This Web site is a component of the SAMHSA Health Information Network.

Synar Regulation Implementation: FY 97 State Compliance Section I: Program Summary February 1998

A Preliminary Analysis of Tobacco Sales to Minors Before and After the Synar Legislation

Throughout the past 25 years, tobacco control efforts have been conducted in numerous States and communities across the United States in order to reduce youth smoking. Although the Federal Government often provided funding and technical assistance through several programs, such as the National Cancer Institute's ASSIST (American Stop Smoking Intervention Study) and the Centers for Disease Control and Prevention's IMPACT (Initiatives to Mobilize for the Prevention and Control of Tobacco Use), tobacco control initiatives and their goals were primarily determined at the local level. Among the measures taken by many States and communities to keep young people from using tobacco was the enactment of legislation prohibiting its sale to minors. However, because enforcement of these laws was rarely undertaken, youth continued to be able to purchase tobacco products readily from both over-the-counter and vending machine outlets.

The minors' access issue changed substantially in January 1996 with the publication of the Synar Regulation. Starting in 1996 (and with the Federal Fiscal Year [FFY] 1997 SAPT Block Grant applications), States were required to enforce their youth access laws, conduct statewide surveys to determine retailer compliance, develop plans to decrease the illegal sales rate to 20 percent or less, and provide documentation of these efforts to SAMHSA. The regulation also established performance targets regarding noncompliance, with financial penalties that might be imposed on States for failing to meet them.

Although it is expected that the implementation of the Synar law will contribute to the reduction of youth smoking, the first measure of the regulation's effectiveness will be whether it has succeeded in its stated goal: to reduce the rate of illegal tobacco sales to minors. In order to gauge whether there has been progress toward meeting this goal in the short time since implementation, this section of the State Compliance Report examines and analyzes various assessments of noncompliance rates taken before and immediately after the Synar Regulation went into effect.

While this comparison is useful as a preliminary guide, the comparison of rates of tobacco sales to minors before and after the implementation of the Synar Legislation is not straightforward. Although about 40 studies conducted before its implementation may be found in the published literature, important differences exist between those studies of youth access to tobacco and the surveys mandated under the regulation. The caveats associated with both sets of studies are reviewed here, along with their general findings. Because of the widespread interest in appraising the immediate effect of the Synar Regulation on retailer compliance, a strong impetus exists for using the baseline rates as postintervention assessments for such an analysis. However, the original purpose of the baseline rates is to provide a standardized, and therefore reliable, initial measurement against which to measure future efforts to control illegal sales of tobacco to youth under 18 years of age.

Pre-Synar Regulation Studies of Sales to Minors

A number of important methodological issues distinguish studies conducted before the Synar Regulation from those conducted later. These differences in how research was conducted can introduce important biases that make comparisons and interpretation of findings across studies difficult. Methodological issues associated with the preregulation studies include:

• Differences in the levels of assessment of noncompliance,

• Variability in sampling procedures,

• Differences in inspection methodologies, and

• Multiple reports of noncompliance within studies.

This difference in the unit of analysis of noncompliance arises because research conducted before 1996 had a different purpose than that of the surveys conducted after the regulation. Compliance surveys conducted before the Synar Regulation were most often undertaken as part of local evaluations to determine whether community-based youth access interventions were effective in reducing sales to minors in those specific localities. In contrast, the purpose of the scientific surveys mandated by the regulation is to document the statewide rate of noncompliance through a random probability sample of outlets throughout the State.

In addition to local evaluations, a few studies of baseline rates of sales to minors were also conducted within a small number of States before the regulation took effect (Centers for Disease Control and Prevention [CDC], 1990, 1993, 1996; Fox, Gothard, & Remington, 1996; Gothard, Fox, Flores de Pierquet, Musial, & Yoast, 1996; Thomson & Toffler, 1990; Young, Walkington, & Ravesloot, 1990). However, with one exception, these baseline studies reported findings that were not based on statewide samples of retailers, but rather on data collection limited to distinct geographical regions of those States (e.g., a single city, such as Portland in the Oregon assessment, or a large city and its surrounding areas, such as the Austin metropolitan area for the report of sales in Texas or suburban Denver and outlying communities for the assessment in Colorado). Looking outside the published literature is unlikely to yield much data on statewide noncompliance rates because very few States (with the exceptions of Florida, Utah, and Vermont) were vigorously enforcing their youth access laws (HHS, 1992; Kirn, 1987), much less conducting statewide surveys prior to the Synar Regulation. Thus, while there may be few if any alternatives to using these substate assessments of noncompliance for determining sales rates prior to the Synar Regulation, their lack of comparability to the statewide sales rates is problematic in analyzing changes over time.

Variability in sampling procedures. Studies conducted prior to the Synar Regulation varied substantially from each other as well as from postregulation studies in terms of the sampling techniques used. These differences in sampling procedures pose threats to the reliability and validity of the results of preimplementation studies.

Sometimes studies conducted before 1996 relied on either convenience and/or very small samples of retailers (Cummings & Coogan, 1992; Jason, Ji, Anes, & Birkhead, 1991; Jason, Ji, Anes, & Xaverious, 1992). For example, to estimate the cigarette sales rate to minors in the Chicago area, Jason et al. (1992) used DePaul University students to recruit minors and observe purchase attempts. Stores were selected from a convenience sample of stores that were located near the undergraduate students' area of residence. Nonrandom and small samples are less likely to be representative of the population to which one wishes to generalize research findings.

Differences in inspection methodologies. Studies conducted before the Synar Regulation used inspection protocols that varied in a number of important ways, further reducing the comparability of these studies to each other as well as to those conducted after the regulation. It is widely recognized that methodological differences, such as the age and gender of the minor making the purchase, influence tobacco sales rates.

Differences among inspection protocols include the age of youth inspectors, whether youth worked in pairs or alone, whether an adult escort was in the store during the purchase attempt, whether youth carried proof-of-age identification into the store, and whether youth were instructed to show identification or state their true age, if asked by a salesclerk. Many studies conducted before the Synar Regulation used a wider age range of youth inspectors, often involving adolescents 13 years old and younger to attempt tobacco purchases (CDC, 1990; DiFranza & Brown, 1992; DiFranza, Carlson, & Caisse, 1992; DiFranza, Norwood, Garner, & Tye, 1987; Forster, Hourigan, & McGovern, 1992; Hoppock & Houston, 1990; Jason et al., 1992; Thomson & Toffler, 1990), compared to the best practice guidance given to the States by SAMHSA to use teenage inspectors 15 to 16 years of age in the surveys required for compliance with the Synar Regulation. With no standardized procedures for inspections, the ability to compare and generalize findings from preregulation studies is more limited than with the postregulation assessments of tobacco sales rates using more consistent protocols.

Multiple reports of noncompliance within studies. A final issue related to studies conducted before the Synar Regulation surrounds the fact that these studies typically contain multiple assessments of sales to minors within a particular study -- preintervention, postintervention, and occasionally followup measures. The question is whether all sets of assessments or the preintervention rates only should be considered to reflect the status quo prior to the regulation. If the evaluations prior to 1996 are seen as a series of "local Synar Regulation pilot tests" because they embody the spirit of the regulation, then for the sake of conceptual consistency, it is necessary to limit our analysis to the pretests exclusively in examining preregulation sales rates, which is the strategy adopted here.*

Findings

With these methodological issues in mind, what does the research indicate about rates of tobacco sales to minors before the Synar Regulation? The majority of studies conducted before 1996 indicate that:

• Noncompliance was typically in the 60- to 90-percent range,

• There is wide variability in noncompliance rates, and

• The retail source of tobacco influences both how high and how variable the rates of sales to minors are.

Variability in noncompliance rates. As is true now, there is wide variability in documented pre-regulation sales rates to minors, which ranged in one study from 0 to 100 percent across nearly 90 communities in the United States (Radecki & Zdunich, 1993). The 69 independent sales rates (data for 22 towns and cities in Illinois were combined) were distributed as shown in figure 1. The bar farthest to the right includes 13 communities that reported 100 percent successful purchase rates by minors. The median was a noncompliance rate of 83 percent (with 34 communities below and above that sales rate), reflecting the highly skewed nature of the distribution with most cities clustered at high noncompliance rates. It is important to note that the noncompliance rates reported by Radecki and Zdunich may be somewhat elevated compared to other studies conducted before the regulation because youth inspectors were told to lie about their age. The wide disparity in sales rates is also evident in a study of more than 40 compliance operations conducted throughout Wisconsin between 1992 and 1995 (Gothard et al., 1996) in which noncompliance across communities ranged from 7 to 76 percent.

Source of tobacco. The retail source of tobacco influences both how high and how variable the rates of sales to minors are. When the source for purchase attempts is vending machines, there is less variability in minors' success rates (82 to 100 percent for vending machines across 9 studies versus 32 to 87 percent for over-the-counter sales across 13 studies), and noncompliance rates are substantially higher (approximate weighted-average sales rate of 88 percent for vending machine purchases versus 67 percent for over-the-counter sales) (HHS, 1994).

Post-Synar Regulation Studies of Sales to Minors

Methodological Issues

Problems with drawing conclusions about changes in sales to minors before and after implementation of the Synar law are also due in part to a number of issues related to the first set of noncompliance rates officially reported by the States in the FFY 1997 block grant applications. Methodological issues associated with these post-Synar Regulation assessments include:

• Delayed applicability,

• Differences in sampling methods,

• Variability in inspection protocols, and

• Co-occurrence of other strong tobacco control initiatives.

Delayed applicability. Delayed applicability. This analysis of baseline rates is limited to a subsample of 43 States and the District of Columbia. It does not include all States and applicable jurisdictions because of the 1-year delayed applicability schedule for seven States and because of the unique circumstances presented by the jurisdictions covered under the Synar Regulation. (See table A2 in the appendix on the requirements for timing of inspections.)

Seven States were unable to pass the required tobacco access laws because their legislatures were not scheduled to meet within the timeframe covered by the Synar Legislation. The law provided for this possible delay, and as a result, Congress delayed the applicability of the baseline sales rate requirement for these States for 1 year. Known as "delayed applicability" States, they are Arkansas, Kentucky, Montana, Nevada, North Dakota, Oregon, and Texas. Their baselines will be based on their FFY 1997 tobacco sales data to be reported in their FFY 1998 block grant applications.

In addition to these States, the jurisdictions face a unique set of geographic, cultural, resource-related, and logistic issues in complying with the Synar Regulation. As a result of these special challenges, their noncompliance rates are outliers and thus not representative of the distribution of sales rates nationwide.

Differences in sampling methods. States used a variety of designs for drawing their samples of tobacco outlets, which resulted in differing levels of precision in their estimated sales rates. A number of challenges faced by States account for their developing different sampling designs. Some States had no list of tobacco vendors, and others had incomplete or out-of-date lists, which made identification of the sample universe of tobacco retailers difficult. In addition, some States were able to avail themselves of consultants who helped develop the sampling plans, while others did not have expert resources on which to draw for help.

Variability in inspection protocols. Differences in inspection protocols used by States also reduce the comparability of their survey results. In some cases, States' legislation mandated different procedures from the guidance in the SAMHSA-promulgated tobacco outlet inspections protocol, such as the prohibition against using minors in compliance surveys. Aside from legal barriers, some States also faced practical limitations on their ability to adhere to the SAMHSA protocol.

Co-occurrence of other strong antitobacco initiatives. Because other strong antitobacco measures occurred simultaneously with the implementation of the Synar Legislation, any reductions in sales to minors may reflect their combined effects and would overstate the impact of the Synar law alone.

The negotiations between State attorneys general and the tobacco companies, as well as the $368.5 billion "tobacco settlement," were widely covered by the media. Class action suits filed by nonsmoking citizens, such as flight attendants who were exposed to environmental tobacco smoke on airplanes, also received significant attention. Additionally, the final rule giving the FDA the authority to regulate tobacco products, which was issued in August 1996 and went into effect in February 1997, established a number of provisions complementary to the Synar Legislation for restricting youth access to tobacco.* The FDA rule also includes provisions to reduce the appeal of advertising to young people and to educate them about the health risks associated with tobacco use in addition to its measures designed to restrict youth access. Because of these other antitobacco activities, it is difficult to know how much of the change in retailers' sales practices may be attributed to the Synar Regulation exclusively.

Findings

What do the baseline noncompliance rates from the FFY 1997 block grant applications indicate about sales to minors after the implementation of the Synar Regulation?

The 43 States and District of Columbia are distributed as follows across baseline noncompliance rates (see figure 2 above and map 1 in the appendix): 4 with rates less than 20 percent, 7 in the 21- to 30-percent range, 13 in the 31- to 40-percent range, 12 in the 41- to 50-percent range, 4 in the 51- to 60-percent range, 3 in the 61- to 70-percent range, and 1 over 70 percent. The median is at 40 percent noncompliance. (The median represents the point at which half of the reported baseline rates fall above and half fall below.)

Conclusion

While it would be tempting to compare the distributions of sales rates and medians from the postregulation baseline surveys with rates found in studies conducted before the regulation took effect, there are too many limitations with both sets of assessments to permit such an analysis. Sales-to-minors rates appear to be moving in the right direction -- from prelegislation studies showing most rates falling in the 60- to 90-percent range to the baseline surveys in which the majority of States are in the 31- to 50-percent range. While sales of tobacco products to youth appear to be diminishing, at this time it is too early to tell how much of this decline may be attributed to implementation of the Synar Regulation alone and to know how stable and enduring these changes will be.

Previous Section || Table of Contents || Next Section

---

Home Page