Are the Federal and State Governments Complying With the Synar Amendment? FREE

IN 1987, ATTENTION focused on the sale of tobacco to children when it was demonstrated that an illegal sale was made to an 11-year-old girl in 75 of 100 attempts to purchase tobacco.¹ At that time, 12 states had no laws concerning the sale of tobacco to children, and in those states that did have laws, enforcement was almost nonexistent.^{1 - 3} Attempts to improve merchant compliance through educational programs have at times produced a transient improvement in merchant behavior, but the modest effects of even the best programs have not reduced the availability of tobacco to minors.^{4 - 7} Success in reducing the availability of tobacco to minors has been achieved only through the frequent inspection of all retail outlets through the use of underage decoys, followed by penalties and reinspection.^{8 - 10} However, since youths know which stores in their communities will sell them tobacco, it has been shown repeatedly that the availability of tobacco to minors and underage tobacco use are reduced only when the proportion of noncompliant merchants falls below 10%.^{9 - 11} In 1 study,¹² reducing merchant noncompliance from 65% to 18% had little impact on the availability of tobacco, and no effect at all on tobacco use. Enforcement is an effective public health intervention only when it is conducted with sufficient vigor to produce high levels of merchant compliance.^{9 - 12}

After 3 studies^{9 - 10 ,13} reported reductions in the prevalence of smoking among adolescents in communities in which a prohibition on the sale of tobacco to minors was enforced, Congress enacted Public Health Service Act 398 in July 1992.¹⁴ This law has become known as the Synar Amendment after its sponsor, the late US Rep Mike Synar (Oklahoma). It makes block grants to states from the Substance Abuse and Mental Health Services Administration (SAMHSA) under the Department of Health and Human Services (DHHS) contingent on states enacting and enforcing a prohibition on the sale of tobacco to minors.¹⁴

Congress allowed states until 1994 to enact a law and enforce it "in a manner that can reasonably be expected to reduce the extent to which tobacco products are available to individuals under the age of 18."¹⁴ To this end, states are required to conduct "random, unannounced inspections to ensure compliance with the law."¹⁴ Congress required DHHS to reduce block grant funding to noncompliant states by 10% for 1994, 20% for 1995, 30% for 1996, and 40% for 1997 and all subsequent years that the state remains noncompliant.¹⁴ States whose legislatures did not convene during either 1993 or 1994, the "delay states" (Arkansas, Kentucky, Montana, Nevada, North Dakota, Oregon, Texas, and Wyoming), were allowed until 1995 to enact and enforce a state law.¹⁴

When the Synar Amendment was adopted, state enforcement of tobacco sales laws was nonexistent, and illegal sales rates averaged 76% across the United States with little variation between states.¹⁵ Congress left it to DHHS to define the criteria to be used to determine if a state was enforcing its law "in a manner that can reasonably be expected to reduce the extent to which tobacco products are available to individuals under the age of 18."¹⁴ Tobacco control workers advocated a uniform national standard that would require all states to reduce illegal sales to 50% of attempted purchases by minors within the first year (1994 or 1995), and further reduce the illegal sales rate by 10% each succeeding year until merchant noncompliance fell below 10%.¹⁶

On August 26, 1993, DHHS issued draft regulations incorporating this model with the exception that the ultimate noncompliance goal was weakened from 10% to 20%.¹⁷ The draft regulations were strongly opposed by the tobacco industry, by US Rep Thomas Bliley (Virginia) (Thomas Bliley, written communication, October 25, 1996), and by some state officials.^{18 - 19} Final regulations were promised by "early 1994" but were not released until January 19, 1996, 3½ years after the Synar Amendment was enacted (Gale A. Held, written communication, August 1993).¹⁹ This delay postponed the implementation of the law by 2 years.

Although states are required to conduct an annual statewide random survey to measure merchant compliance, the final regulations do not define effective enforcement.¹⁹ The uniform goals and timelines in the draft regulations were eliminated, leaving it to the discretion of DHHS to determine on an annual and individual basis if a state met undefined criteria. Individual compliance goals and timelines are to be negotiated with each state.¹⁹ Although Congress mandated that DHHS "shall" withhold funding from noncompliant states, the regulations provide that DHHS can refrain from applying sanctions if it determines that undefined "extraordinary circumstances" exist.^{14 ,19}

When DHHS officials announced that they were "confident that no states will lose financing," it raised concerns that DHHS did not plan to withhold funding from noncompliant states.²⁰ To determine if the states and DHHS were in compliance with the statutory requirements of the Synar Amendment and the implementing regulations, an independent audit of each state's fiscal year 1997 application was performed. The DHHS regulations and actions were examined to determine if these were in compliance with the statutory requirements of the Synar Amendment.

Throughout this article, the term "state" shall include the District of Columbia and 8 US territories (American Samoa, Micronesia, Guam, Marshall Islands, Northern Mariana Islands, Palau, Puerto Rico, and the Virgin Islands of the United States). All states file an annual block grant application with SAMHSA.²¹ States must provide the methods and results of their statewide survey, information concerning how the state law was enforced during the preceding fiscal year, plans for enforcement in the upcoming year, and the timeline the state has negotiated with DHHS to reduce merchant noncompliance.²¹ States must also certify each year that they have in effect a law prohibiting the sale of tobacco to minors younger than 18 years and that the state is enforcing the law effectively.²¹

In spring 1997, the author informed the secretary of the DHHS of the purpose of this study and filed a Freedom of Information Act request for each state's fiscal year 1997 application and for correspondence between SAMHSA and the states. These applications cover state activities during 1995 and 1996. States are asked to describe their activities during the previous year, but since some states file their applications earlier than others, there are no uniform cutoff dates for the activities covered by these reports. Additional information was obtained from SAMHSA's "Report to Congress," which summarizes and comments on each state report and provides the performance guidelines negotiated between DHHS and the state.²² Telephone calls and letters to voluntary agencies and state, county, and local officials were made when there was doubt about whether any law enforcement had been conducted.

The following 4 criteria were used to judge state compliance. The first 3 are requirements of the Synar Amendment.¹⁴ The last is required by the implementing regulations.¹⁹

The DHHS actions were audited to determine compliance with 3 criteria:

All materials were audited by 1 researcher (J.R.D.). Before publication, this manuscript was shared with DHHS to allow them the opportunity to identify any factual inaccuracies or to supply additional information (J.R.D., written communication, September 3, 1998).

Applications containing 1400 pages from 59 states were obtained and reviewed.

State reports ranged in length from a few pages (Marshall Islands) to a few hundred pages (South Carolina). Data quality was generally poor. Data presented in tables and text were frequently inconsistent without explanation (eg, North Dakota, Ohio, Oklahoma), and descriptions of the survey methods and results were often scanty, contradictory, or absent (eg, Arkansas, Delaware, Missouri, Oklahoma). Descriptions of enforcement efforts were typically quite vague, and often concealed the fact that there was no enforcement to describe.

The state reports were audited for compliance with the following 4 criteria.

1. A law prohibiting the distribution of tobacco products to minors. All 59 states had enacted laws prohibiting the sale of tobacco to minors by 1996 (Table 1). However, 2 states provided exceptions that undermined the intent of the law (Maryland and Northern Mariana Islands). The Northern Mariana Islands law "allows a vendor to sell tobacco to a minor so long as the tobacco is not ‘for the minor's own use.'" Reflecting the difficulty of enforcing or complying with this law, sales to minors were made by 95% of the islands' merchants. Maryland's law prohibits vending machine operators from being penalized for illegal sales provided the machines display a warning sign. In a letter to the governor dated February 9, 1995, SAMHSA determined that "Maryland's law is not in compliance with the requirements of Section 1926 of the PHSA [Public Health Service Act] and this could result in a loss of SAPT [Substance Abuse Prevention and Treatment] Block Grant funding" (Elaine M. Johnson, PhD). However, in a letter dated May 18, 1995, SAMHSA reversed its decision in response to a technical argument that the Synar Amendment requires states to prohibit the sale of tobacco to minors but does not specify that there must be penalties for breaking the law (Nelba Chavez, PhD). The effect of this loophole in the Maryland law is seen in the illegal sales rates reported for vending machines of 95.5% for 1996 and 87.1% for 1997.

TWENTY-SEVEN STATES have in effect provisions that could hinder the effective enforcement of their laws (Table 1). A common tobacco industry strategy to hinder enforcement is to promote state legislation that preempts local governments from enacting or enforcing local restrictions.¹⁸ Many states enacted preemptive laws while DHHS was considering the Synar regulations. However, in a few states, preemption was written into the state constitution and was not targeted at tobacco sales laws. Several states have made the successful prosecution of violators more difficult, if not impossible, by requiring prosecutors to prove that a merchant "knowingly" sold tobacco to a minor.¹⁸ Some states have hampered enforcement by limiting enforcement authority to selected state officials, or by proscribing the use of underage purchase attempts.^{18 ,22 ,26}

2. Inspections conducted to enforce the law. The block grant application requires states to "describe in detail the State's activities used to enforce the law(s) in FFY [Federal fiscal year] 1996."²¹ Despite this requirement, it was frequently impossible to determine from the states' responses if any inspections had been conducted that resulted in merchants being penalized. Fifteen states failed to provide SAMHSA with evidence that any inspections had been coupled with prosecution (Table 1). Florida conducted the most underage buyer inspections (n=9819), followed by Massachusetts (n=8000). The study protocol included a plan to correlate the number of enforcement inspections (test purchases) per vendor with the state's compliance rate. Unfortunately, this could not be done because less than one third of states reported the number of enforcement inspections, less than half provided reliable estimates of the number of vendors in the state, and variations in the methods used to conduct the compliance surveys rendered comparisons invalid.

3. Penalties as evidence of enforcement. Most state reports were seriously lacking in detail concerning the citation and prosecution of violators of the law. Eighteen states failed to provide evidence that any merchant had been prosecuted successfully for breaking the law during the preceding fiscal year (Table 1). The Marshall Islands reported no inspections. In Puerto Rico, 12 previous attempts to prosecute merchants were thrown out by the court system,

due to the unwillingness of the judges to place importance on violations for tobacco sales to minors as compared to the violations of drug and alcohol sales and the limitations of the law in terms of providing options for the sanctioning of the violators such as fines or repeal of licenses for the sale of tobacco products.

With the exceptions of the Marshall Islands and Puerto Rico, all of the states that failed to prosecute merchants had observed merchants violating the law but did not prosecute. Compliance surveys in which merchants were provided information regarding their performance or given an informal warning were frequently offered as evidence of effective enforcement (Kentucky, Mississippi, Missouri, Montana, and Virginia).

Tennessee and Puerto Rico attempted to enforce the law through "passive inspections" that entailed officials checking signs and observing for illegal sales while conducting other inspections. During fiscal year 1995, Tennessee Department of Agriculture officials inspected 33,449 establishments. Signs regarding the law were observed in 87% of establishments, suggesting a very high rate of merchant awareness. Despite this, minors were able to purchase at a rate of 628 per 1000 merchants in a January 1997 state survey. Illegal sales were observed in only 1 per 1000 passive inspections. Passive inspections failed to detect 99.8% of law breakers (627/628). The state policy is to give a written warning for a first offense. The chance of a violator getting caught in 2 consecutive inspections can be estimated at 1 in 400,000 (1/628×1/628).

In Puerto Rico, while 90.7% of merchants made illegal sales during test purchases, 6000 passive inspections by police and treasury agents failed to detect a single violation.

Most of the 41 states that had prosecuted lawbreakers were not enforcing their laws in a manner that could even remotely be expected to reduce the availability of tobacco to minors. While the Synar Amendment does not require the state law to be enforced by state officials, often states claimed credit for local enforcement efforts that were conducted without state funding and often in spite of state laws that hindered local efforts (eg, Indiana, Iowa, Maryland, Michigan, Minnesota, North Carolina, North Dakota, New Mexico, South Carolina). In many cases, local officials enforced only local laws, because enforcing the state law was cumbersome (eg, Illinois, Massachusetts, Michigan, Minnesota, New Jersey). Only a handful of states reported budgeting funds to enforce the state law (eg, Alabama, California, Florida, Massachusetts, New Jersey, South Dakota, Utah, Washington). Even fewer had implemented enforcement programs that could reasonably be expected to affect the availability of tobacco to minors (eg, Alabama, Florida, Massachusetts, South Dakota, Nevada). Perhaps three quarters of all states had not implemented effective enforcement strategies and did not describe plans to do so. The components of an effective enforcement strategy have been identified, and include an enforceable law with penalties sufficiently severe to deter potential violators, a practical mechanism for prosecuting violators, and a designated enforcement agency with a mandate and resources adequate to allow it to inspect all merchants through the use of underage decoys with sufficient frequency to reduce the violation rate below 10%.⁸

4. Compliance survey. Compliance surveys were conducted by every state except the Marshall Islands. In several cases, the surveys were not conducted until the deadline for applications had passed (eg, Louisiana, Missouri, Tennessee, Wisconsin). Missouri did not provide a description of its survey that would allow an evaluation of its quality. Eighteen states conducted surveys that may not have accurately measured statewide compliance rates because of a failure to use a process of randomization, an inadequate sample size, or both. One survey included only 10 merchants (Palau).

Bias in the direction of better compliance rates may have occurred by using young buyers, conducting surveys immediately after an intervention, or limiting data collection to areas where enforcement was active (Vermont, Virginia, and Guam). Since this survey was to be used as a baseline against which future progress would be measured, several states argued that their violation rates were actually worse than their survey would indicate (Colorado, Vermont, Virginia, Wyoming, and Guam). Several states were allowed by SAMHSA to adjust their baseline violation rates upward (Colorado, Vermont, Wyoming, and Guam).

The proportion of merchants who violated the law during the state survey varied from 7.2% (Florida) to 95.1% (Northern Mariana Islands), with only Florida reporting a violation rate below 10%. States are not required by SAMHSA to follow a standard survey protocol. As a result, state-to-state comparisons of violation rates are invalid. A listing of state violation rates is omitted herein for that reason. States used youths from the age of 10 years (Guam) to the age of 19 years (District of Columbia and Wyoming) to purchase tobacco. In 15 states, it was not possible to determine the exact age range of the youths used. The strong confounding impact of age is well established,⁵ and is seen in an analysis of the Massachusetts data performed by the author, where a difference of only 1 year in the mean age of the youths produced an absolute difference in violation rates of 8.6 percentage points (25.2% vs 16.6%) greater than the confidence interval of plus or minus 5 percentage points requested by SAMHSA.^{23 - 25}

Other important variations in methods include the exclusion (Alabama, Kansas, Minnesota, and Oregon) or limitation (Virginia and West Virginia) of the number of vending machines, the exclusion of liquor establishments (North Dakota), the presence of adults in the store (Florida, Illinois, Texas, and Virginia), and the requirement that youths appear obviously underage or not appear older than average (Georgia, Kansas, Kentucky, Oklahoma, Oregon, Tennessee, and Micronesia). No state allowed youths to lie about their age. New York did not allow youths to present proof of age "to avoid confusing a clerk who may not know the birth date to purchase." This may have lowered New York's violation rate since South Dakota reported that 9% of the time, minors were sold tobacco after showing proof of their true age.

For this audit, any state that submitted a survey, no matter how small, flawed, biased, or tardy, was credited with fulfilling this obligation.

Nineteen states failed 1 or more of the 4 state criteria (Table 1). Although not statistically significant (P=.11), compliance with the Synar Amendment was more common among the 17 states receiving funding from the federal American Stop Smoking Intervention Study for cancer prevention (88%) than among the 34 nonparticipating states (68%). Territories were more likely to fail than states, but no other regional patterns in compliance were evident.

The actions of the DHHS were audited for compliance with the following 3 criteria.

1. Are DHHS regulations, as implemented, requiring states to adopt and enforce laws in a manner that can reasonably be expected to reduce the extent to which tobacco products are available to individuals younger than 18 years? In the case of Maryland, DHHS has interpreted the Synar Amendment to require states to adopt laws but not to establish any punishment for breaking the law. This interpretation of the Synar Amendment is in conflict with Congress' mandate that the law be enforced, for how can a law be enforced if there is no penalty? The Northern Mariana Islands' law permits the sale of tobacco to minors provided they do not subsequently use it. In this case, DHHS has approved of a law that does not prohibit the sale of tobacco to minors. Many other states have adopted laws that hamper enforcement or render the prosecution of violators impractical if not impossible (Puerto Rico).¹⁸ The DHHS is not requiring states to adopt enforceable laws prohibiting the sale of tobacco to minors.

The DHHS negotiated timelines with each state for a plan to reduce violation rates to 20% of attempted purchases. Experimental data indicate that minors encounter little difficulty in purchasing tobacco in their own communities when the violation rate is 18%.¹² All of the studies^{9 - 11} that have demonstrated a decreased availability of tobacco to minors, as evidenced by a reduced prevalence of tobacco use, have achieved violation rates below 10%. Thus, based on the scientific evidence, the 20% goals negotiated by DHHS cannot be reasonably expected to reduce the availability to minors of tobacco from commercial sources.

Congress requires states to conduct "unannounced inspections" to "ensure compliance" with the law.¹⁴ Eighteen states did not report any instance in which a merchant was penalized for violating the law, and 15 of these states did not report conducting a single enforcement inspection involving an underage buyer. The remaining 41 states are enforcing the law, but most are not doing so in a manner that could be reasonably expected to reduce the availability of tobacco to minors. None of the states that failed to enforce their laws were sanctioned.

2. Has DHHS implemented the Synar Amendment according to the time frame mandated by Congress? Congress required states to implement effective enforcement by 1994 or 1995. It can be reasonably argued that were a state to institute an effective enforcement program by 1994 or 1995, it could still take time for violation rates to fall. According to SAMHSA, communities can "reduce their violation rates approximately 15-20 percentage points per year, until they reach 40 percent" and then by 10% each year.²² This drawn out timeline is inconsistent with the published data^{9 - 10} from communities where effective enforcement programs produced violation rates under 10% within 1 to 2 years. For example, on St Thomas, a Virgin Island of the United States, violation rates fell from 90% in 1994 to 6% in 1996.

Violation rates in all states were comparable when the Synar Amendment was adopted in 1992.¹⁵ Some states implemented enforcement plans promptly (Florida), while others did nothing before (or after) (Wyoming) the final regulations were issued in 1996. Had states implemented effective enforcement programs when the Synar Amendment was enacted, most could have achieved effective violation rates by 1994 or 1995. The DHHS did not require states to make any progress in reducing violation rates between 1992 and 1996. The 1996 violation rates serve only as a baseline for future comparisons. The delay states were given until 1997 to perform a baseline survey and begin enforcement, even though Congress mandated that these states achieve enforcement by 1995. Since the timelines negotiated by DHHS provide states with more time in proportion to their baseline 1996 violation rates, the states that ignored the congressional mandate to enforce the law from 1992 to 1996 are rewarded with additional time to achieve reductions in their violation rates. Eighteen states were allowed 4 or 5 years beyond the congressional 1994 deadline, 26 were allowed 6 or 7 additional years, and 3 were allowed 8 or 9 additional years to reach the 20% goal. Only 4 states had already achieved the 20% goal (Florida, Maine, New Hampshire, and Washington). The DHHS has not implemented the Synar Amendment according to the time frame mandated by Congress.

3. Has DHHS applied the statutorily mandated funding sanctions to states that failed to meet the statutory requirements of the Synar Amendment or the implementing regulations previously described?

Nineteen states failed to meet 1 or more statutory requirements. All received full block grant funding, although a few experienced some delay until their application was completed by providing survey results (eg, Missouri). The DHHS has not sanctioned any state for failure to comply with the Synar Amendment.

The perceived availability of tobacco has been monitored in a national survey.²⁷ When the Synar Amendment was enacted in 1992, 89.1% of 10th graders believed that tobacco would be "fairly easy" or "very easy" to obtain. In 1997, 89.6% believed this.²⁷ Clearly, Congress' goal of reducing the availability of tobacco to minors in all states by 1995 has not been accomplished. The DHHS regulations, as written and implemented, are in conflict with the letter and spirit of the Synar Amendment in that they do not require states to enforce their tobacco access laws or to achieve violation rates low enough to reduce the availability of tobacco to minors. The DHHS has postponed the implementation of the Synar Amendment; with the DHHS policy, it is doubtful that, through this legislation, the availability of tobacco to youth will be reduced in most states. With each year of DHHS' failure to implement the Synar Amendment, adolescent tobacco use rates have risen progressively.²⁷

As documented by this audit, state violations of the Synar Amendment are widespread, but DHHS has not applied sanctions as required by law. Eighteen states failed to provide a single example of a violator being penalized but provided the federal government with legal certification that the law is enforced. Others certified that the state law was enforced when, in fact, only local ordinances had been enforced (eg, Indiana, North Dakota). While local ordinances were almost uniformly superior to the state laws, the state bears the responsibility to enforce a state law.

The prediction by DHHS officials that no state would lose financing proved accurate.²⁰ Although well-intentioned, the Synar Amendment was clumsily drafted in that it placed the threat of reduced funding on mental health agencies that had no authority or funding to enforce tobacco access laws. The DHHS compounded the problem by prohibiting those programs from using block grant funds for enforcement.¹⁹ The DHHS was given the conflicting duties of promoting and sanctioning these programs, and apparently decided, before the fact, that it would not sanction states.²⁰

On reviewing an earlier draft of this report, DHHS stated in reference to the criterion of a requirement to enforce the law that there was no conclusive proof as to "what constellation of interventions will most effectively reduce youth access to or use of tobacco products" (Nelba Chavez, PhD, written communication, October 6, 1998). The ineffectiveness of merchant education programs in the absence of enforcement is confirmed by the fact that they were implemented in nearly every state, yet violation rates in almost every state remained unacceptably high.^{4 - 7} Very high levels of merchant awareness often coexist with very high violation rates, suggesting that merchants have little fear of punishment (Connecticut and Tennessee). Enforcement programs that did not use underage decoys were completely ineffective (Connecticut, Tennessee, and Puerto Rico). Inspections for proper placement of signs and vending machines had little or no impact (Connecticut and Tennessee). Despite this dismal record, DHHS still considers merchant education and passive inspections effective enforcement.²²

The DHHS also expressed its belief that states should only be required to do the best they can under the constraints of their current tobacco laws: "States would be required to implement enforcement systems commensurate with their individual and disparate laws" (Nelba Chavez, PhD, written communication, October 6, 1998). With the failure of DHHS to sanction states that do not enact or enforce effective laws, legislators in many states have repeatedly rejected public health initiatives designed to implement enforcement and have instead passed laws protecting tobacco merchants from prosecution (Arizona, Georgia, Louisiana, Maryland, New Mexico, North Carolina, Oregon, Texas, Virginia, and Wisconsin).¹⁸ As of 1996, 27 states had laws containing provisions that could hamper enforcement (Table 1). Many nonenforcing states choose to educate or warn violators rather than issue citations because prosecution under the current state law is not a practical option. Clearly, the intent of Congress was not to encourage "a race to the bottom" as states enact legislation to disable their ability to enforce the law, but this is what has happened through tobacco industry lobbying and the failure of DHHS to set minimum standards for laws and their enforcement.

The DHHS may be reluctant to enforce the Synar Amendment because it represents an unfunded federal mandate. However, for the next 25 years, every state will receive a windfall in uncommitted revenue from the tobacco industry as reimbursement for tobacco-related expenses incurred in part through the federally funded Medicare program.²⁸ The federal government could negotiate to relinquish its claim to a share of these settlement dollars if a designated proportion of those funds are used to enforce the state restriction on the sale of tobacco to minors. The Synar Amendment has failed thus far because it has never, in fact, been put into effect. The solution is for DHHS to put some teeth into the federal enforcement of the Synar Amendment now that states are awash in tobacco settlement dollars.

For the most part, the failure of the designated state agency to conduct a scientific survey can be attributed to the failure of state governments to allocate funding for this activity. The DHHS does not require sufficient standardization in the conduct or reporting of the surveys to allow for the valid comparison of the year-to-year performance that forms the heart of the DHHS implementation strategy.²¹ A small difference in the age of the youths used to purchase tobacco can make a significant difference in the measured violation rate.⁵ Age standardization across states could be achieved by DHHS by simply changing the reporting form to request age-specific violation rates. The failure of DHHS to require a standardized method of conducting the statewide surveys represents a tremendous lost research opportunity. Standardized methods across states would have permitted a comparison of the effectiveness of various enforcement methods and an evaluation of the impact of different violation rates on the prevalence of tobacco use among minors.

The minimum frequency with which merchants must be inspected to reduce youth access to tobacco has not been established, but the successful community-level enforcement programs have used active testing of all merchants several times annually.^{9 - 10 ,29} Only South Dakota and Massachusetts came close to testing each merchant a single time. Effective enforcement requires, at a minimum, an enforceable law, an adequately funded and staffed enforcement agency, and a mandate to test all merchants. Few states indicated an understanding of what effective enforcement is, or how they will move beyond ineffective merchant education programs. Given the complete lack of reported enforcement in many states, and only sporadic local enforcement in many others, it appears that roughly half of all states had never expended a penny to enforce their law despite their repeated certifications that they were effectively doing so. The vast majority of states take in more tax revenue from cigarettes consumed by minors than they spend to enforce their tobacco access laws.^{30 - 32}

The criterion used to determine if states were enforcing their laws was extremely generous since it is unreasonable to believe that penalizing a single merchant constitutes effective statewide enforcement. Others may disagree with the author's assertion that the Synar Amendment requires states to enforce the law by conducting inspections. Since every state that failed to conduct enforcement inspections also failed to provide evidence of any other enforcement activity, this point is moot (Table 1). It is possible that some states had penalized merchants without reporting it to DHHS. Indeed, it is likely that a comprehensive polling of every community in every noncompliant state would turn up isolated examples of merchants being penalized for violating the law. This would not alter the conclusions of this study about the lack of effective statewide enforcement programs or the assertion that DHHS is certifying states without evidence that the law had been enforced.

No states allowed their decoys to lie about their age as real underage smokers may do. South Carolina confirmed the author's personal observation that store clerks sometimes coach youths to lie about their age. If the youth lies, the clerk can be sure that the youth is not a decoy. In a recent survey in Vermont, when youths were permitted to misrepresent their ages if asked, illegal sales rates increased "by up to 8%" (Thomas Radecki, MD, JD, written communication, September 12, 1998). This casts doubt on the validity of using surrogates who cannot lie to measure the availability of tobacco to underage tobacco users who can lie.

Finally, Smoke Free Pennsylvania has repeatedly expressed concerns about the veracity of the information supplied in state applications (William Godshall, MPH, oral communication, September 17, 1998). States have a strong financial incentive to paint a rosy picture. The poor quality of the state reports that were analyzed for this audit, and relied on by DHHS, is also of concern. If the poor data quality introduced bias in this article, it is most likely in the direction of overestimating the extent to which states are enforcing their tobacco access laws.

Accepted for publication February 23, 1999.

This study was supported by a grant from The Robert Wood Johnson Foundation, Princeton, NJ.

Editor's Note: Another good law goes up in smoke because of failure to enforce it.—Catherine D. DeAngelis, MD

Reprints: Joseph R. DiFranza, MD, Department of Family Medicine and Community Health, University of Massachusetts Medical Center, 55 Lake Ave, Worcester, MA 01655 (e-mail: Joe.Difranza@banyan.ummed.edu).