of our survey
instrument, we randomly selected 24 stores in the
Wellington City area and carried out a repeat
assessment by a different observer.
We used the Statistics New Zealand’s online map
45 (www.stats.govt.nz) to identify the location and
census area unit of each store, and we used the
2001 census to gather information about the level of
deprivation, ethnicity, population density, and age
distribution of the census area unit where each store
50 was located. The level of deprivation was based on
the NZDep 2001 Deprivation Index, which combines
nine census variables from the 2001 census that
reflect aspects of material and social deprivation in
the census area unit (Crampton, Salmond, &
55 Kirkpatrick, 2004). Gathering of area unit informa-
tion was done after the data collection in the stores
so that the observers were blind to this information.
Nicotine & Tobacco Research ntr157423.3d
The Charlesworth Group, Wakefield +44(0)1924 369598 -
NICOTINE & TOBACCO RESEARCH 3
Ethics approval was obtained through the 60
University of Otago ethical review system.
Data analyses
We used Intercooled Stata version 9.1 for the analyses. 65
During analyses, we grouped stores according to the
following potential determinants of compliance:
N By store type. Dairies and convenience stores were
grouped together in most analyses, because these 70
store types are similar in size and range of
products sold and their frequency of violations
of the point-of-sale regulations were similar.
N By decile of deprivation. Stores were grouped by
the NZDep score of their census area unit: (a) 75
deciles 1–4 (low deprivation), (b) deciles 5–7
(medium deprivation), and (c) deciles 8–10 (high
deprivation).
N By quartiles of the percentage of Maori and
Pacific Islanders in the population of the store 80
census area units.
N By quartiles of the percentage of children (aged
,19 years) in the population of the store census
area units.
N As urban or nonurban, accepting the definitions 85
of Statistics New Zealand for ‘‘main urban’’ and
anything else as ‘‘nonurban’’.
We calculated descriptive statistics for the dis-
tribution of breaches of point-of-sale regulations by
90
potential determinants of compliance. We investi-
gated the association between potential determinants
and compliance (any violations, >3 violations, and
the three most common individual violations) by
calculating crude and adjusted odds ratios using
95
logistic regression analysis.
We assessed the level of agreement and chance
corrected agreement (kappa statistic) in the reliability
study for each category of violation and the total
number of violations in each store. 100
Results
Response and store characteristics
105
We surveyed 288 out of 296 (97.3%) of eligible stores
that had tobacco products displayed for sale. Eight
stores were not surveyed due to observer error. The
288 stores were located in 102 census area units
across the lower North Island. The characteristics of 110
the stores are shown in Table 2.
Results of the reliability study
The overall level of agreement between observers for 115
each of the nine categories of possible violations
assessed varied between 70% and 100%. Analyses
27/6/08 20:19:31
Rev 7.51n/W (Jan 20 2003) 324053
 4 <
0 Table 2. Characteristics of the stores in the survey. The most common breaches of the regulations 60
were (a) failure to appropriately display a ‘‘Smoking
Deprivation level of census Number of stores
area unit of store (percent) Kills’’ sign (30% of all stores), (b) point-of-sale
display visible from outside the store (25%), (c)
Low (NZ Dep Classes 1–4) 81 (28)
tobacco products displayed within 1m of children’s
Medium (NZ Dep Classes 5–7) 88 (31)
5 High (NZ Dep Classes 8–10) 119 (41) products (24%), and (d) displaying more than 100 65
Urban vs. nonurban Number of stores (percent) packets per point of sale (16%). There was very low
Urban 225 (78)
Nonurban 63 (22)
noncompliance with the prohibition of countertop
Store type Number of stores (percent) displays (0.7%), copackaging or ‘‘bundling’’ of
Dairy 141 (49) tobacco (0.7%), or offering additional incentives to
10 Convenience store 35 (12) 70
Service station 83 (29) buy tobacco products (1.4%).
Supermarket 29 (10)
Ethnicity of store census area Median percent (range)
units Compliance with point-of-sale regulations by store type
Proportion of Maori 10.9 (4.9–38.5)
Proportion of Pacific Islander 4.1 (0.0–67.8) Compliance with the point-of-sale regulations was
15 Age distribution in store census Median percent (range) worst among dairies and convenience stores 75
area units
Proportion of children (age , 28.1 (12.1–44.5) (Table 3). This pattern was consistent across all
19 years) individual regulations except for the display of a
‘‘Smoking Kills’’ sign and for having more than two
packs displayed at the same point of sale, where
20 using kappa statistics were difficult to interpret due supermarkets were similarly noncompliant with 80
to the small number of violations in each category; dairies and convenience stores.
hence, high agreement was expected by chance. The The results of the logistic regression analysis for
most important finding was that the number of the association between the potential determinants
violations detected in the 24 stores increased from 27 and overall compliance with the point-of-sale regula-
25 in the main study to 38 in the re-survey. The increase 85
tions are shown in Table 4. The main finding was
was due to greater reporting of violations for the that the strongest predictor of any violations of the
categories ‘‘tobacco products on display within 1m of point-of-sale regulations was store type. Dairies and
children’s products’’ and ‘‘greater than two packs of convenience stores were much more likely to have
the same kind at the same point of sale’’. The first one or more breaches. The proportion of children in
30 category showed an increase of six, and the second an 90
the census area unit where the store was situated also
increase of five violations. was a statistically significant predictor when included
in the model as a linear term. Store type and
Compliance among all store types combined proportion of children in the census area unit were
35
predictors of stores having three or more violations. 95
Overall, 185 out of 288 stores surveyed (64.2%) had Dairies and convenience stores were much more
at least one breach of the point-of-sale regulations. likely to have multiple breaches of the regulations.
Half (92/185) of the stores with any noncompliance Ethnicity (as measured by the proportion of Pacific
breached one regulation, 29% (54/185) breached two Islanders and Maori in the census area unit) and area
regulations, and 21% (39/185) breached three or deprivation level were not significantly associated
40 100
more regulations. Five stores had five breaches with overall compliance.
each.

Table 3. Noncompliance by store type.

45 Convenience stores Dairies Service stations Supermarkets 105

(n535) (n5141) (n583) (n529) All stores

Degree of noncompliance
None 6 (18) 33 (24) 48 (56) 16 (53) 103 (36)
1 11 (32) 43 (31) 28 (33) 10 (33) 92 (32)
2 9 (26) 36 (26) 7 (8) 2 (7) 54 (18)
50 >3 8 (24) 27 (19) 2 (2) 2 (7) 39 (14) 110
Total 34 139 85 30 288
Type of noncompliancea
‘‘Smoking Kills’’ sign not displayed 11 (32) 51 (37) 12(14) 12 (40) 109 (38)
Visible from outside store 13 (38) 47 (34) 13 (15) 0 (0) 72 (25)
.100 packages 26 (32) 26 (19) 8 (9) 1 (3) 45 (16)
55 .2 packages in same display 7 (20) 16 (11) 3 (3) 5 (17) 30 (10) 115
,1 m to children’s products 13 (38) 47 (34) 6 (7) 2 (7) 66 (23)

Note. All values are numbers of stores with percentages. aOnly those categories with more than 10 total noncompliances are shown.

Nicotine & Tobacco Research ntr157423.3d 27/6/08 20:19:31

The Charlesworth Group, Wakefield +44(0)1924 369598 - Rev 7.51n/W (Jan 20 2003) 324053
 NICOTINE & TOBACCO RESEARCH 5

0 Table 4. Predictors of overall compliance with point-of-sale regulations. 60

Any violation At least three violations

Crude OR (95% CI) Adjusted ORa (95% CI) Crude OR (95% CI) Adjusted ORa (95% CI)

5 Store type 65
Service station 1.0 1.0 1.0 1.0
Supermarket 1.1 (0.5–2.6) 1.2 (0.5–2.9) 3.0 (0.4–22.0) 3.1 (0.4–24.0)
Dairy or convenience store 4.5 (2.6–7.8) 4.2 (2.3–7.6)* 10.5 (2.5–44.9) 7.3 (1.7–32.4)*
Deprivation level of census area unit
Low 1.0 1.0 1.0 1.0
High 0.8 (0.4–1.5) 1.0 (0.4–2.4) 1.7 (0.7–4.0) 1.6 (0.4–5.8)
10 Proportion of children in census area unit 70
1st quartile 1.0 1.0 1.0 1.0
4th quartile 1.9 (0.9–4.0) 1.7 (0.7–4.2)b 6.1 (2.2–17.2) 6.1 (1.6–22.7)*
Proportion of Maori in census area unit
1st quartile 1.0 1.0 1.0 1.0
4th quartile 0.6 (0.3–1.2) 1.1 (0.4–3.0) 0.8 (0.3–1.9) 0.7 (0.2–2.2)
Proportion of Pacific Islanders in census area unit
15 1st quartile 1.0 1.0 1.0 1.0 75
4th quartile 0.8 (0.4–1.6) 0.7 (0.3–1.9) 3.2 (1.2–8.8) 1.4 (0.3–6.1)

Note. OR5odds ratio; CI5confidence interval. aAdjusted odds ratios are adjusted for all other variables included in the table. bThe
proportion of children in the census area unit was a significant predictor of any violations when entered as a linear term in the adjusted
model (p5.04).
*Statistically significant associations in the adjusted odds ratios.
20 80

The results for the logistic regression analysis, for
predictors of compliance with the three regulations
that were breached most frequently, revealed a
25 similar pattern (Table 5).
For violation of the proximity of tobacco displays
to children’s products, the odds of a violation were
increased significantly for dairies and convenience
stores, and also for stores situated in census area
30 units with the highest proportion of children. The
odds of a violation in the visibility of tobacco
displays from outside the store were significantly
increased among dairies and convenience stores. For
appropriate display of a ‘‘Smoking Kills’’ health
35 warning sign, the odds of a violation were increased
among dairies and convenience stores, and for
supermarkets in comparison to service stations. The
odds of a violation of this regulation also were
increased in communities with an increased propor- 85
tion of Maori, although there was no association
between overall compliance and the proportion of
Maori in the census area unit (see Table 4).
We found no significant differences in the like-
lihood of breaches of the point-of-sale regulations 90
among shops (all subtypes) in urban (68%) versus
nonurban areas (63%).
As an illustration of these effects, the proportion
of stores noncompliant with three or more regula-
tions varied from 1 out of 43 (2%) service stations 95

Table 5. Predictors of compliance with commonly breached point-of-sale regulations.

Proximity to children’s products Visibility of display from outside ‘‘Smoking Kills’’ sign not visible
40 100
Crude OR Adjusted ORa Crude OR Adjusted ORa Crude OR Adjusted ORa
(95% CI) (95% CI) (95% CI) (95% CI) (95% CI) (95% CI)

Store type
Service station 1.0 1.0 1.0 1.0 1.0 1.0
Supermarket 0.9 (0.2–4.9) 1.0 (0.2–5.6) None visible None visible 4.1 (1.6–10.5) 4.3 (1.6–11.6)*
45 Dairy or convenience 7.0 2.9–17.0) 5.7 (2.3–14.3)* 2.9 (1.5–5.7) 3.4 (1.6–6.9)* 3.4 (1.7–6.7) 3.9 (1.9–8.1)* 105
store
Deprivation level of census area unit
Low 1.0 1.0 1.0 1.0 1.0 1.0
High 1.1 (0.6–2.1) 1.3 (0.5–3.7) 1.4 (0.7–2.6) 1.2 (0.5–3.2) 0.9 (0.5–1.7) 0.7 (0.3–1.7)
Proportion of children in census area unit
1st quartile 1.0 1.0 1.0 1.0 1.0 1.0
50 4th quartile 3.7 (1.7–7.9) 3.6 (1.4–9.3)* 0.9 (0.4–1.9) 0.7 (0.3–1.8) 1.2 (0.6–2.5) 0.8 (0.3–2.0) 110
Proportion of Maori in census area unit
1st quartile 1.0 1.0 1.0 1.0 1.0 1.0
4th quartile 0.5 (0.2–1.0) 0.4 (0.1–1.0) 1.3 (0.6–2.8) 2.1 (0.7–5.8) 2.3 (1.1–4.7) 5.6 (1.9–16.2)*
Proportion of Pacific Islanders in census area unit
1st quartile 1.0 1.0 1.0 1.0 1.0 1.0
55 4th quartile 1.8 (0.8–3.7) 1.2 (0.4–3.6) 1.2 (0.6–2.4) 0.9 (0.3–2.5) 0.6 (0.3–1.2) 0.5 (0.2–1.3) 115

Note. OR5odds ratio; CI5confidence interval. aAdjusted odds ratios are adjusted for all other variables included in the table.
*Statistically significant associations in the adjusted odds ratios.

Nicotine & Tobacco Research ntr157423.3d 27/6/08 20:19:32

The Charlesworth Group, Wakefield +44(0)1924 369598 - Rev 7.51n/W (Jan 20 2003) 324053
 6 <
0 and supermarkets, to 5 out of 43 (12%) dairies and
convenience stores in census area units with the
lowest quartile proportion of children, to 24 out of 56
(39%) dairies and convenience stores in census area
units with the highest quartile proportion of children.
5 The equivalent figures for violation of the regulation
on displaying tobacco products within 1m of
children’s products were 5/43 (12%), 10/43 (23%),
and 31/56 (55%), respectively.
10
Discussion
Main findings
In the present study all retailers surveyed were found
15 to be displaying tobacco products. This finding
illustrates the wide availability and visibility of
tobacco products in the New Zealand retail environ-
ment. The 64% rate of noncompliance with regula-
tions among the tobacco retailers we surveyed
20 suggests that the current point-of-sale regulations
are ineffective. Although the present study was
limited to the lower North Island, discussion with
enforcement officers from other jurisdictions in New
Zealand indicated that these results reflected obser-
25
vations in their own areas.
Interpreting the findings
We suggest that in this jurisdiction, two main factors
30
are contributing to the high rate of noncompliance: a
lack of awareness among retailers and the general
public regarding the tobacco point-of-sale regula-
tions, and an almost total absence of enforcement of
the act. At present, enforcement officers act only in
35
response to complaints from the public. This passive
enforcement system gives rise to a situation where
noncompliance is common and reporting of viola-
tions is unlikely. A license to sell tobacco is not
40
required in New Zealand; hence, there is no readily
available list of retailers, which makes systematic
enforcement difficult to carry out.
An interpretation of the findings with more
general application across similar jurisdictions is that
45 where tobacco retail displays are allowed, attempts
to control and limit them will be complicated, and
active enforcement may be expensive. Many aspects
of such displays need to be considered in making and
implementing such partial controls.
50 In the present study, dairies and convenience
stores were far less compliant than supermarkets
and service stations. This could be due to the
inherent differences between these two groups.
Supermarkets and service stations are part of larger
55 companies or franchises (as opposed to dairies and
convenience stores, which are often independently
owned); therefore, they are likely to have stricter
Nicotine & Tobacco Research ntr157423.3d 27/6/08 20:19:32
The Charlesworth Group, Wakefield +44(0)1924 369598 - Rev 7.51n/W (Jan 20 2003)
compliance standards expected of them by the 60
franchising company. Supermarkets and service
stations may have stronger incentives to comply with
all laws, so as not to generate bad publicity for their
parent company. In addition, tobacco sales may
make up a higher proportion of income for dairies 65
and convenience stores; thus they may have a greater
incentive to promote tobacco sales.
Our results demonstrated that dairies and conve-
nience stores in areas with the highest percentage of
children had the highest rates of noncompliance. 70
Whether this is by accident or by design is not clear
from the study. However, this finding is almost
irrelevant, given that the end result is the same: the
current New Zealand retail environment allows for
the widespread marketing of tobacco to children. 75
Research in other jurisdictions that suggests that the
density of retail outlets may increase the risk of youth
smoking, and that displays may be a particular
danger for youth, provides further support to these
findings (Feighery et al., 2001; Novak, Reardon, 80
Raudenbush, & Buka, 2006; Wakefield et al., 2006).
The present results are similar to those from an
observational study following the implementation of
retail tobacco promotion and display regulations in
85
Victoria, Australia. That study found high compli-
ance with many aspects of the new law 3 months
after its introduction but poor compliance with
regulations on the size of price tickets (38%
noncompliance) and the overall size of the tobacco
90
product displays (26% noncompliance), and moder-
ate (12%–13%) noncompliance with the display of
health warnings and signs to deter underage pur-
chases (Letcher et al., 2002). Interestingly, in the
Victoria study, more than half of the tobacco
95
products were displayed next to or behind confec-
tionary products.
Strengths and weaknesses
100
An important strength of the present study is the 97%
coverage rate within the sampling frame. We collected
data using a piloted checklist developed with the
assistance of enforcement officers who were experts in
the interpretation and monitoring of the SEAA 105
regulations. To minimize the introduction of observer
bias, surveyors were blind to the census area unit
demographics of each store at the time of the survey.
The study also had some potential weaknesses. We
excluded some possible participants by selecting the 110
study population from the yellow pages, which is not
a complete list of retailers. Potentially more partici-
pants were missed in this way than the eight missed
by observer error. However, we have little reason to
believe that any of those few retailers would be more 115
or less compliant, and these exclusions are unlikely to
have introduced significant bias.
324053

0 Another potential source for error in the study is
the use of visual estimates in some aspects of data
collection, particularly for proximity of tobacco and
children’s products displays in stores. However, the
results of the reliability study suggested that there
5 was a learning effect for observers during the study
and, if anything, we underestimated the degree of
noncompliance with the point-of-sale regulations in
the main survey.
Because the study excluded tobacco retailers in the
10 hospitality setting, such as bars, our ability to
extrapolate the study findings to this sector is limited.
However, only 5% of British American Tobacco sales
in New Zealand in 1997 were in hospitality outlets
(Wills New Zealand, 1997), and due to restrictions on
15 vending machines and smoking in bars since then,
this is unlikely to have increased.
Inadequacies identified in current New Zealand point-
of-sale legislation
20
In addition to widespread noncompliance, we identi-
fied two important loopholes during the fieldwork
that further undermine the purpose of the SEAA.
First, the law requires that no more than 100
25
packages be displayed per point of sale. However,
we observed many retailers with much larger displays
that were close to more than one point of sale and
thus did not contravene the letter of the regulations.
Several retailers with displays larger than 100 packs
30 appeared to be using a deliberate tactic to get around
the regulations over the size of displays, as they had a
redundant cash register, which was obviously not
plugged in or turned on, presumably so that it could
be designated as a second point of sale.
35 Second, in an attempt to dissociate children’s
products from tobacco, the SEAA states that
tobacco products cannot be within 1m of these
products. However, not only is this regulation
frequently breached, but even if a store is compliant,
40 the regulation does not prevent tobacco and chil-
dren’s products from being framed in the same visual
field. This allows a visual association and therefore
defeats the purpose of the 1m separation requirement
(Figure 1).
45
Policy implications
We believe the present findings have implications for
50 policies to regulate retail display of tobacco in New
Zealand and internationally. We found that, despite
considerable strengthening of the legislation on
point-of-sale displays, tobacco products were almost
ubiquitously on prominent display in key retail
55 environments and that there was widespread non-
compliance with even the limited point-of-sale dis-
play restrictions in place.
Nicotine & Tobacco Research ntr157423.3d
The Charlesworth Group, Wakefield +44(0)1924 369598 -
NICOTINE & TOBACCO RESEARCH 7
60
65
70
E-COLOUR
FIGURE
Figure 1. Example of a tobacco display that is not within 75
1m of children’s products but is within the same visual
field as the children’s products.
We therefore propose two frameworks to
strengthen the current New Zealand regulation of 80
tobacco retail displays, and which might also be
relevant in other countries and jurisdictions.
Framework 1. Permit point-of-sale displays but 85
strengthen the regulation and enforcement mechan-
isms and retailer education. Suggested changes could
include the following:
N Prohibit the combination of tobacco displays from
two or more points of sale. 90
N Dissociate tobacco more clearly from children’s
products by introduction of a clause prohibiting
tobacco and children’s products from being in the
same line of sight at point of sale, and increasing
95
the distance to at least 2m.
N Communicate more about the regulations to the
public and retailers, and introduce active enforce-
ment of the act with appropriate resource allocation.
This suggested framework would be an improve- 100
ment on the current situation; however, it would be
practically difficult and expensive to implement and
enforce this policy (particularly ensuring tobacco
products and children’s products are not in the same
105
line of sight), and tobacco products would remain
displayed in the most visited retail environments.
Framework 2. Introduce a complete ban on point-of-
sale displays. We suggest that this would be an easier 110
framework to enforce and would remove any
ambiguity regarding the legality of a display. There
are similar frameworks in Iceland, Thailand, and
several Canadian provinces such as Saskatchewan
and Manitoba (Hamann, 2005; Tilson, 2004). This 115
approach may not be possible in all countries, given
constitutional and other legal constraints, but its
27/6/08 20:19:33
Rev 7.51n/W (Jan 20 2003) 324053
 8 <
0 successful implementation in these jurisdictions sets a
precedent for many countries.
In both frameworks we recommend the introduc-
tion of tobacco sale licensing. This would facilitate
easier monitoring of compliance and enforcement of
5 the regulations, and it would allow for licenses to be
withdrawn from retailers who breach the regulations.
Licensing fees also could be used to fund implemen-
tation, education, and enforcement, thus increasing
compliance and making the legislation sustainable. If
10 licensing were used to limit the types or numbers of
outlets selling tobacco, then the research findings
indicate that controlling displays or a display ban
may be easier in outlets that are part of larger
companies or franchises.
15
Conclusion
Our findings suggest that where point-of-sale regula-
20
tions are introduced, they need an appropriate
monitoring and enforcement mechanism. Otherwise,
noncompliance will be high. Given the contradiction
between allowing tobacco product displays and
preventing youth exposure to tobacco marketing,
25
we believe that the best policy option is to ban all
retail tobacco product displays and introduce a
licensing system for tobacco retailers.
30 Acknowledgments
The authors thank Pam Smith of the Hutt Valley DHB Regional
Health Unit for her valuable contributions, and Janet Hoek of Massey
University, for her perspectives on marketing. Funding was provided
by the New Zealand Cancer Society to cover transport costs of the
observers. No other financial contribution was made, and there was no
influence on the study design or implementation.
35
Declaration of interest: The authors report no conflicts of interest. The
; authors alone are responsible for the content and writing of the paper.
References
40
Apte, U., & Viswanathan, S. (2002). Strategic and technological
innovations in supply chain management. International Journal of
Manufacturing Technology and Management, 4, 264–282.
45
50
55
Nicotine & Tobacco Research ntr157423.3d 27/6/08 20:19:36
The Charlesworth Group, Wakefield +44(0)1924 369598 - Rev 7.51n/W (Jan 20 2003)
Bloom, P. (2001). Role of slotting fees and trade promotions in shaping 60
how tobacco is marketed in retail stores. Tobacco Control, 10,
340–344.
Carter, S. M. (2003). New frontier, new power: The retail environment in
Australia’s dark market. Tobacco Control, 12(Suppl. 3), iii95–iii101.
Crampton, P., Salmond, C., & Kirkpatrick, R. (2004). Degrees of
deprivation in New Zealand: An atlas of socioeconomic difference.
Auckland, New Zealand: David Bateman Ltd. 65
Dewhirst, T. (2004). POP goes the power wall? Taking aim at tobacco
promotional strategies utilised at retail. Tobacco Control, 13,
209–210.
Feighery, E. C., Ribisl, K. M., Clark, P. I., & Haladjian, H. H. (2003).
How tobacco companies ensure prime placement of their advertising
and products in stores: Interviews with retailers about tobacco
company incentive programmes. Tobacco Control, 12, 184–188. 70
Feighery, E. C., Ribisl, K. M., Schleicher, N., Lee, R. E., &
Halvorson, S. (2001). Cigarette advertising and promotional
strategies in retail outlets: Results of a statewide survey in
California. Tobacco Control, 10, 184–188.
Forbes, S. (2005). Big cash for smoke-sellers. Western Leader.
[Auckland], May 10. 75
Hamann, S. (2005). Thailand’s point of sale ban complete [e-letter].
Tobacco Control, December 16.
J Gottheil Marketing Communications Inc. (2005). The influence of
tobacco powerwall advertising on children. Retrieved May 10, 2007,
from www.nsra-adnf.ca/cms/file/pdf/tobacco_powerwalls.pdf
Laugesen, M. (1999). Rothmans pays retailer $1700 yearly to display
cigarettes. NZ Smokefree News, 3(April 19). 80
Lavack, A. M., & Toth, G. (2006). Tobacco point-of-purchase
promotion: Examining tobacco industry documents. Tobacco
Control, 15, 377–384.
Letcher, T., Cameron, M., Inglis, G., & Wakefield, M. (2002). Bans on
tobacco advertising at point of sale. Quit Victoria Research and
Evaluation Studies No. 11: 2000–2001. Melbourne: Quit Victoria.
85
Masters, C. (1999). Cigarette displays have Ash fuming. New Zealand
Herald [Auckland], May 3, p. A1.
New Zealand Ministry of Health. (1990). Smoke-free Environments Act
1990. Wellington: Author.
New Zealand Ministry of Health. (2003). Smoke-free Environments
Amendment Act 2003. Wellington: Author.
Novak, S. P., Reardon, S. F., Raudenbush, S. W., & Buka, S. L. 90
(2006). Retail tobacco outlet density and youth cigarette smoking: A
propensity-modeling approach. American Journal of Public Health,
96, 670–676.
Statistics New Zealand. (2007). Regional council by sex and age group for
the census usually resident population count: 2006 Census. Retrieved
October 2, 2007, from www.stats.govt.nz/nr/rdonlyres/19d6a4b9-7d98-
4af9-970d-5bbd139794ca/0/regionalsummarytablesregionalcouncil.xls 95
Tilson, M. (2004). Restrictions on the retail display of tobacco products.
Retrieved September 19, 2007, from www.smokefreens.ca/
displayban.pdf
Wakefield, M., Germain, D., Durkin, S., & Henriksen, L. (2006). An
experimental study of effects on schoolchildren of exposure to
point-of-sale cigarette advertising and pack displays. Health
Education Research, 21, 338–347. 100
Wills New Zealand. (1997). Wills New Zealand trade marketing plan
1997–1999. Retrieved April 20, 2007, from http://bat.library.ucsf.edu/
tid/atr53a99
105
110
115
324053