Background Under the 2009 Family Smoking Prevention and Tobacco Control Act, the Food and Drug Administration (FDA) has been routinely inspecting tobacco retailers' compliance with under-age sales laws. We seek to identify factors associated with Retail Violation Rate for sale to minors (RVRm).
Methods We collected the tobacco retailer inspection data for 2015 from the FDA compliance check database. RVRm was calculated at the census tract level and overlaid with tobacco regulations and youth smoking prevalence at the state level. Multi-level spatial analysis was performed to examine the impacts of tobacco jurisdiction variations, youth smoking rates and neighbourhood social characteristics on RVRm.
Results A total of 136 816 compliance checks involving minors conducted by the FDA in 2015 were analysed. A higher RVRm was associated with higher youth smoking prevalence (aRR=1.04, p<0.0001). Tobacco regulations show significant relationships with RVRm. For every one dollar increase in cigarette tax per pack, the likelihood of retail violations was reduced by 2% (aRR=0.98, p=0.03). For every 10% increase in tobacco prevention spending towards Centers for Disease Control recommended funding targets, the likelihood of retail violations was reduced by 1% (aRR=0.99, p=0.01). RVRm increased in states that enacted stronger smoke-free air policies (aRR=1.08, p<0.0001).
Conclusion We observed associations of tobacco regulations and neighbourhood social characteristics with tobacco retailers’ compliance with under-age sales laws. This study provides evidence to support stronger tobacco regulations and control policies in reducing youth access to tobacco products.
- Priority/special populat
Statistics from Altmetric.com
Restricting tobacco sales to minors is an important component of tobacco control and prevention strategies to reduce youth smoking and access to cigarettes. Tobacco use often starts during adolescence1 and 9 out of 10 smokers tried their first cigarette by age 18.2 Smoking at an early age can affect brain and lung development. Although the 2009 Family Smoking Prevention and Tobacco Control Act (FSPTCA) restricts tobacco marketing and sales to youth,3 the sale of cigarettes to minors remains a problem. Over 60% of respondents from the 2014 National Youth Tobacco Survey4 reported that it was ‘easy’ or ‘somewhat easy’ to buy tobacco products in stores. A previous study has summarised the inspection results by state for the Food and Drug Administration (FDA) compliance check program,5 but this study did not assess factors that might be associated with variations in inspection results. Lee and colleagues6 assessed the neighbourhood inequities in tobacco retailer sales to minors in 2015. However, tobacco control policies were not included in their analysis and little is known about effects of these policies on violations of under-age sales. This study seeks to evaluate neighbourhood and state-level contextual predictors of tobacco retailers' compliance with youth access by analysing the FDA compliance check database. First, youth smoking prevalence is impacted by state-level tobacco control policies (ie, cigarette tax, tobacco prevention spending and smoke-free laws).7 8 Smoking prevalence has been found to be higher in regions with weaker tobacco control regulations, such as the South or Midwest.9 10 Thus, we hypothesise that tobacco regulation policies along with youth smoking prevalence could impact retailers’ violations of under-age sales laws. Second, past studies have shown that retail stores selling traditional tobacco products are disproportionally concentrated in vulnerable communities (ie, young, ethnic minority and low-income communities).11–16 We hypothesise that tobacco retail sale violations might exhibit similar neighbourhood inequities.
Data and method
We collected the tobacco retail compliance data for under-age sales in 2015 from the FDA compliance check database17 and included only inspections indicated as ‘minor involved’. The Retail Violation Rate for sale to minors (RVRm) is calculated by dividing the number of violations by the number of inspections involving minors at the census tract level. Prevalence of current smoking by youth was collected from Campaign for Tobacco-Free Kids, which compiles state-level data from the Youth Risk Behavioral Surveillance, Youth Tobacco Surveillance and state-specific surveys.18 State cigarette tax rate data were collected from the Centers for Disease Control and Prevention (CDC)19 and state smoke-free air law data were collected from the American Lung Association.9 Census tract-level predictor variables include socio-economic status (SES) variables (ie, gender, race, age, education and poverty) collected from the 2014 American Community Survey (ACS) and urbanicity data collected from the United States Department of Agriculture Economic Research Service.20
ArcGIS, V.10.3.1, was used to geocode tobacco retailers’ addresses and overlay RVRm with the tobacco control policies, census tract SES and urbanicity. A multi-level model was developed to incorporate predictors at the state and census tract levels. As geographically neighbouring census tracts might be more similar than distant ones, we tested the spatial autocorrelation using SAS ‘Glimmix’ procedure.21 Since many census tracts had no violation, a zero inflated negative binomial (ZINB) regression model was used to predict RVRm.22 In ZINB, the dependent variable is the count of violation at each census tract, and log(RVRm) is further derived by introducing the number of inspections as an offset variable under log transformation. Adjusted relative risk (aRR) and 95% CI were reported. Statistical analyses were performed using SAS 9.4 (Cary, North Carolina) and p value <0.05 was considered statistically significant.
Inspections involving minors (n=1 36 816) were mapped to 32 124 census tracts (see online Supplementary Table 1 for RVRm by state). A higher RVRm was associated with higher youth smoking prevalence (aRR=1.04, p<0.0001). For every one dollar increase in cigarette tax per pack, the likelihood of retail violations was reduced by 2% (aRR=0.98, p=0.03). For every 10% increase in tobacco prevention spending towards the CDC’s recommended target level, the likelihood of retail violations was reduced by 1% (aRR=0.99, p=0.01). RVRm was higher in states enacting stronger smoke-free air policies (aRR=1.08, p<0.0001, see table 1).
Supplementary table 1
This analysis has been adjusted by neighbourhood SES as tract-level covariates. We observed higher risks of RVRm among census tracts with a higher proportion of males (aRR=1.6, p=0.01), African Americans (aRR=1.6, p<0.0001), and Hispanics (aRR=1.5, p<0.0001). RVRm also varied by age distribution and was higher in urban areas (vs non-urban areas) and census tracts with a higher poverty rate.
Summary and discussion
We found significant associations between the RVRm and the youth smoking rates and tobacco control policies. Strong enforcement at the community level and youth access interventions have been shown to have effects on reducing adolescent smoking.23–25 Compliance checks using under-age decoys can significantly reduce illegal sales to minors.26 High youth smoking prevalence could trigger more attempts by youth to purchase cigarettes from retail stores. Studies show that regular youth smokers are more likely to purchase cigarettes in stores.27–30 Our finding confirms that under-age sale violations were more likely to occur in states with high youth smoking rates. Both cigarette tax and tobacco prevention spending as percentage of the CDC target have been proven to be effective in reducing youth smoking prevalence.8 31–33 Our study further demonstrates the negative relationship between these programs and RVRm. Surprisingly, RVRm was higher in states that enacted stronger smoke-free air policies than other states. One possible reason is that the states with stronger smoke-free air policies more strictly enforce compliance inspections, resulting in a higher RVRm. Another plausible reason is that youth located in states with stronger smoke-free air policies are less likely to obtain tobacco products illicitly from social sources (ie, friends or family members), and thus they are more likely to purchase cigarettes from retail stores.
Consistent with past studies,6 34 our study identified neighbourhood social inequities related to FDA compliance inspections for under-age sales. A higher RVRm was associated with poverty and larger proportions of African Americans and Hispanics. RVRm was lower in neighbourhoods with a larger proportion of youth. A possible reason for these age inequities is that tobacco retailers located in areas with a higher proportion of youth might be more cautious in complying with the FSPTCA and less likely to sell tobacco products to youth. Consequently, to improve their chances of getting tobacco products, youth might intentionally choose tobacco retailers located in areas with a lower proportion of youth.27 Oversampling tobacco retailers in high-risk communities could be implemented to reduce neighbourhood inequities in under-age sales.
Although the FDA manages the routine inspections of tobacco retailers, it subcontracts with each state to conduct the inspections. Thus, enforcement might be different across geographic regions. In 2015, the number of inspections ranged from 55 (OH) to 13 955 (FL) and the RVRm ranged from 1.4% (MT) to 29.4% (NH) (see online Supplementary Table 1). Our findings are consistent with previous studies5 35 and indicate that the protocol for inspections might be inconsistent across states. Continued efforts to increase the number of compliance checks across states and development of an optimal sampling design to improve enforcement efficiency are needed.
Our findings should be interpreted with caution in light of limitations. First, since the FDA does not release the compliance inspection sampling method and weights,36 the FDA compliance checks might not be nationally representative and there might be sampling bias in calculating the RVRm. Some states might selectively oversample in areas with higher likelihood of illegal sales to increase compliance. Second, RVRm was calculated at the level of each inspection and has been challenged as an inaccurate measure of the true violation rate,37 since youth could self-select tobacco retailers known to sell to minors. In addition, repeated inspections were included in the analysis and this might introduce bias in calculating true RVRm. Third, this study only included the census tracts with compliance checks in 2015 as the census tracts with no compliance checks could be the ones with no tobacco outlets or the ones with tobacco retailers but no compliance checks. Finally, the analysis was performed at the census tract level and some census tracts with a small number of inspections might have an unstable RVRm.
Our cross-sectional study demonstrates that tobacco control policies were associated with retailers’ violations of under-age sales laws. Through ‘Tobacco 21’ campaigns, two states (Hawaii and California) and at least 145 localities in 11 states have raised the minimum smoking age from 18 to 21 years.38 These new regulations will have additional impacts on enforcing retailers’ compliance with under-age sales laws. Thus, continuous surveillance of the impact of jurisdiction variations on retailers’ compliance is needed.
What this paper adds
Restricting tobacco sales to minors is an important component of tobacco control strategies to prevent youth from smoking. The neighbourhood inequities have been found in tobacco retailer sales to minors.
This study evaluated neighbourhood and state-level contextual predictors of tobacco retailers' compliance with youth access. We found significant associations between tobacco retailers’ violations of under-age sales and youth smoking rates and tobacco control policies.
This study provides evidence to support stronger tobacco regulations and control policies in reducing youth access to tobacco products.
Contributors HD conceptualised the study. JH acquired data. HD and JH performed analyses. Both authors contributed to writing the manuscript. We thank the Medical Writing Center at Children’s Mercy Hospital for editing this manuscript.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement Data are publicly available.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.