Internet Alcohol Sales to Minors

Objectives To determine whether minors can successfully purchase alcohol online and to examine age verification procedures at the points of order and delivery.

Design A cross-sectional study evaluated underage alcohol purchase attempts from 100 popular Internet vendors.

Setting The study was conducted at the University of North Carolina at Chapel Hill, July 14-27, 2011.

Participants Eight 18- to 20-year-old individuals participated.

Outcome Measures Rates of successful sales to minors and use of age verification procedures at order and delivery were determined.

Results Of the 100 orders placed by the underage buyers, 45% were successfully received; 28% were rejected as the result of age verification. Most vendors (59%) used weak, if any, age verification at the point of order, and, of 45 successful orders, 23 (51%) used none. Age verification at delivery was inconsistently conducted and, when attempted, failed about half of the time.

Conclusions Age verification procedures used by Internet alcohol vendors do not adequately prevent online sales to minors. Shipping companies should work with their staff to improve administration of age verification at delivery, and vendors should use rigorous age verification at order and delivery. Further research should determine the proportion of minors who buy alcohol online and test purchases from more vendors to inform enforcement of existing policies and creation of new policies to reduce youth access to alcohol online.

Internet alcohol sales have received relatively little legislative attention in recent years compared with Internet cigarette sales, which are now regulated by 34 state laws,¹ 2 federal laws,^2,3 and 4 federal agreements.^4-7 One reason that Internet cigarette sales garnered such public health and regulatory attention was that few Internet cigarette vendors (ICVs) properly verified buyers' ages, resulting in as many as 92% selling to minors.^8,9 The issue of age verification is potentially a concern for the Internet alcohol vendor (IAV) industry as well, but has not been the focus of what little regulatory attention has been given to this $2.4-billion-a-year industry (verbal and written communication, Wanda V. Vega-Garcia, BS, Service Sector Statistics Division, Retail Census Branch, US Census Bureau, January 27, 2012).¹⁰ Instead, the legislative focus for Internet alcohol sales has primarily been on loosening commerce restrictions to allow interstate shipment of wine from vineyards^11,12 rather than on youth access prevention.

This issue came before the US Supreme Court in 2005 in Granholm v Heald.¹³ The court ruled that laws in New York and Michigan allowing in-state but not out-of-state wineries to ship wine to consumers were unconstitutional because they interfered with free trade and interstate commerce by giving an unfair economic advantage to in-state wineries. The states claimed that these sales undermined their ability to police underage drinking, asserting that minors have easy access to credit cards and are likely to purchase alcohol online, but the court found that

“The States, aided by several amici, claim that allowing direct shipment from out-of-state wineries undermines their ability to police underage drinking. Minors, the States argue, have easy access to credit cards and the Internet and are likely to take advantage of direct wine shipments as a means of obtaining alcohol illegally. The States provide little evidence that the purchase of wine over the Internet by minors is a problem.”^13(p26)

No peer-reviewed studies have examined the sales and age verification practices of IAVs and assessed whether minors can purchase alcohol. Several anecdotal reports, however, have shown that youth can easily obtain alcohol from IAVs. In 2003 and 2004, student studies at Gonzaga University¹⁴ and the University of Tennessee¹⁵ reported underage buyers receiving online alcohol orders without being questioned about their age. In 2006, an investigative reporter detailed 2 cases in which a 15-year-old individual received alcohol without age verification, even though one of the packages was clearly labeled as wine and in need of an adult signature.¹⁶ In addition, a 2004 undercover operation in Massachusetts led to lawsuits against 4 IAVs for selling to minors.¹⁷

The goals of this study were to determine whether minors can successfully purchase alcohol from Internet vendors and to examine age verification procedures at the points of order and delivery.

One hundred popular alcohol vendor websites composed the study sample. A private sector online risk-monitoring firm (Cyveillance, Inc) helped to develop the strategy for identifying IAV websites.¹⁸ Specially developed search algorithms and intelligent web spiders reviewed more than 40 million websites, postings to approximately 100 000 message boards and newsgroups, and 1 million spam e-mail messages to identify websites that were likely to be IAVs based on key words appearing in text and features indicative of online retailers (eg, prices and secure shopping carts). Each website on the list of possible IAVs was reviewed manually by trained data collectors to determine whether it was an English-language website direct shipping alcohol to consumers. In cases in which websites on the list turned out to be online shopping portals for IAVs, all links were reviewed as potential IAVs.

A similar approach for a longitudinal study of ICVs¹⁸ yielded at its peak 775 websites, and the present study was modeled after that one. Using the same procedures to identify IAVs, however, yielded a larger number of websites, and, because of budgetary constraints, only the first 5000 sites identified could be screened. Alexa.com traffic-ranking data based on the number of unique monthly visitors to each site were used to identify the 100 most popular IAVs after applying the exclusion criteria detailed in Table 1 (eg, vendor does not ship to North Carolina).

The study buyers were eight 18- to 20-year-old English-speaking individuals who lived in the county where the study was conducted. At this age, they were adults but still minors according to the legal age to purchase alcohol. Purchases were divided among multiple buyers (with each making 11-14 purchase attempts) to minimize the chances that delivery drivers' age verification attempts might be biased by a growing recognition of recipients. A letter of immunity from prosecution was obtained from the local district attorney to protect all staff and buyers involved in the study.

In July 2011, under one-on-one staff supervision and using procedures approved by the University of North Carolina institutional review board, buyers visited the study websites and attempted to purchase the minimum order of the lowest-priced alcohol available, tracking the process in the study's online data collection system. The buyers recorded details of the order (alcohol type, brand, volume, cost, shipper, age warnings, and age verification attempts) and the delivery (shipping company, whether there was human interaction, and age verification attempts).

A large proportion of the IAVs in the study sample exclusively sell wine, and FedEx and UPS have policies and procedures for age verification at delivery (AVAD) for wine shipments; it is against FedEx and UPS policies to ship any other type of alcohol (eg, beer, liquor). Both companies have official stickers for labeling packages as alcohol requiring AVAD, as well as labeling embedded in (and printed underneath) their shipping label barcodes to trigger AVAD by delivery drivers (the second feature can be used for AVAD of nonalcohol products without the alcohol sticker). It is against federal law to ship any alcohol via the United States Postal Service.¹⁹ When nonwine products were available, buyers ordered them to test how they would be shipped and to maximize the data available on sales of these products. To ensure unbiased representation of all shipping companies, in cases in which vendors offered multiple carrier options, the shipper was randomly selected. Purchases were made using Visa debit gift cards purchased with cash.

When encountering age verification, buyers were allowed to misrepresent their age by clicking a button or checkbox or typing a false birth date. However, if they were required to provide identification (ID) to complete a purchase, they provided their real North Carolina driver’s license, which uses 3 visual cues to clearly identify individuals younger than 21: vertical orientation, a colored border around the photo, and text in that border indicating the date that they will become 21.²⁰ Buyers were not allowed to alter their own ID, use a friend's or relative's ID, or use any other strategies to thwart age verification. Although youth outside a study setting could use these strategies, this protocol measured the extent to which minors can successfully obtain alcohol without engaging in these illegal activities.

When packages were delivered, buyers recorded the date, by whom the package was delivered, details of age verification attempts, and whether the package was labeled for AVAD.

Of the 100 online alcohol purchase attempts made by underage buyers in this study, 45 were successfully received (Figure). Of the remaining orders, 12 were rejected at or shortly after the point of order as a result of age verification and 16 were rejected after the delivery driver checked the recipient's ID. Eleven more failed at the point of delivery because no one was home to receive the order, although it is unclear whether these IAVs had requested a signature or AVAD. The remaining 16 orders failed for other reasons that appear to be unrelated to age verification (eg, technical and communication problems with vendors).

Most (65%) of the websites sold wine exclusively. When other categories of alcohol were available, they were purchased in lieu of wine (Table 2). The IAVs that sold nonwine products were more likely than wine-only vendors to sell to minors, but the differences were not statistically significant (P = .50). While 42% of wine orders (n = 27) were successfully received, 53% of liquor orders (n = 9) and 57% of beer orders (n = 8) were successfully received.

There was very little use of age verification at the point of order (Table 3). In fact, age warnings were infrequent (18%) on the 100 IAV websites, especially by those that sold to minors (9%). Most vendors (59%) used weak age verification, if any, relying on checkboxes or buttons (31%) or spurious claims that by merely submitting an order, users were legally certifying their age (23%). Many vendors (41%) did not address age verification at the point of order at all. Of the orders successfully received, 71% (n = 32) did not use rigorous age verification at the point of order and 51% (n = 23) used none. Orders from vendors that used weak or no age verification at the point of order were significantly more likely to be successful than were those from vendors using more rigorous age verification (P = .03).

Buyers entered a false date of birth for the 39 vendors that requested their birth date. Only 3 of those orders (7.7%) were rejected for age verification reasons; those vendors may have used an online age verification service to determine that the date of birth did not match that in the buyer's government records.

In all 5 cases in which the vendor collected a driver's license number, they rejected the sale after receiving the buyer's actual license number, indicating that the number was used to verify the buyer's age. At the point of order, 2 vendors requested that the user enter their driver's license number and 1 asked the user to send a copy of the license. Two more contacted the buyer after the order was placed and requested a copy of the license.

Only 2 vendors indicated use of online age verification services on their websites. One was unspecified and blocked the underage sale. The other used IDology, and not only failed to block the sale at the point of order but also failed to block the sale at delivery; the FedEx delivery driver handed the recipient the package after merely asking whether the recipient was older than 21 years.

Age verification at delivery was inconsistently administered and, when used, failed about half the time (Table 4). A total of 47 packages arrived labeled as requiring AVAD. Twenty-three featured the carrier's AVAD and alcohol labeling and 7 used the carrier's AVAD barcode labeling without the alcohol sticker. One was labeled as requiring age verification but not using the carrier's official labeling standards. Because delivery was refused on the basis of age verification and the packages were returned to the senders, it is unknown how the remaining 16 packages were labeled.

Many deliveries did not include any attempts to verify age. Sixteen orders (35.5% of successfully received orders, 10.6% of orders labeled for AVAD) were left at the door with no human interaction, and 14 (31.1% of successful orders, 25.5% of orders labeled for AVAD) were handed to the recipient with no attempt to verify age.

Delivery staff attempted to verify the recipient's age for 30 orders but blocked only 16 of those. In 14 cases (31.1% of successful orders), the alcohol was delivered to underage recipients after taking their word that they were older than 21 years or after examining their vertical driver license that clearly labeled them as being younger than 21. In one case, the driver “looked it over, claimed it was a new license and he didn't know how to read it, looked at it a few seconds longer, then had me sign for the delivery.”

FedEx did a poorer job than UPS of verifying age at delivery. About a third of each company's deliveries labeled for AVAD were returned to the sender after delivery was refused on the basis of age verification. However, in cases in which delivery staff attempted to perform AVAD, FedEx packages were significantly more likely to be delivered to the underage buyer (P = .007).

Four orders (3 FedEx, 1 UPS) were picked up by recipients at the company's package distribution centers. All were labeled for AVAD and in each case, despite showing an underage ID, the minor was given the package. One buyer said, “The clerk checked my ID, pointed at it right where it said WILL TURN 21 IN 2014 and said ‘OK!,’ and gave me the package.”

In another case, after making several attempts to deliver a package labeled for AVAD while a buyer was out of town, the driver simply left the package at the door rather than returning it to the sender. On a separate delivery, because the buyer was not home, the driver delivered 2 packages to the buyer's middle-aged neighbor, who in turn delivered the packages to the buyer, unaware of what was in them.

With 45 of 100 alcohol purchase attempts successfully received by underage buyers, the age verification procedures in use by IAVs clearly do not adequately prevent online alcohol sales to minors. This sales rate is comparable to the rate of offline retail sales in communities where there is little or no enforcement. In such areas, individuals who look younger than 21 years can buy alcohol without showing ID in 45% to 50% of attempts.^21-23 However, compliance checks and active enforcement of youth access laws in retail stores have been shown to reduce underage sales by as much as half.^24-27 Enforcement operations that replicate our findings and cite vendors and or shipping companies might also reduce sales to minors.

For online stores, AVAD is the only opportunity for face-to-face age verification (the standard in retail stores), yet fewer than half of the vendors used this option. Furthermore, in cases in which the vendor used AVAD, it was frequently implemented improperly by delivery services, resulting in a substantial number of successfully received orders. In 36.1% of cases in which the vendor paid for AVAD, delivery staff failed to administer it. In the cases in which the delivery staff did attempt AVAD, they failed to do so properly about half the time. Delivery staff examining recipients' driver licenses should have been able to consistently reject the deliveries; each buyer's license used no less than 3 strategies to very clearly label the owner as younger than 21 years.

It is very important that shipping companies work with their staff to ensure more faithful execution of their policies and procedures for alcohol shipments. In this study, shipping companies frequently delivered beer and liquor in violation of their own policies, and AVAD was often executed poorly, if at all. Furthermore, it is important that alcohol be delivered only to the individual who ordered it, not to neighbors who may unwittingly participate in the delivery of alcohol to underage buyers. While the delivery driver technically delivered the packages to an adult, they also facilitated the underage buyer in obtaining alcohol. Also, although AVAD is very important, IAVs should not rely so heavily on using only AVAD to prevent youth access. Because age verification left to delivery drivers was frequently done incorrectly, IAVs should also use rigorous age verification at the point of order.

It is important to note that buyers in this study were barred from using a strategy to bypass age verification that minors overwhelmingly say they have access to and are very willing to use: a parent's driver license.²⁸ Future studies should assess the ease with which youth in real-world circumstances (ie, with access to a friend's or parent's license) can bypass age verification, as well as vendors' ability to thwart such purchases by using rigorous age verification at both the points of order and delivery.

This study's purchases were made using prepaid Visa gift cards, which necessarily limited the study sample to domestic websites (these cards were not usable for international purchases). They were selected after we concluded that they would provide the greatest protection against buyers being linked to their study participation. Future studies should include international vendors because, in the ICV industry, many vendors ship from overseas to offer lower prices by avoiding payment of excise taxes, and it is important to determine whether this is also an issue for IAV sales.

Prepaid gift cards are easy for youth to obtain and are difficult to track. The cards were purchased with cash without ID, were managed online, and nothing was ever mailed to the user; it would be easy for minors to buy and maintain such cards for online purchases without their parents being aware.

This study was limited to the 100 most popular IAVs, which were disproportionately wine vendors (vs beer, liquor, or other alcohol, which are more frequently used by youth²⁹). Although this study provides important evidence that these IAVs do a poor job of preventing youth access, further research is needed to determine the proportion of minors who buy alcohol online and to guide the formation of public health policy on online alcohol sales and age verification, with samples including larger numbers of vendors selling nonwine alcohols.

At the peak of the ICV industry, the methods described herein identified 775 ICVs.¹⁸ Owing in large part to extensive regulatory attention, that number dwindled to 392 in 2009 before beginning to rise again (R.S.W., unpublished data, 2009). Considering that there are generally fewer cost advantages to buying alcohol online as compared with cigarettes (because of lower excise taxes and higher shipping cost for alcohol vs cigarettes), we expected to find substantially fewer IAVs than ICVs. However, we had to stop counting when we reached 5000 because of budgetary constraints—clearly there were more to be found. The community of IAVs is far larger than the community of ICVs; future research should more completely identify the population and capture data on sales and age verification practices from a greater proportion of the IAV population.

The application of similar approaches that have been used to regulate ICV sales to IAV sales may be effective in regulating this industry. As with ICVs, issues of jurisdiction, interstate sales, and the sheer number of vendors make it difficult for state officials to intervene directly with the IAVs. However, working at the federal level to cut off vendors from their established shipping and payment-processing partners could, as it did with ICVs, lead to an increase in vendors going out of business and a substantial decrease in vendors using banned shippers and payment processors.³⁰ Greater enforcement of existing policies, or perhaps new policies, are needed to effectively reduce youth access to alcohol online.

Correspondence: Rebecca S. Williams, MHS, PhD, Center for Health Promotion and Disease Prevention, University of North Carolina at Chapel Hill, 200 N Greensboro St, Ste D-13, Carrboro, NC 27510 (rebeccawilliams@unc.edu).

Accepted for Publication: February 29, 2012.

Published Online: May 7, 2012. doi:10.1001/archpediatrics.2012.265

Author Contributions: Dr Williams had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Williams. Acquisition of data: Williams. Analysis and interpretation of data: Williams and Ribisl. Drafting of the manuscript: Williams and Ribisl. Critical revision of the manuscript for important intellectual content: Williams and Ribisl. Statistical analysis: Williams. Obtained funding: Williams and Ribisl. Administrative, technical, and material support: Williams. Study supervision: Williams.

Funding/Support: This work was funded by grant 64747 from the Robert Wood Johnson Foundation Substance Abuse Policy Research Program.

Role of the Sponsor: The funding source had no role in the design and conduct of the study; in the collection, analysis, and interpretation of the data; or in the preparation, review, or approval of the manuscript.

Additional Information: This study of the practices of Internet alcohol vendors was a part of the authors' larger longitudinal study of Internet cigarette vendors ongoing since 1999 and funded from 2008-2011 by a grant from the Robert Wood Johnson Foundation Substance Abuse Policy Research Program. The grant funded ongoing surveillance of Internet cigarette vendors as well as exploratory research to identify Internet alcohol vendors, analyze their website content, compare their practices with those of Internet cigarette vendors, and examine the extent to which Internet alcohol vendors prevent sales to minors.