Evaluation of Retailers' Behaviour Towards Certain Youth Access-to-Tobacco Restrictions (Final Report of Findings: 2006)

Introduction

Preface

This is the latest in a series of annual studies designed to evaluate the behaviour of retailers towards youth access-to-tobacco restrictions. Health Canada has been using independent researchers since 1995 to evaluate retailer behaviour towards youth access restrictions of the federal and provincial tobacco laws.

The federal Tobacco Act sets the minimum federal standards in Canada, prohibiting the furnishing of tobacco products to minors. Among key provisions, the legislation makes it illegal for retailers to furnish tobacco products to anyone under the age of 18 and provides for defence of due diligence for retailers who require photo identification to confirm a customer's age. The Act also restricts the distribution of advertising and sets minimum mandatory sign requirements at point of sale.

Individual provinces have in place their own regulations prohibiting the sale of tobacco products to minors. These may stipulate rules for the display of tobacco products differently from the federal minimum, may require sign-posting requirements different from the federal provision, and may restrict the sale of tobacco products to minors older than 18. Indeed, except for Quebec, Manitoba, Alberta and Saskatchewan, retailers in all other provinces are restricted by local law from selling tobacco products to anyone under 19 years of age.

At the time of this research, Alberta, Manitoba and Quebec were the last remaining provinces where tobacco retailers are monitored against the basic federal guidelines, even though Quebec put its own tobacco legislation in place several years ago. Observers in these three provinces were instructed to monitor retailer compliance consistent with the age and sign provisions of the federal Tobacco Act. Any additional signs were noted, if found.

In March 2002, Saskatchewan's  Tobacco Control Act took effect and removed retail displays from stores. The law required vendors to use curtains and cabinets to hide cigarettes, cigars and chewing tobacco behind the counter, or ban minors from their stores. The legislation banning power walls was unsuccessfully challenged in the Saskatchewan Court of Appeal in 2004. Therefore, after a brief reprieve from the enforcement of the law that year, the law was re-instated in January 2005. Observations made of retailers in Saskatchewan for 2006 continued to be completed in the manner in which they were conducted in 2005. Manitoba also adapted a similar new legislation in 2005 with regard to retail displays and was observed accordingly in 2006.

This study was designed to permit direct comparisons with the results of last year and, as much as possible, with similar data collected in years prior. As such, the current data can be compared with statistical validity with past results. Where adjustments to the sample were made or where legislative changes necessitated adjustments to work requirements, these differences affecting the historical trend are noted at the appropriate places in this document.

Research Objectives

Health Canada's fundamental objective in commissioning this research is to maintain and enhance the health and safety of young Canadians. The data we have collected is strictly for information and evaluation purposes, and has not been commissioned for purposes of enforcement.

The work is important within the context of the Federal Tobacco Control Strategy (FTCS). The FTCS embodies the federal government's latest initiatives to combat tobacco use in Canada. Among these, it sets clear and ambitious ten-year targets for tobacco control, including the goal of increasing retailer compliance with tobacco-sales-to-youth laws to 80%. The FTCS specifies the on-going annual evaluation of retailer compliance as one of five strategic objectives for the coming years. The findings from the current study build on the results of annual retailer behaviour measurements since 1995, at the same time establishing new benchmarks against which future progress can be measured and compared.

This year's study had three specific measurement objectives in mind:

1. To monitor the behaviour of retailers relative to the sales-to-minors provisions of the Tobacco Act and related provincial legislation, including aspects of the laws relating to the posting of mandatory age restriction/health-warning signs;
   
   
2. To collect information on tobacco point-of-sale (POS) merchandising; and
   
   
3. To determine the in-store display of "Operation ID" materials.

Under each individual objective, estimates were collected of the following:

1. Sales-to-Minors and Sign-Posting Provisions of the Tobacco Laws
   
   
   • number and percent of retailers refusing to sell tobacco to persons under legal age;
     
     
   • number and percent of retailers asking customers for proper identification as required by the law;
     
     
   • number and percent of retailers posting all mandatory age advisory and/or health-warning signs prescribed by the law;
     
     
   • number and percent of retailers posting each of these signs in the proper manner and location prescribed by the law, although position was not taken into account for compliance; and
     
     
   • number and percent of retailers in partial or complete non-compliance with the sign-posting provisions of applicable federal or provincial laws, including provinces where some signs are unauthorized to be posted, with an indication of the number and percent of retailers posting which signs by type.
     
     
2. Retail Point-of-Sale (POS) Merchandise
   
   
   • in all trade classes, point-in-time information regarding the distribution, number and type of retail POS materials in support of the tobacco category as a whole; and
     
     
   • in convenience stores (both chain and independent) and gas stores/kiosks, the above-indicated POS measures by brand as follows: Belvedere, Benson & Hedges, Canadian Classics, du Maurier, Export A, Export A Smooth, Remaining Export A, Matinee, Players, Rothmans, Sportsman, "Other".
     
     
3. "Operation ID" Displays
   
   
   • the number and percentage of establishments displaying "Operation ID" POS material of any sort.

Methodology

Teams made up of two evaluators, one a minor (15, 16 or 17 years of age) and the other an adult over 19 years of age, were sent into a randomly selected, stratified sample of 5,605 retail establishments in 30 cities and towns across Canada. Stores were visited over nine weeks, from June 19, 2006 to August 14, 2006.

Team members entered stores at different times. They gave no indication of being together. Each carried out specifically assigned tasks:

• The responsibility of the teen researcher was to gauge the willingness of retailers to sell him/her tobacco by actually attempting to buy a 20 or 25 size pack of name-brand cigarettes. The teenagers did not carry identification. During the attempted transaction, minors made no misleading statements other than if asked their age. If asked their age, they were not truthful, but rather claimed to be 18 or 19 years old, depending on the minimum age requirements of that province. Under no circumstances did they make a purchase. They were given clear instructions about how to casually back out of any attempted sale; and
  
  
• The senior member of the research team was responsible for supervising the younger partner and for carrying out a visual inspection of the retailer's place of business for the purpose of observing and recording compliance with the posting of mandatory signs under the federal Tobacco Act or similar provincial legislation. These people were also responsible for collecting information on in-store tobacco advertising and promotions and "Operation ID" displays.

In order to measure the influence of age and sex on retailers' propensity to comply with sales-to-minors legislation, visits were organized so as to ensure the following minimum national requirements:

• that store visits were conducted by a mix of both male and female teens in roughly equal proportions; and
  
  
• that approximately half of all store visits were completed by a minor 17 years of age, with the rest of the sample being divided about equally between 15 and 16 year olds.

These distribution targets are the same as those established for studies conducted since 1998.

Table i confirms national completion rates by age and sex of teen researchers for the current study and preceding ones to 2002.

Table i - National Sample Dispersion by Age and Sex of Teen Researcher

	2002	2003	2004	2005	2006
Sample Size (# Store Visits)	5,550	5,452	5,516	5,561	5,605
Male	49.4%	49.6%	49.7%	50.1%	50.0%
Female	50.6%	50.4%	50.3%	49.9%	50.0%
15 year old	23.7%	23.8%	24.7%	26.6%	25.5%
16 year old	29.3%	29.4%	24.4%	26.5%	26.4%
17 year old	46.9%	46.8%	50.9%	46.9%	48.1%

As much as possible, we tried to respect these targets within individual cities and across retail classes of trade. However, as in past years, there were cities and store types this year where the number of sample stores was too small to accommodate a full mix of boys and girls across each age group. In these cases, the sample was assigned to one or more teens of pre-determined age and sex. The distributions were similar to the criteria used in the last study of this type, in 2005.

Scope

Retail stores in 30 urban markets were visited. The list of all cities includes at least one city from each province in Canada.

Sample

A targeted national sample of retailers was selected from best-available universe estimates across each of the chosen cities. A completely new sample was randomly chosen for the latest study.

This year, pharmacies were removed from the sample due to the increased number of provinces banning the sale of tobacco in pharmacies and the voluntary ban in pharmacies in provinces where it is still allowed. The sample was redistributed among the remaining four classes of trade.

The sample frame was designed to ensure equal chances of representation for retailers in each of four classes of trade:

• Chain convenience
• Gas stores/kiosks
• Grocery
• Independent convenience

Sample selection was guided by defined statistical procedures used in previous studies to select samples the first time in each city. For each city and for each trade class, the sample frame was organized by banner, where appropriate, and geographically distributed across each city based on postal codes. The frame was then divided into sub-strata requiring a minimum sample of one, and a store was randomly selected from each sub-stratum.

In all cities for 2006, our sampling objective was to match sample distributions by city and class of trade as closely as possible to those established in previous years. Samples were drawn to ensure representation of the census metropolitan area (CMA) of the largest cities. For Vancouver, it was requested that statistically reliable measures be provided for the entire CMA and separately for each of four different health territories. This required the area to be over-sampled.

The four health regions within the Greater Vancouver Area were defined by the boundaries of communities as follows:

The four health regions within the Greater Vancouver Area

Region 1	Region 2	Region 3	Region 4
Vancouver	Burnaby	Delta	North Vancouver
Richmond	Coquitlam	Surrey	West Vancouver
	Port Moody	Langley
	Port Coquitlam	North Delta
	Maple Ridge	White Rock
	Whonnock
	Pitt Meadows
	New Westminster

Once fieldwork began, some selected sample stores were found to have closed or to no longer be selling tobacco products. In these cases, the stores were replaced with other retail establishments of the same type located in close proximity to the original store(s). When additional stores of the same type that sold tobacco could not be found, the original sample store was replaced by a store from another class of trade in the same area.

In the end, we completed visits to 5,605 stores, geographically dispersed largely in a manner consistent with our pre-established targets. Table ii, at the end of this section, shows the actual number of completes across the nation.

Sample Weighting

Due to the stratified random sample, four class trades per region, raw level data from our sample outlets has been weighted statistically to reflect the distribution of total outlets, in the universe within the 30 communities and, subsequently, the provinces and across the country. Weighted data is an estimate of conditions we likely would have found had we visited every retail outlet in the 30 cities instead of just the outlets in our sample. In no instance has an attempt been made to project results beyond the retail universe in the urban areas sampled.

Unless otherwise mentioned, all the percentages quoted in this report are results after weightings have been applied.

NOTE: Readers attempting to replicate weighted percentage figures using raw sample sizes will yield different results from those generally reported.

Understanding this Report

The survey is not intended as a rating of cross-jurisdictional performance. The design of this research is intended to yield reasonable estimates of retailer compliance under specified conditions at the national level. The regional data is useful for understanding the national trend, but one must be careful about drawing comparisons between regions.

For practical reasons, it was not possible to impose in all cities the same controls for age and sex of teens that we did nationally. This will affect the results in the regions. At the regional level, the findings should be used for other considerations:

• to debate best practices between regions;
• to track the progress and general direction of retailer compliance with the tobacco laws over 11 surveys since 1995;
• to learn from the national results what variables motivate retailer behaviour; and
• to consider ways to apply these lessons in the provinces.