Health Education Research Advance Access originally published online on May 4, 2006
Health Education Research 2006 21(3):378-385; doi:10.1093/her/cyl007
Compliance with a minimum-age law of 18 for the purchase of tobacco—the case of Sweden
1 Public Health, Karlstad University, SE-651 88 Karlstad, Sweden and
2 Public Health, Karlstad University, SE-651 88 Karlstad, Sweden
*Correspondence to: M. Sundh. E-mail: mona.sundh{at}kau.se
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
The purpose was 2-fold: first to analyse how compliance with the minimum-age law of 18 for the purchase of tobacco has changed over time, second to determine what factors affect the possibility for adolescents to buy cigarettes. Three series (1996, 1999 and 2002) of test purchases of tobacco were conducted in three regions of Sweden. Altogether, 36 test purchasers carried out 1867 test purchases. In 1996, 93% of all attempted purchases resulted in the purchase of cigarettes. The corresponding figures for 1999 were 80% and for 2002 63%. The results indicate significant regional differences in the test purchasers' opportunities to buy cigarettes. This paper demonstrates that compliance with the minimum-age law has had an effect on the ease with which adolescents purchase cigarettes, and further that such purchase has become more difficult. The study also indicates great potential for local efforts to induce more retailers to comply with the law.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Measures to prevent adolescents from starting to smoke, and to encourage those who do smoke to stop, are important parts of the anti-tobacco campaigns being waged in many countries. Reducing the availability of tobacco products, high tobacco prices, clean indoor-air laws, advertising restrictions and health education are key elements in this strategy.
The availability of tobacco products affects whether adolescents start smoking [1], and an effective minimum-age law thus serves a vital function by making it more difficult for adolescents to obtain tobacco. Such a law may influence both the onset of smoking and the smoking habits of young frequent smokers [2, 3]. In many settings, tobacco sales restrictions have worked well to reduce youth smoking, but in other settings they have not [4–6]. Tobacco vending machines may be an obstacle to controlling compliance with the law [7]. In Sweden, however, tobacco vending machines represent only a small percentage of total sales. One reason why the age-limit law does not work as intended may be the opportunities available to adolescents to obtain tobacco in ways other than from shops [8, 9].
Empirical reports on reduced youth access to tobacco are based largely on interventions and experience in the United States (for a review of the US case, see Levy and Friend [7]). Although all the Nordic countries have a minimum-age law for purchasing tobacco, few studies have evaluated the efficiency of these laws.
In the 1990s, the demand for restrictions on tobacco sale increased significantly in Sweden and in 1996 two large supermarket chains introduced a voluntary age limit of 18. The following January, the age limit of 18 became law, covering both cigarettes and smokeless tobacco (Swedish ‘snus’, which is moist tobacco tucked under the lip). The Swedish minimum-age law stipulates that ‘Tobacco products may not be sold or in any other way conveyed commercially to those who are under 18 years of age. Those who supply tobacco products shall ensure that the recipients have reached the stated age. If there is a special reason to believe that a tobacco product is intended for distribution to anyone who has not reached 18 years of age, the product shall not be supplied’ (Swedish tobacco law 1993:581 
12). After 1997, the law has been made strict on several occasions.
Supervision of the law rests with the Swedish National Institute of Public Health (centrally), the county administrative boards (regionally) and the municipalities (locally). A minor who attempts to purchase tobacco cannot be punished. The responsibility for age control rests solely on the cashier, who can be sentenced to a fine or imprisonment for a maximum of 6 months.
An earlier Swedish study [10] concerning adolescents' opportunities to purchase tobacco before, and 2 years after, the introduction of the minimum-age law showed a limited effect of the law on the possibility for adolescents to purchase tobacco. The compliance with the law had obvious shortcomings, in particular with respect to the procedure of checking the age of adolescents who wants to buy tobacco. In 1999, age checks were carried out only in 22% of the purchase attempts. The present study provides data on the implementation of the law 5 years after its introduction in 1997, permitting evaluation of longer-term effects.
The purpose of this study was 2-fold: to analyse how compliance with the minimum-age law for the purchase of tobacco has changed over time and to determine what factors affect adolescents' opportunities to buy cigarettes.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Materials
Test purchases were used to collect the data in this study. The method used is based mainly on one developed in Finland [11], which in turn was influenced by a method developed in the United States [12].
Study objects
Three series of test purchases, in 1996, 1999 and 2002, were conducted in three regions of Sweden: the city of Malmö in southern Sweden (267 000 inhabitants), the rural county of Västernorrland in mid-northern Sweden (244 000 inhabitants) and Värmland county in mid-southern Sweden (273 000 inhabitants). These regions were chosen in order to include different kinds of regions at different geographical locations. Malmö is Sweden's third largest city, located across the Sound some kilometres from Copenhagen. Värmland is located 
500 km north of Malmö and Västernorrland is located 1000 km northeast of Malmö.
Each year, 250 test purchases of tobacco were carried out in each region, giving a total of 2250 test purchases. Throughout, attempts were made to make purchases on weekdays, in different places and from various types of sales outlet. There were too few test purchasers to target all municipalities in the regions. Therefore, municipalities located far away from the regional centres were not targeted. In Värmland, 11 of 16 and in Västernorrland five of seven municipalities were targeted. For the same reasons, not all districts in Malmö were covered. No complete registers of tobacco sales outlets were available but the strategy was to visit all sales outlets (for example, all department stores, all after-hour supermarkets and all petrol stations) that were recognized in the municipalities and districts targeted. In all cases, the coordinator (a research assistant) was present outside the location when the test purchase was made.
Selection of test purchasers
Two boys and two girls were used as test purchasers in each region in all the test purchase series, i.e. a total of 36 test purchasers participated. The participants in 1996, before the minimum-age law was introduced, were <18 years but their apparent ages were not estimated. The adolescents who participated in 1999 and 2002, i.e. after the introduction of the minimum-age law, were in fact 18 years of age or just over, since the use of minors as test purchasers is not in accordance with the Swedish law. The majority of the participants looked <18 years. The method used to examine the test purchaser's age was developed by Rehnman [13] in a study intended to determine the extent to which adolescents are asked to present identification when they order strong beer at restaurants.
The 1996 test purchasers were recruited with help from non-governmental organizations. In 1999 and 2002, they were recruited through advertising at upper-secondary schools in each region. Those who were interested in participating in 1999 and 2002 were invited to meet a reference group (in each region) consisting of a police officer, a door security guard, an upper-secondary schoolteacher, a school nurse and a cashier. At the meeting, the adolescents gave their names. After that, the members of the reference group estimated the age of each boy and girl to the nearest 6 months. There were no discernible occupational patterns in to how the reference group members estimated the adolescents' ages. The reference group's estimations were written down and of the two girls and the two boys who appeared, the youngest were selected for each of the three regions (see Sundh et al. [14] for a more detailed account). The actual ages and the average estimated ages of the test purchasers in all three series are presented in Table I.
|
The 18 participating girls attempted 1140 purchases. The actual age of the girls who participated in 1996 varied between 15 and 17 years. The majority of participating females in 1999 and 2002 was 18 years of age but with a younger appearance. The 18 boy participants attempted 1110 purchases. The actual age of the boys who participated in 1996 varied from 14 to 17 years. The actual ages of the male participants in 1999 and 2002 were between 18 and 20 years but the majority has a younger estimated age. Note that some of the test purchasers in 1999 had somewhat higher estimated ages than those who participated in 2002.
Data collection—test purchases
The first series of test purchases was conducted in October–November 1996. The second series was in March 1999 and the third in November–December 2002. Within each municipality and district targeted, the coordinator organized the test purchases. The participating adolescents selected the tobacco brands.
In each attempted purchase, the test purchaser entered the shop alone, since the presence of friends could conceivably affect the outcome. The participants were provided with money in advance to purchase tobacco. They were not permitted to carry any ID card but, if asked were to answer that it was not available. During the visit, the adolescents had to observe a number of important factors, such as the type of shop, whether they were asked for identification, whether there were any signs present informing customers of the minimum-age law. After each completed test purchase, the adolescents filled in a survey sheet, and all the tobacco purchased were handed over to the co-ordinator. The test purchases were made in the same way in all the three series in all the three regions, and the co-ordinator was the same person on all occasions.
Before the test purchases were conducted, the adolescents participating within each region were trained in how they should handle situations and questions that could conceivably arise during the purchases. The sales outlets were not informed of the test purchases either beforehand or afterwards.
Analysis
The present analyses are confined to the attempts to purchase cigarettes. Thus, of 2250 test purchases, 383 attempts to buy Swedish snus were excluded, leaving 1867 test purchases for analysis.
Contingency table analysis was used in Tables II and III. In Table II, the analyses are based on data from 1996 as well as data from 1999 and 2002. The results, which are presented by region, include all the test purchases, distributed according to whether there was a voluntarily imposed age limit. The z test for differences in proportions was used to test differences across years. In Table III, the analyses are based on data from 1999 and 2002 and concern the relationship between ID control and the outcome from the test purchases.
|
|
In addition, multiple logistic regression analysis (presented in Table IV) based on data from 1999 and 2002 was carried out. Data from 1996 were not used in this analysis because the age of the adolescents was not estimated this year. The purpose of the analysis was to estimate the effects of each of nine independent variables (for example, year of study, type of shop and region for purchase attempts) on the dependent variable, i.e. whether or not the test purchaser was able to buy cigarettes. ID control was not included in the logistic regression analysis, since it was considered a mediating factor. In anticipating the presentation of the results from the logistic regression analysis, it is reasonable to hypothesize that ID control and outcome from test purchases at the face-to-face situation with the cashier are measuring about the same thing.
|
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Table II shows the results of the test purchases distributed by year.
It became significantly more difficult over time for the participating adolescents to purchase tobacco. In 1996, 93% of the attempts made in shops that had not voluntarily imposed an age limit and 84% of the attempts made in shops with an age limit resulted in a purchase. The corresponding proportion in 1999 was 80% and in 2002, 63%. In all the regions, it was more difficult for test purchasers to buy cigarettes in 2002 than in earlier years. Comparison of the results of test purchases in 1999 and 2002 in the three regions also reveals significant differences.
In parallel with the increasing proportion of rejected test purchases over time, the proportion of ID controls also increased. In 1996, the adolescents were asked for ID card in 16% of the test purchases that were made in shops with a voluntary age limit. Corresponding figures for 1999 and 2002 were 24 and 45%, respectively.
Table III shows the proportion of test purchasers who were able to buy cigarettes, distributed according to whether ID card was requested or not.
Table III shows a very strong relationship between request for ID card and the outcome of the test purchases. When the test purchasers were asked for ID card, almost all test purchases failed which is statistically reflected by extremely high odds ratios (ORs) (1999:600; 2002:500).
Table IV shows the results of the multiple logistic regression analysis.
The results presented in Table IV show that the odds of being able to purchase cigarettes in 2002 were roughly half as high as the corresponding odds in 1999, controlling for the remaining eight factors in the model. Test purchasers with an estimated age of 18 years or over (=median value) had significantly higher odds of being able to buy cigarettes than the odds for test purchasers with median estimated ages <18 years. The odds of test purchasers being able to buy cigarettes were significantly lower where signs providing information about the minimum-age law were present than where there were no signs. The odds of the test purchasers being able to buy cigarettes in Västernorrland and Malmö were significantly lower than the corresponding odds for Värmland.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
The results from three series of attempted purchases presented in this paper concern the effect of the minimum-age law on the ease with which adolescents can buy cigarettes. In 1996, 93% of all the attempted purchases resulted in the participating adolescents being able to buy cigarettes. The corresponding figures for 1999 and 2002 were 80 and 62%, respectively. Thus, the results indicate that the introduction of the minimum-age law has affected the opportunities for adolescents to purchase cigarettes, even though compliance with the law must still be improved. A major difference between 1999 and 2002 is the frequency of age controls. In 2002, considerably more test purchasers than in 1999 were asked for identification, which resulted a number of attempted purchases failing.
The reduced availability of cigarettes to adolescents via personal purchases in shops reported in this study accords with statistics presented in a nationwide study [15]. In parallel with these decreasing figures on availability, the study also shows that cigarette smoking in Sweden has declined among adolescents. The proportion of students in Year nine (adolescents 15–16 years of age) who smoke daily or almost daily has decreased since 1996, i.e. the year before the minimum-age law was introduced. In 1996, the proportion among ninth year students who smoke daily or almost daily was 12% (boys) and 17% (girls). The corresponding figures for 2003 were 7% (boys) and 13% (girls). A similar trend is shown in an analysis based on Swedish data [16] from the international WHO study of health behaviour among schoolchildren [17].
The present analyses of the effect of different factors and conditions on the outcomes of the attempted purchases show that request for identification is crucial for the ease with which adolescents buy tobacco. This conclusion is consistent with results reported from other studies [10, 18]. DiFranza et al. [19] showed that, in those test purchases that resulted in the adolescents being able to buy cigarettes, identification had been asked for in only 1.5% of the cases.
Under Swedish law, the cashier selling tobacco is obliged to ascertain that the would-be purchaser of tobacco is at least 18 years of age. If the age law is to work more effectively, those who sell tobacco must make sure through ID checks that adolescents with a somewhat older appearance are in fact 18 years or over. The state liquor shops in Sweden, where all those <25 years are required to show proof of age could function as a role model for age controls in connection with tobacco sales.
Moreover, information and guidelines for staff may be improved through closer collaboration between shopkeepers and local authorities. Also crucial are, of course, staff attitudes towards the minimum-age law.
The results also indicate regional differences with respect to how compliance with the minimum-age law is working. For instance, it was considerably easier for adolescents in Värmland to buy cigarettes than it was for adolescents in the two other regions. Reference data from structured telephone interviews with key individuals responsible for tobacco-related issues in each region show that the results of the purchase attempts accord at least in part with the level of initiative taken in implementing and checking compliance with the law. In Västernorrland, the region where it was hardest for adolescents to buy cigarettes, active measures were apparently implemented when the minimum-age law was introduced, including information and personal visits in an attempt to induce shopkeepers to comply. In Värmland, where it was easiest for adolescents to purchase cigarettes, the interviews indicated that little effort had been made in connection with age-law implementation. Given the results reported in prior studies, the strong link between the outcomes from the attempted purchases and implementation of the law does not come as a surprise. It does, however, indicate a great potential for local efforts to induce more shopkeepers to comply with the law. This view is also supported by US studies [2–3] reporting much higher compliance rates with minimum-age laws compared with those indicated by the Swedish data.
In addition to implementation of the minimum-age law, the overall strategy for tobacco prevention turns out to be very important. The need for broad approaches and combinations of approaches has been stressed by Staff et al. [6] and Levy et al. [20].
The combined effects of clean indoor air laws, mass media policies and minimum age for the purchase of tobacco may also change attitudes to smoking. This in turn may directly influence not only the opportunities for adolescents to obtain tobacco (commercially or otherwise) but also their smoking habits.
Finally, a few methodological remarks about the limitations of the study. Although different characteristics of test purchasers as well as the cashiers were taken into account in the multi-variate analysis, personal characteristics not controlled for may have affected the results. For example, effects on the outcomes due to the behaviour of the test purchasers cannot be ruled out. Although the smoking frequencies among adolescents in Sweden have decreased since the introduction of the minimum-age law, other initiatives and conditions may have contributed to the declining smoking among adolescents. In order to more closely elucidate possible explanations to the decreasing smoking rates, the impact of local and regional initiatives will be analysed in a separate paper.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
Our profound gratitude is extended to the National Institute of Public Health in Sweden which provided the financial support that made this study possible. We also wish to thank Professors Arja Rimpelä and Matti Rimpelä for their excellent cooperation.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionAcknowledgementsReferences |
|---|
1. Rimpelä M, Aaro LE, Rimpelä A. The effects of tobacco sales promotion on initiation of smoking—experiences from Finland and Norway. Scand J Soc Med 1993 49:5–23 (Suppl.).
2. Forster JL, Murray DM, Wolfson M, et al. The effects of community policies to reduce youth access to tobacco. Am J Public Health 1998 88:1193–8.
3. Cummings KM, Hyland A, Perla J, et al. Is the prevalence of youth smoking affected by efforts to increase retailer compliance with a minor's access law? Nicotine Tob Res 2003 5:465–71.[Medline]
4. Fichtenberg CM and Glantz SA. Youth access interventions do not affect youth smoking. Pediatrics 2002 109:1088–92.
5. Rigotti NA, DiFranza JR, Chang Y, et al. The effect of enforcing tobacco sales laws on adolescents' access to tobacco and smoking behavior. N Engl J Med 1997 15:1044–51.
6. Staff M, Bennet CM, Angel P. Is restricting tobacco sales the answer to adolescent smoking? Prev Med 2003 37:529–33.[CrossRef][Web of Science][Medline]
7. Levy DT and Friend KB. Strategies for reducing youth access to tobacco: a framework for understanding empirical findings on youth access policies. Drug Educ Prev Policy 2002 3:285–303.
8. Croghan E, Aveyard P, Griffi C, et al. The importance of social sources of cigarettes to school students. Tob Control 2003 12:67–73.
9. DiFranza JR and Coleman M. Sources of tobacco for youths in communities with strong enforcement of youth access laws. Tob Control 2001 10:323–8.
10. Sundh M and Hagquist C. The importance of a minimum age law for the possibility of purchase of tobacco by adolescents: a study based on Swedish experiences. Scand J Public Health 2004 32:68–74.
11. Koskinen M and Leppänen K. Lasten suojelu päihteilä tupakkatuotteiden ja alkoholijuominen myyminen alle 18-vuotiaille Jyväskylässä ja Jyväskylän maalaiskunnassa. Esitutkimus ostokoemenetelmän kehittämiseksi [Sale of tobacco and alcohol products to children in the region of Jyväskylä. An exploratory study to improve a method for test purchasing]. In Rimpelä A (Ed.). Jyveskylä: Jyväskylän lääninhallitus ja Stakes, sisäinen raportti [Internal Report] 1996.
12. DiFranza JR, Savageau JA, Bouchard J. Is the standard compliance check-protocol a valid measure of the accessibility of tobacco to underage smokers? Tob Control 2001 10:227–32.
13. Rehnman C. En stor stark tack! En studie av legitimationskontroll på restauranger i Stockholm under våren och hösten 1996 [A large strong beer please! A study of identification controls in restaurants in Stockholm during the spring and autumn of 1996]. Stockholm: Psykiatriska Beroendekliniken, St Göran, 1996.
14. Sundh M, Hagquist C, Rimpelä A, et al. Ungdomars möjlighet att köpa tobak. Resultat från provinköpsstudier genomförda före respektive efter införandet av 18-årsgräns för tobaksinköp. [The possibility of purchase of tobacco by adolescents. Results from purchase attempts before and after the introduction of the minimum age laws of 18 years]. Karlstad: Karlstad University StudiesVolume 200127.
15. Hvitfeldt T, Andersson B, Hibell B. Skolelevers drogvanor 2003 [Drug abuse among students 2003]. Stockholm: The Swedish Council for Information on Alcohol and Narcotics (CAN), Report No.77, 2004.
16. Danielsson M. Svenska skolbarns hälsovanor 2001 /02. Grundrapport 2003:50 [Health behaviour in school-aged children in Swedish 2001/02. A WHO collaborative study]. Stockholm: Swedish National Institute of Public Health 2003.
17. Currie C, Roberts C, Morgan A, et al. Young People's Health in Context. Health Behaviour in School-Aged Children (HBSC) Study: International Report from the 2001/2002 Survey. Copenhagen: World Health Organization Regional Office for Europe 2004.
18. Hyland A, Cummings KM, Seiwell M. The impact of untruthful age reporting during tobacco compliance checks. J Public Health Manage 2000 6:115–8.
19. DiFranza JR, Celebucki CC, Mowery PD. Measuring statewide merchant compliance with tobacco minimum age laws: the Massachusetts experience. Am J Public Health 2001 91:1124–5.[Abstract]
20. Levy DT, Chaloupka F, Gitchell J. The effects of tobacco control policies on smoking rates: a tobacco control scorecard. J Public Health Manage 2004 10:338–53.
Received on May 11, 2005; accepted on March 1, 2006
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||