Notice No. 917
Submitted by the Center for Science in the Public Interest
August 20, 2001

Since 1982, the Center for Science in the Public Interest (CSPI) has promoted enhanced consumer information about alcoholic beverages. In the mid- and late-1980s, CSPI participated in efforts that led to the passage of the Alcohol Beverage Labeling Act of 1988, which mandated health warning labels on alcoholic-beverage containers and the development of regulations implementing that Act. In 1999, together with 121 other organizations and four members of Congress, CSPI petitioned the Bureau of Alcohol, Tobacco, and Firearms to issue rules to improve the placement, legibility, and noticeability of the congressionally mandated warning label. We are pleased to offer the following comments in response to BATF's Advance Notice of Proposed Rulemaking (ANPRM) of May 22, 2001.


In order to comment meaningfully and comprehensively to the numerous questions posed by the Bureau in its ANPRM, CSPI commissioned the research firm, Penn, Schoen, and Berland Associates, to conduct a national survey of public opinion on issues concerning the noticeability, legibility, size, conspicuousness, and effectiveness of the current health warning labels found on alcoholic-beverage containers. That poll, the results of which are attached as Exhibit I and incorporated herein, was conducted from July 29 to July 31, 2001 and reports the opinions of a nationally representative sample of 801 adult Americans, segmented by gender and drinking status.

Overall, the Penn, Schoen poll revealed that, although approximately two-thirds of the respondents were aware that a health warning statement is required by law, few people notice the statement or look for it on containers of alcoholic beverages. Most are not very familiar with the information contained in the warning, and a majority reported that the warning statement can be difficult to read because the "words are too small." Nearly half of the drinkers who responded to the survey said that they sometimes find the warning statement on alcoholic beverages difficult to notice, a third of them reporting that the statement was "very difficult" to notice. In addition to asserting that the statement was too small, respondents strongly indicated other deficiencies in the appearance of most labels, including their inconspicuous placement, lack of a contrasting background, and illegibility under ordinary conditions.

Overwhelmingly (74%), American consumers (including 71% of drinkers) agree that people would be more aware of what the warning message stated if it were easier to read, and agree that the messages sometimes appear in the least prominent place on containers (74% of drinkers), making them difficult to notice and read. Significantly, in an implicit indictment of current government requirements on warning labels, a strong majority of 72% thought that warning labels would be more visible and readable if the government was actually serious about people noticing them. Even stronger majorities -- including drinkers -- believed that many of the label improvements proposed by CSPI and others would be effective in making them more noticeable and readable.


1. Are consumers aware that the health warning statement must appear on the label of all alcohol beverages sold in the U.S.?

Overall, nearly two-thirds (65%) of Americans report awareness of the warning label requirement, with drinkers registering higher awareness (72%) and women, less (61%). Barely half (54%) of non-drinking consumers know that a warning statement must appear on alcohol beverages sold in the U.S.

2. Do consumers look for the warning statement on alcohol beverage containers?

Most drinkers (62%) either never look for the warning statement or seldom look for it (24%) when they buy alcoholic beverages. Only 7% of all adult consumers report that they look for it every time, frequently, or sometimes. Significantly, far more men than women report that they do not look for the warning statement, 67% of men compared to 52% of women.

3. Do consumers notice the health warning statement on alcohol beverage containers? Explain.

For most consumers, the warning notice rarely attracts attention. Among drinkers, who apparently handle the containers of alcoholic beverages on a more regular basis, only 34% usually notice the warning statement. Fewer than one-third (32%) of all adult consumers notice the warning statement. When asked how often they noticed the warning statement, only 18% of drinkers responded "all of the time" or "most of the time." Fifty-two percent said "rarely" or "never."

Not surprisingly, many drinkers (48%) -- as many who noticed the label as didn't -- reported that they sometimes find the warning statement difficult to notice, and more than one-third (37%) among those consumers found it "very difficult" to notice. Among the most common reasons cited for this difficulty: the warning statement is too small, the color of the message blends into the background, and the statement is not in a prominent place on the bottle or can. In fact, only one in four drinkers (25%) -- and 20% of all consumers -- say that the warning labels are located on containers in a conspicuous and prominent place "all the time," compared with 40% who say "not very often" or "hardly ever." Among female consumers, 44% say the labels are "not very often" or "hardly ever" conspicuous.

Drinkers overwhelmingly agree -- by a three to one margin (74% agree vs. 25% disagree) -- that warning messages sometimes appear in the least prominent place on bottles and cans and that this makes them difficult to notice and read. Similarly, two in three drinkers (65%) agree that warning messages often appear on a background color that makes them hard to read.

4. Do consumers read the warning statement on labels of alcohol beverages? Why or why not?

Fewer than half (47%) of all consumers and 54% of drinkers claim that they have ever read the entire wording of the health warning statement. Even though the label notice is brief, that is not surprising, given what we learned about the difficulty that consumers have in noticing it, and the fact that relatively few consumers even look for the statement on a regular basis. Additionally, more than half of the drinkers polled (58%) and nearly two-thirds (64%) of all consumers think the warning statement can be difficult to read because the words are too small. Another 10% cite the statement's location on the label as an impediment to reading.

The poll also found that only one in five drinkers (21%) and 18% of all consumers think that warning labels appear all the time with a contrasting background that makes them stand out and makes them easy to read. More than half (52%) of the drinkers said that the labels stand out "hardly ever" or "not very often." Only 40% of drinkers and barely one-third of American consumers (34%) -- and only 27% of females -- think that the words on health warning labels are always clearly legible or readable under ordinary conditions.

Respondents offered numerous suggestions for changes in health warning labels to make them easier to read. The top suggestions included using bigger type and making the warning bigger, putting the warning in a more noticeable place on the container, using bolder type, and requiring brighter and contrasting colors.

5. Are consumers familiar with the information contained in the alcohol health warning statement?

Based on responses in the poll we conducted, relatively few consumers consider themselves familiar with the information in the alcohol health warning statement. Half of all consumers (49%) reported that they are either "not very familiar" or "not at all familiar" with the information, compared to just 17% who said they were "very familiar." Even among drinkers, who should have the greatest familiarity, 42% said they were either "not very familiar" or "not familiar at all" with the health warning information.

When we probed further to determine what consumers actually knew about the information in the warning label, we discovered that their recollection of the label warnings is quite limited. Thirty percent (30%) of drinkers didn't recall anything about the label or didn't know. Only 40% of drinkers and 37% of all respondents recalled the statement that "women should not drink during pregnancy"; 29% of drinkers and 26% of all consumers remembered "women should not drink during pregnancy because of birth defects." When asked whether the statement warned any type of person not to drink alcoholic beverages at all, and asked to identify that person, one in four drinkers and almost one-third (31%) of all respondents didn't know; only 60% of drinkers and barely half (53%) of all consumers named "women who are pregnant."

Consumers, however, revealed greater knowledge of the risk of alcohol-impaired driving, with 69% of drinkers and 64% of respondents overall recalling that warning on the label. From these and other findings (provided in the attached polling data) we can only conclude that relatively few consumers are familiar with the information contained in the alcohol health warning statement.

6. Do consumers find the warning statement on alcohol beverages difficult to read? Explain.

More than one in five consumers (22% of drinkers) sometimes find the health warning statement on alcoholic beverages difficult to read, and only three out of five (62% of drinkers) find reading it not difficult. We assume that if one were to notice the label and focus on it, most consumers could read the simple language, as if they were reading a newspaper. The problem is that so few actually notice the label. Other findings in the poll suggest, however, that consumers do have difficulty in reading the label. Two-thirds (64%) of respondents -- 58% of drinkers -- reported that the warning statement "can be difficult to read" because "the words are too small." Another 13% of drinkers cited the "location on the label" as an impediment to reading the label. Two in three consumers (66%) think that health warning labels don't appear clearly legible or readable under ordinary conditions all of the time, including more than half (58%) of the drinkers.

In its November 1999 petition to BATF, CSPI provided numerous examples of label warning statements that were either obscured by a non-contrasting background, too small and crowded into other information, or hidden in a non-prominent area on the container (even the neck ring of a beer bottle). All were printed in all upper-case letters, squashing the message and making it more difficult to read. The facts speak for themselves. Such labels are difficult to read. Our survey confirms that consumers have experience with the same problems we found.

7. Do consumers have examples of alcohol beverages where the warning statement is legible and noticeable? What makes the warning statement legible and noticeable?

When we asked drinkers (in an open-ended question) to identify aspects of the warning label that could be improved to make it more noticeable and easier to read, one in three drinkers (29%) suggested making the type and warning bigger. Others recommended putting the notice in a more noticeable place, using contrasting and brighter colors, bolder and clearer type. Those suggestions echo the results of warning label research, which has established numerous criteria (cited in CSPI's November 1999 petition) for the presentation of effective warning label statements.

Further inquiry clearly indicates that consumers strongly think various improvements to the current labels would make them easier to notice and read. For example, 84% of drinkers say that placing the warning label in a prominent place on the front of alcohol containers would be effective, more than half (53%) saying "very effective"; and 78% of drinkers say that placing the message in the same relative position on all bottles and cans would make them easier to notice and read. Other findings demonstrate that, by strong majorities, consumers believe that all the improvements proposed in the CSPI petition, which BATF set out in its May 22 ANPRM, will be effective in making the warning statement more noticeable and readable. Please see the attached poll for further details.

8. Do consumers believe the regulations need to be amended to make the warning statement more legible? Explain.

Unquestionably. Barely one-third of consumers (34%) and only two in five drinkers (40%) think that health warning labels are clearly legible or readable under ordinary conditions all of the time. Nearly half the drinkers (whether they usually notice the warning or not) say the labels are not clearly legible or readable all of the time. Only one in five drinkers think that the warning appears with a contrasting background that makes it stand out and easy to read all the time. Three in four drinkers say the warning doesn't appear with a contrasting background all of the time.

As the poll reflects, consumers identify numerous other problems with current labels -- for example, their lack of prominence and small type -- and suggest rational means of correcting those deficiencies. In addition, they question whether the current government requirements regarding visibility and readability reflect a serious government intent that consumers notice the warning message. Consumers know that there is a problem and they know what to do to fix it. They question the seriousness of government standards for health warning labels and strongly believe that BATF would require the labels to be more visible and readable if it really wanted to carry out the intent of Congress to inform the American public about serious risks related to alcohol consumption. Americans believe that more people would be aware of those risks if the warning statement were routinely noticeable and readable.

9. What would be the costs associated with adopting any or all of the changes recommended by the petitioners, to the industry and, ultimately, the consumer?

Although we do not have the capacity to calculate the exact cost to industry, we are convinced that the costs can easily be borne by industry and will result in no cost to consumers.  In fact, if an improved warning on the dangers of drinking during pregnancy were to prevent the birth of just one person from a life with fetal alcohol syndrome, the human benefit would be incalculable and society would be spared lifetime costs of care for that person of at least $1.4 million(1). If a noticeable warning label helped avoid just one fatality in a highway crash, both the economic and human benefit would outweigh the one-time cost to most producers of improving the label. In addition, the label allows government, without cost to itself, and with minimal cost to industry, the opportunity to provide consumers with important health and safety information about unhealthy and unsafe alcohol consumption, one of this country's most widespread and costly public health problems. Certainly, an equivalent national media campaign or other sufficient education effort to alert consumers to the risks of alcohol consumption would be far more costly than the overall one-time expense of preparing a new, more prominent and readable label statement. Naturally, these are not alternative prevention approaches; warning statements on alcoholic-beverage labels should be only one component of a broader prevention campaign.

In assessing costs to industry, BATF must also recognize that alcohol producers routinely develop new labels and alter or vary the design of their product labels, whenever, apparently, it suits a marketing purpose. For example, beer companies print commemorative labels to celebrate college football teams (likely incurring development and printing costs for a relatively small run of labels). They develop labels for seasonal or new products (Oktoberfest - Becks) and for special events and markets; to create a "new identity;" even to celebrate the Millennium (Budweiser). The cost of redesigning a label never seems to be an issue, even though the production run is far less than what would be required were BATF to require improved warning messages. The costs of changing a label to comply with a new warning requirement would be spread over far more production and a longer period of time.

Likewise, BATF should look to the experience of other agencies that have calculated the costs of product label changes. One recent example concerns relatively new Food and Drug Administration (FDA) requirements for nutrition labeling of trans fatty acids in food(2). FDA based its estimate of the cost of changing the nutrition information panel on certain food products on the expectation of a three-color change and a 2-year compliance period. It concluded that some 38,670 product labels would be changed (for foods such as cereals, breads, crackers, cookies, chip type snacks, etc.), at a total relabeling cost of about $30 million. That cost comes to less than $1,000 per label. The per-unit cost of label changes for alcoholic beverages must be similarly insignificant.


Improvements in the label could not come at a better time. After years of decline, drinking by women during pregnancy has increased dramatically since 1991, in particular, frequent drinking(3). According to the Substance Abuse and Mental Health Services Administration of the Department of Health and Human Services(4), among women aged 15 to 44 who were pregnant in 1999 (when the survey was conducted), 13 percent reported drinking alcohol in the past month, with 3 percent reporting "binge" alcohol use (five or more on the same occasion at least once in the past 30 days). That translates into 316,000 pregnant monthly drinkers and 80,000 pregnant women who were binge drinkers. SAMHSA estimates that 6,000 pregnant women engaged in heavy drinking, consuming five or more drinks of alcohol on the same occasion on each of five days within the past 30.

Although the prevalence of Fetal Alcohol Syndrome (FAS) is not known, estimates range from 3 to 22 cases per 10,000 births. Between 1,300 and 8,000 children are born with FAS and many more are born with alcohol-related neurodevelopmental disorder (ARND)(5). The estimated health care costs of FAS totaled $2.8 billion in 1998(6).

According to the survey CSPI conducted, only 44 percent of women could recall (unaided) the warning statement that says women should not drink during pregnancy, and barely half (54%) could identify pregnant women as a person the warning label specifically advises not to drink alcoholic beverages at all. Effective warning labels may not deter all pregnant women from drinking, but they would certainly help improve the chances that alcohol-consuming women of child-bearing age will be more aware of the risks than they now are.

Based on the experience of other public health campaigns, in particular the effort to reduce cigarette smoking, evidence shows that well-designed warning messages can be part of a broader, comprehensive prevention effort. CSPI's November 1999 petition referenced substantial research that demonstrates that well-designed warning messages can help to increase awareness of health and safety risks.

Americans agree that the current warning messages on alcoholic-beverage containers are not to be taken seriously. They also agree on changes that can be made to make them more effective in increasing awareness of the health and safety risks related to alcohol consumption. Furthermore, dozens of public health, medical, safety, and addiction specialists agree that the warning messages should be improved to make them more noticeable, conspicuous, prominent, legible, and readable(7). Researchers who study label information also believe that the standards for warning statements need to be strengthened(8). Congress never intended the search for a warning label on alcoholic beverages to be an Easter egg hunt.

Based on a detailed survey of a representative sample of Americans, health warning statements on alcoholic beverages fail miserably at attracting consumer notice and effectively communicating important information about serious risks of alcohol consumption. Americans overwhelmingly agree that improving the physical presentational aspects of the label -- its prominence, placement, size, conspicuousness, legibility, color, and design -- would make the warning labels more noticeable and readable. Research findings previously submitted also strongly indicate that conspicuous and clear warning messages will better convey information to consumers. Therefore, we strongly recommend that the Bureau of Alcohol, Tobacco, and Firearms recognize its responsibility to Congress to require clear and conspicuous, noticeable, and legible health warning labels on alcohol beverage containers. We strongly support the development of new rules to improve the physical appearance of health warning statements on alcoholic-beverage labels.


1. Centers for Disease Control (1998) Preventing secondary conditions in children with fetal alcohol syndrome. World Wide Web document http:// (accessed on 3/17/99).

2. 21 CFR Part 101, [Docket No. 94P-0036], RIN 0910-AB66, basic formula for cost estimation, Table 13. -- Number of Information Panels Changed and Cost of Reprinting.

3. Alcohol Consumption by Pregnant Women in the United States During 1988 - 1995, Ebrahim,, Journal of Obstetrics & Gynecology, August 1998.

4. The NHSDA (National Household Survey on Drug Abuse) Report, Office of Applied Studies, Substance Abuse and Mental Health Services Administration, July 20, 2001.

5. Fetal Alcohol Syndrome, The National Center on Birth Defects and Developmental Disabilities, Centers for Disease Control, World Wide Web posting at (accessed 8/13/01).

6. 10th Special Report to Congress on Alcohol and Health, U.S. Department of Health and Human Services, June 2000, page 368.

7. A statement signed by more than 50 health officials and experts, including former Surgeon General C. Everett Koop, is attached to these comments as Exhibit 2.

8. A statement signed by two prominent human factors researchers who have studied alcohol warning labels extensively is attached to these comments as Exhibit 3.

For more information on alcohol warning labels, please click here.