Tobacco Industry Marketing at Point of Purchase After the 1998 MSA Billboard Advertising Ban

Point-of-purchase marketing has become increasingly important for the tobacco industry in the United States.1 In the wake of the 1998 master settlement agreement (MSA) that required tobacco advertising on billboards across the country to end on April 24, 1999, the point-of-purchase environment is likely to assume even greater importance in the industry's marketing efforts.

One goal of the billboard advertising ban (as well as other MSA advertising and promotion restrictions) was to reduce youth exposure to cigarette advertising. However, previous research suggests that the tobacco industry is able to compensate for an inability to advertise in one medium by transferring advertising dollars to other marketing activities.2–6 Accordingly, there is concern that the MSA billboard advertising ban may merely shift tobacco advertising funding to other efforts, such as point-of-purchase marketing. In this study, we used data from a unique national sample of retailers to explore changes in the point-of-purchase environment after implementation of the billboard tobacco advertising ban.

METHODS

Sample Selection

Data were obtained through the activities of ImpacTeen, a policy research partnership (supported by the Robert Wood Johnson Foundation) focused on reducing substance use among young people. The sampling strategy (described elsewhere7) for the 193 communities involved in the partnership was primarily determined by the locations of public schools drawn as part of nationally representative samples of students in the 8th, 10th, and 12 grades in the coterminous United States.

Catchment areas (communities) reflecting the area from which each school drew the majority of its student population were defined. Up to 30 retailers in each community were randomly selected for observation; in instances in which there were fewer than 30 retailers, all were included. The final sample of retailers (n = 3553) varied from 2 to 34 (mean = 18.4) per community. Of these establishments, 36 were excluded because they did not sell tobacco, and a further 55 were excluded owing to missing covariate values; thus, analyses involved a maximum of 3462 tobacco retailers in 191 communities.

Data Collection and Analysis

Observations were conducted between February 16 and June 23, 1999. Trained field staff teams unobtrusively collected information on tobacco product placement, extent of related advertising and promotions, and prices.

Levels of store interior, store exterior, and parking lot tobacco advertising were measured with a 4-point scale ranging from no advertising to advertising covering most of the store or storefront. Prevalence of low-height (less than 3.5 ft [105 cm]) ads was noted, and the number of tobacco-related functional objects (i.e., items owned by the store, such as clocks and shopping baskets, that are labeled with a cigarette brand) was recorded. The prevalence of a variety of promotions, including multipack discounts and gifts offered with purchases, was recorded.

SAS (version 6.12) was used in conducting logistic regression analyses that attempted to determine the effect of the ban on the prevalence of store interior, store exterior, and parking lot tobacco advertising, tobacco promotions, and functional objects. Cumulative logit analyses were used to examine the relationship between ban date and extent of interior, exterior, and parking lot advertising and number of functional objects.

RESULTS

Table 1 ▶ shows characteristics of the sample. Observations were made in 1484 stores (43%) before April 24, 1999, and 1978 stores (57%) were observed on or after that date.

TABLE 1.

—Characteristics of Communities and Retailers

Characteristic	Retailers, % (n=3462)
Population density
Large city	16.3
Midsize city	11.0
Urban fringe, large city	33.0
Urban fringe, midsize city	12.7
Large/small town	15.9
Rural	11.0
Tobacco control
States without programs	70.5
States with programsa	29.5
Store type
Convenience	12.4
Convenience/gas	34.5
Gas station	7.5
Mom/pop store	3.3
Grocery store	9.1
Supermarket	9.0
Drug store	9.6
Liquor store	8.5
Tobacco store	2.0
Other	4.1
Store size, No. of cash registers
1	61.0
2	17.2
≥ 3	21.7

^aArizona, California, Florida, Massachusetts, Maine, Oregon.

Table 2 ▶ indicates that, in comparison with tobacco retailers observed before the billboard ban, those observed thereafter evidenced significant increases in the prevalence of interior tobacco advertising and the prevalence and extent of exterior tobacco advertising. Highly significant increases were found in the prevalence of all 3 types of cigarette promotions and the number of tobacco-related functional objects. However, there was no significant change in the extensiveness of interior advertising, low-height advertising, or parking lot advertising.

TABLE 2.

—Regression Analyses: Association of Tobacco Advertising and Promotion Variables With Date of Ban

Dependent Variable	No. of Stores	Preban, %	Postban, %	Unadjusted OR (95% CI)	P	Adjusted ORa(95% CI)	P
Interior ads, any vs none	3424	76.0	79.6	1.23 (1.05, 1.45)	.012	1.27 (1.06, 1.52)	.011
Interior ads	3424			1.09 (0.95, 1.25)	.210	1.08 (0.94, 1.25)	.292
Free of any ads		24.0	20.4
Ads limited to where sold		57.5	62.1
High levels of adsb		18.6	17.6
Low-height tobacco ads	2646	44.3	43.3	0.96 (0.82, 1.12)	.592	1.02 (0.86, 1.20)	.846
Exterior ads, any vs none	3401	55.2	59.9	1.22 (1.06, 1.40)	.006	1.22 (1.03, 1.44)	.020
Exterior ads	3401			1.29 (1.13, 1.46)	.000	1.30 (1.12, 1.50)	.001
Free of any ads		44.8	40.1
< 5 ads, each < 30 cm in any dimension			20.3	17.5
High levels of adsc		34.9	42.4
Parking lot ads, any vs noned	1421	41.1	39.8	0.95 (0.76, 1.17)	.613	1.02 (0.81, 1.27)	.877
Parking lot adsd	1421			1.10 (0.89, 1.35)	.396	1.19 (0.95, 1.47)	.128
Free of any ads		58.9	60.3
< 5 ads, each < 30 cm in any dimension			15.6	6.2
High levels of adsc		25.5	33.6
Promotions, any vs none	3414	43.3	52.0	1.42 (1.24, 1.63)	.000	1.65 (1.42, 1.92)	.000
Promotions, specific types
Multipack promotions, 1 or more	3424	23.4	27.1	1.22 (1.04, 1.43)	.014	1.38 (1.17, 1.64)	.000
Gift-with-purchase promotions, 1 or more	3423	3.8	8.5	2.36 (1.73, 3.23)	.000	2.51 (1.81, 3.47)	.000
Cents-off promotions, 1 or more	3415	32.3	40.4	1.43 (1.24, 1.64)	.000	1.65 (1.41, 1.92)	.000
Functional objects, any vs none	3434	65.9	72.8	1.38 (1.19, 1.60)	.000	1.63 (1.38, 1.93)	.000
Functional objects	3434			1.36 (1.20, 1.54)	.000	1.57 (1.38, 1.79)	.000
None		34.1	27.2
1–2		33.7	33.6
3–4		17.7	20.9
≥ 5		14.6	18.2

Note. OR = odds ratio; CI = confidence interval.

^aAdjusted for store type, number of cash registers (proxy for store size), urbanicity of community, and presence of statewide tobacco control program in April 1999.

^bCombination of “has ads in sections of the store distinctly separate from where product sold” and “has ads covering almost all available space throughout the store.”

^cCombination of “had less than 5 ads, but one or more is larger than 30 cm in any dimension” and “has 5 or more ads.”

^dIncludes only stores with parking lots (gas stations and convenience stores selling gas), n = 1454.

DISCUSSION

The increases observed here in point-of-purchase marketing over a relatively short period of time probably reflect responses to changes in the tobacco marketing regulatory environment, particularly the prohibition of billboard advertising and restrictions on event sponsorships and related advertising (e.g., auto racing, concerts) after implementation of the MSA provisions. Our results are consistent with those of other research demonstrating that as long as tobacco advertising restrictions are incomplete, significant reductions in overall marketing efforts are unlikely to be achieved. This explains why only comprehensive advertising bans are associated with reductions in smoking.6 Point-of-purchase marketing organizations have noted that the billboard ban is expected to translate into millions of extra revenue dollars for point-of-purchase marketers.8

Point-of-purchase advertising increases are of particular concern to those seeking reductions in smoking among teenagers. There is growing evidence that cigarette advertising and promotions increase youth smoking9–15 and that youths are more responsive to such marketing than are adults.4 Research shows that tobacco advertising has both predisposing and reinforcing effects on youth smoking, acting as an inducement to experimentation with smoking as well as reinforcing continued progression toward regular smoking.12 For example, one study showed that, in comparison with students who saw pictures of stores with no tobacco advertising, students exposed to photographs of stores with tobacco ads perceived that tobacco was significantly easier to acquire, believed more of their peers had tried and approved of smoking, and expressed weaker support for tobacco control policies such as advertising restrictions and cigarette price increases.16

Also, a merchant intervention study conducted in Baltimore, Md, showed that youths were more likely to attempt cigarette purchases in stores with exterior cigarette advertising depicting models who were youthful in appearance than in stores without similar ads.17 In that 3 of 4 teenagers visit a convenience store at least once per week,18 these research studies suggest that the point-ofpurchase environment may have important influences on youths in terms of making tobacco use seem normative and, ultimately, increasing the likelihood of smoking initiation.

In conclusion, evidence suggests that point-of-purchase advertising and promotions have increased since implementation of the MSA billboard tobacco advertising ban. These increases, at least in part, are likely to have resulted from the shifting of resources once spent on billboard advertising to other marketing efforts. As a result of this shift, the intended effect of the billboard advertising ban may not be realized, because overall exposure to advertising and promotions may not be reduced. Further research is needed to examine the impact of the billboard ban and other MSA restrictions on tobacco company marketing strategies and on youth and adult smoking.