|Home | About | Journals | Submit | Contact Us | Français|
The average US adolescent is exposed to 34 references to alcohol in popular music daily. Although brand recognition is an independent, potent risk factor for alcohol outcomes among adolescents, alcohol brand appearances in popular music have not been systematically assessed. We aimed to determine the prevalence of and contextual elements associated with alcohol brand appearances in U.S. popular music.
Qualitative content analysis.
We used Billboard Magazine to identify songs to which US adolescents were most exposed in 2005-2007. For each of the 793 songs, two trained coders independently analyzed the lyrics of each song for references to alcohol and alcohol brand appearances. Subsequent in-depth assessments utilised Atlas.ti to determine contextual factors associated with each of the alcohol brand appearances.
Our final code book contained 27 relevant codes representing 6 categories: alcohol types, consequences, emotional states, activities, status, and objects.
Average inter-rater reliability was high (κ=0.80), and all differences were easily adjudicated. Of the 793 songs in our sample, 169 (21.3%) explicitly referred to alcohol, and of those, 41 (24.3%) contained an alcohol brand appearance. Consequences associated with alcohol were more often positive than negative (41.5% vs. 17.1%, P<.001). Alcohol brand appearances were commonly associated with wealth (63.4%), sex (58.5%), luxury objects (51.2%), partying (48.8%), other drugs (43.9%), and vehicles (39.0%).
One-in-five songs sampled from U.S. popular music had explicit references to alcohol, and one quarter of these mentioned a specific alcohol brand. These alcohol brand appearances are commonly associated with a luxury lifestyle characterised by wealth, sex, partying, and other drugs.
Alcohol consumption is the leading root cause of mortality and serious morbidity in adolescence (1-3). About 70% of deaths in this age group are due to four major causes: motor vehicle accidents, other accidents, homicide, and suicide (1, 3), and it is estimated that these deaths are related to alcohol as much as 40-50% of the time (2-4). Alcohol is also considered the leading cause of morbidity in this population, due to its established association with nonfatal injuries (3, 4), other substance use (3, 5), risky sexual behavior (1, 6), academic failure (3, 7), physical and sexual assault (4), and alcohol dependence (8).
Despite the negative impact of alcohol use on adolescents, consumption remains high in the U.S. among this population: 45% of youth in grades 9-12 are current drinkers, defined as having a complete alcoholic drink during the past 30 days (1). Furthermore, over one-fourth (26%) of these adolescents are current “binge” drinkers—defined as having had 4-5 alcoholic beverages in a single sitting at least once in the past 30 days (1)—and nearly one-third of adolescents (29%) have, within the past 30 days, ridden in a car driven by someone who has used alcohol (1, 3). These early exposures are particularly concerning in light of the fact that the odds of future alcohol abuse or dependence increase by 7% for each year of age below 21 that alcohol consumption begins (8).
Alcohol use is associated with multiple sociodemographic (9, 10), environmental (10-13), and personal factors (10, 12-14). However, a growing body of literature suggests that exposure to certain mass media representations of alcohol—both narrative (e.g., movies) and persuasive (e.g., advertisements)—may be among the strongest risk factors for adolescent alcohol use (15-27). These media exposures are common; 83% of contemporary films (including 57% of G/PG films) depict alcohol use, exposing the average US youth 10-14 years of age to 5.6 hours of movie alcohol use and 244 alcohol brand appearances annually (28). Additionally, data from 2009 suggest that, during their formative years, U.S. youth view on average 366 alcohol advertisements on television annually (29).
Little research, however, has focused on alcohol representations in popular music, which has emerged as the most frequent alcohol-related media exposure (30). A previous assessment found an average 13.7 instances of alcohol use per song-hour (31), with wide variation by genre. American adolescents spend an average of 2.5 hours per day listening to music (32); depending on what musical genre they prefer, exposure could be quite substantial. Moreover, prior research shows that these references are commonly associated with consequences that adolescents find particularly compelling, such as sex, popularity, and partying (31). Thus, music could serve to link alcohol with these outcomes and enhance positive expectancies for alcohol use.
Prior research has not, however, examined alcohol brand appearances in popular music. Brand appearances are important to assess because they may function as advertising, whether or not they are paid for or sanctioned by the alcohol industry. Developing brand recognition is a crucial step in the marketing of any product, and companies go to great lengths to create positive associations with their brands through product placement activities. Moreover, for both cigarettes and alcohol, brand recognition and having a favorite brand are independent, potent risk factors for the initiation and maintenance of the use of these substances among adolescents (22, 33, 34). Therefore, the purpose of this study was to assess the prevalence and contextual elements of alcohol brand appearances in popular songs in the U.S.
We used Billboard Magazine (35) to identify the most popular songs in the US of 2005-2007, the most recent complete available data when the study was begun. Billboard annually uses an algorithm that integrates data from both sales and airplay to determine the top songs according to exposure. Sales data for this algorithm are compiled by Nielsen SoundScan from merchants representing over 90% of the US music market, including sales from music stores, direct-to-consumer transactions, and Internet sales and downloads. Billboard’s airplay data utilise Nielsen Broadcast Data Systems, which electronically monitors radio stations in more than 120 representative markets across the United States. Integrating these data, Billboard reported the following youth-relevant lists of popular song titles for each of the years from 2005-2007: the “Pop 100” (N=100) the “Billboard Hot 100” (N=100), “Hot Country Tracks” (N=60), “Hot R&B/Hip Hop Songs” (N=100), “Hot Rap Tracks” (N=25), “Mainstream Rock Tracks” (N=40), and “Modern Rock Tracks” (N=40). The lists are “closed-out” at year-end, after which time the song rankings do not change. Some songs were included on more than one chart, leaving 793 unique songs in this sample. Additional popular charts, such as the “Adult Top 40,” were also available but were not included in this analysis focusing on youth exposure to popular music.
For each of the 793 songs, two trained coders familiar with popular music independently analyzed the lyrics of each song for references to alcohol and alcohol brand appearances. As part of a formal training, each coder was given lists of frequently used slang terms related to alcohol use. We computed inter-rater agreement and kappa statistics (36) for each of the data elements coded and found excellent agreement between coders on each of these measures, such as the specific number of times alcohol was mentioned in a song. Minimum agreement was 74% and the few rare inter-rater disagreements were easily adjudicated between the coders (31).
The process described above resulted in 169 of 793 (21.3%) of songs that explicitly referred to alcohol. Of these songs with references to alcohol, 41 (24.3%) of songs specifically contained alcohol brand appearances. Two qualitative coders (E.N. and K.R.) then used Atlas.ti version 5.2 (37) to assess these particular songs in more depth. We selected a qualitative approach for this phase of the research in order to obtain a more in-depth understanding of the contextual associations with alcohol brand appearances.
Selected texts included any paragraph (e.g., verse or chorus) in which an alcohol brand appearance was found. We selected these specific paragraphs (rather than the entire song) in order to focus on elements specifically associated with the alcohol brand appearances. We selected the entire verse and/or chorus in order to maximise consistency.
The coders initially assessed the first 20% of the sample (N = 9 songs) using the in vivo coding feature of Atlas.ti. This feature allows coders to select any relevant text during a tentative coding process, allowing theory to emerge from the data (38). This method of building rather than testing theory is consistent with our selected analytic framework of grounded theory (please see analytic considerations below) (38). Based on these initial assessments and on a discussion process between the two coders and the principal investigator (B.P.), we developed a final study code book, condensing similar or redundant codes and using a system of open codes that incorporated new themes as they emerged from the text. The two coders then independently coded data from all 41 songs using the final codebook. We assessed the inter-rater reliability of our coding using Cohen’s kappa scores and adjudicated any discrepancies using an iterative process.
Our final code book contained 27 relevant codes representing 6 different categories: alcohol types, consequences, emotional states, activities, status, and objects (Table 2).
Nine alcohol-type codes identified the specific type of alcohol (‘NBBEER’, ‘NBRUM’, ‘NBWINE’, etc.) and whether that brand was a “luxury” brand (‘NBLUX’). For the purposes of this analysis, we defined a “luxury” brand as one which costs on average more than twice as much as a generic brand of the same alcohol type, as determined from a routine Internet search.
A pair of codes were used to indicate either positive (‘ALCPOS’) or negative (‘ALCNEG’) consequences of the alcohol use as suggested by song lyrics. When consequences were not discussed or neutral, no code was used.
Two codes were used to define emotional states. ‘FEEL GOOD’ was indicated when the subject expressed a positive emotional state. ‘COPING’ indicated the use of alcohol to modify mood, in an attempt to deal with an emotional issue.
Another series of codes were related to activities found juxtaposed with alcohol brand appearances. ‘SEX’ indicated a reference to sex or a sexualised act, while ‘DEGSEX,’ defined as degrading sex, was applied to any ‘SEX’ code in which both of the following criteria were satisfied: (1) the sexual act had no emotional component (i.e., it was completely physical) and (2) a power differential was conspicuous between the sexual partners.(39-41) ‘DANCE’ described an act of dancing or a dance move and ‘PARTY’ described the act of socializing with the goal of mutual enjoyment. ‘COMMUNITY’ indicated a social activity that invoked an intimate sense of belonging or community. ‘CRIME’ referred to criminal or illegal activity, while ‘DEAL’ specifically referred to the criminal act of buying or selling drugs.
Codes related to status were ‘WEALTH’, ‘ELITE’ and ‘STREET’. ‘WEALTH’ coded references, often to money or luxury items, that indicated substantial financial resources. ‘ELITE’ indicated an individual was socially elite, popular or desirable. ‘STREET’ referenced an individual’s origins in, or lifestyle associated with, “the streets,” which often represented humble beginnings.
Finally, several codes related to objects mentioned in conjunction with brand-name alcohol. ‘WEAPON’ coded a reference to a weapon, ‘OTHERDRUG’ indicated a drug or drug use that was not alcohol-related, but which excluded references to dealing, ‘VEHIC’ indicated a car or a reference to one of its components, such as rims, and ‘LUX’ was used to identify luxury items other than alcohol. Again, we used a price point more than double the generic to define a “luxury” item.
The average kappa score for each of the above variables was 0.80, which Landis and Koch (42) describe as a “nearly perfect” level of inter-rater reliability. Every discrepancy between the two coders was discussed among the team and easily adjudicated, resulting in a final data set.
As discussed above, we used grounded theory to analyze our data. The process of grounded theory analysis is systematic and moves from “basic description” (in vivo coding, described above) to “conceptual ordering” described as “organizing data into discrete categories according to their properties and dimensions and then using description to elucidate those categories (38).” We deemed this to be the most appropriate approach because of (1) the relative lack of prior theoretical work in this area and (2) our goal of obtaining an in-depth assessment of possible textual meanings.
Sample size in qualitative studies is guided by thematic saturation, a process by which a researcher collects and analyzes data until he or she is not learning anything new (43, 44). Although we first analyzed only songs from 2005 (31), at that time thematic saturation was not sufficient due to relatively few examples of brand names related to alcohol. Only after two additional years of data collection (2006 and 2007) was saturation deemed to be sufficient.
The frequencies of the codes were assessed using a quasi-statistical qualitative methodology (45). We summed the number of counts for each code and used Stata Statistical Software (46) to assess statistical significance of differences in code counts using chi-square analyses. We defined statistical significance with a two-tailed alpha of 0.05 a priori.
Of the 793 songs in our sample, 169 (21.3%) explicitly referred to alcohol, and of those, 41 (24.3%) contained an alcohol brand appearance. Over three-fourths (78.1%) of alcohol brand appearances involved luxury brands, and 51.2% of songs contained references to other luxury items as well. Tequila (Patron) and cognac (mostly Hennessy) were the most frequently mentioned types of branded alcohol, at a prevalence of 29.3% each, with vodka (mostly Grey Goose) following at 26.8% (Table 1). Of the songs with alcohol and an alcohol brand appearance, the majority of songs were Rap (63%), R&B/Hip-Hop (24%), or country (12%). No pop or rock songs mentioned name brands.
Associations between presence of a brand name and other song characteristics are presented in Table 2. While 17.1% of alcohol brand appearances were associated with negative consequences, 41.5% were associated with positive consequences (P<.001). We also determined if there were associations between presence of a brand name and types of consequences such as mental, physical, or social consequences. These analyses demonstrated that, compared with songs mentioning alcohol but without a brand name, presence of a brand name was associated with more positive social (P=.03) and sexual (P=.04) consequences but more negative legal consequences (P<.001). Mental, emotional, physical, and monetary consequences were no different in these different types of songs.
Qualitatively coded examples are presented in Table 3. One in 8 (12.2%) brand-name alcohol references were made in the context of a positive emotional state not the direct consequence of alcohol. Only two out of 41 songs (4.9%) referenced branded alcohol as a means of coping with negative emotional issues.
Alcohol brand appearances were commonly associated with sex (58.5%) – the majority of which was classified as degrading (75.2%) – and partying (48.8%). Alcohol brand appearances were also mentioned in songs that referenced dancing (31.7%) and crime (22.0%) or drug dealing (12.2%). A small portion of songs (4.9%) associated with an alcohol brand appearance invoked a sense of community or intimate social belonging, as seen in the lines, “Beer and bonfires / Wide open throttle, Coors in a bottle / It’s all for one and one for all y’all.”
Most songs were associated with wealth (63.4%) and over a third of songs (36.6%) made reference to the singer’s origins or lifestyle being associated with “the streets.” A substantial proportion of songs (43.9%) also make reference to the singer’s status as part of the social elite, as exemplified by the lyrics, “I’m on that Patron so get like me / … Er’body love me I’m so fly.”
Weapons (26.8%), other drugs (43.9%) and luxury objects (51.2%), other than luxury brand alcohol, appeared often in songs containing alcohol brand appearances. Vehicles (39.0%) were also commonly mentioned in these songs.
This qualitative assessment of popular music finds that alcohol is frequently referenced in popular music, and that when it is referenced a brand name is supplied about 25% of the time. This represents about 3.4 alcohol brand appearances per song-hour (31). Given that the average adolescent is exposed to about 2.5 hours of popular music per day (32), he or she will be substantially exposed to alcohol brand references in popular music on an annual basis. Furthermore, specific brand names mentioned are frequently luxury distilled spirits brands and are associated with a lifestyle characterised by degrading sexual activity, wealth, partying, and violence. Branded alcohol references were most common in Rap and R&B/Hip-Hop.
Frequent exposure of young people to brand-name references in popular music is of concern. Brand recognition is an independent and potent risk factor for the initiation and maintenance of substance use among adolescents (22, 33, 34), and references to alcohol brand-names in popular music may be particularly potent because they combine the persuasive appeal of advertising with the narrative appeal of an embedded and naturalistic representation of alcohol. This study suggests that brand-name references to alcohol are strongly associated with positive feelings and associations, which are often the goal of advertisements. Moreover, the brands found in music represent the same distilled spirits brands that are increasingly named as favorites by underage drinkers, especially females (47).
The relatively high level of brand-name appearances related to alcohol may be because of strengthening ties between the alcohol and music industries (Figure 1). An early example of alcohol advertising in the music industry involved the campaign launched by St. Ides Malt Liquor in the early 1990s, featuring TV and radio spots that showcased major rap artists, including 2Pac, Ice Cube, and Dr. Dre. While advertisers continue to engage in such traditional advertising ventures, many companies have turned to less traditional methods, including sponsoring music tours or hosting events in clubs (48). Alcohol companies have also more formally entered the music industry; from 1995 to 2001 alcohol industry giant Seagram’s acquired Universal and Polygram Records, merging the two labels to form the world’s largest music conglomerate. Though Seagram’s sold the label in 2001, the two companies maintained the synergistic marketing relationship they had established while jointly owned, continuing to sponsor music tours and use individual artists in their promotions. Finally, individual artists, particularly those in the rap and hip hop community, have begun to establish and promote their own alcohol lines, including Lil’ Jon (Little Jonathan Wineries, 2008), Ludacris (Conjure Vodka, 2009), Jay-Z (Armadale Vodka, 2002), Snoop Dogg (Landy Cognac, 2008), TI (Remy Martin Cognac, 2010) and Sean “P Diddy” Combs (Ciroc Vodka, 2001) (49-53). The fact that these artists primarily represent Rap and R&B/Hip-Hop labels may explain the preponderance of branded alcohol references in these two genres (Figure 1).
Most instances of brand-name references in song lyrics seem to be unsolicited and unpaid for by advertising companies. However, the line between paid advertising and brand references as an artistic choice has become difficult to distinguish, as advertising companies have begun to retroactively reward artists with product, sponsorship, or endorsement deals after a song containing their product’s name becomes popular. An example of retroactive paid product placement is Busta Rhymes and P. Diddy’s 2002 hit “Pass the Courvoisier.” Following the song’s release, the cognac’s sales jumped 18.9%, and Courvoisier’s parent company, France’s Allied Domecq, reached a lucrative promotional deal with Busta and P. Diddy’s management company, Violator (54).
Although rap and hip-hop music are popular among youth of all races (32), these genres still seem to be most heavily preferred by African-Americans in the U.S. For example, it is estimated that about 30-40% of those who purchase rap music are African-American (55), while African-Americans comprise only about 13% of the U.S. population (56). Our finding that alcohol brand references are most prominent in Rap and R&B/Hip-Hop (Table 2) suggests, therefore, that African-Americans are heavily exposed to these potentially influential messages. This exposure may be a potentially changeable source of the substantial alcohol-related health disparities between African-Americans and Caucasians (57, 58). However, it is not entirely clear whether African-Americans are indeed exposed to increased rap and/or hip-hop music compared with Caucasians, in part because Nielsen data have not systematically examined racial makeup of popular music audiences (55). Therefore, it would be valuable for future work to confirm whether African-Americans are disproportionately exposed to alcohol brand references and to examine the potential influence of these messages.
Compared with unbranded references to alcohol, branded references were more commonly associated with positive social consequences, positive sexual consequences, and negative legal consequences. However these associations are likely related to the fact that branded references were commonly associated with Rap and R&B/Hip-Hop genres, which commonly portray idealised social and sexual situations as well as confrontation with police (31).
Organizations of alcohol producers, such as The Distilled Spirits Council of the United States (DISCUS), have developed self-regulating codes regarding appropriate marketing practices (59). One of these guidelines states that it is unethical to target marketing to audiences below legal drinking age (59). However, Rap music is usually found to be a favorite type of music among high school students (60, 61). Therefore, it is probable that campaigns which center upon Rap artists are not in compliance with these guidelines. It may be valuable for future analyses to specifically address whether marketing campaigns associated with popular music, such as Rap and Hip-Hop, are consistent with marketing guidelines established by the industry.
One limitation of any qualitative methodology is that interpretation and bias introduced by the researcher poses a threat to internal consistency (43). We sought to minimize this by using a team of two researchers to independently code the data, comparing their responses with reliability scores, and involving a third researcher in all discussions related to codebook development. Another limitation of this analysis is that we only assessed songs from 2005-2007. Because of recent announcements of newly-forming relationships between the music and alcohol industries (49-53), it may be valuable to revisit these analyses in upcoming years. Finally, we coded songs deemed to be the most popular among US adolescents. Thus, although music popular in the US is likely to be roughly similar to that in other parts of the world, these findings may not be generalisable to those in other nations.
This study demonstrates when alcohol is referenced in popular music—which occurs in about 20% of songs—a brand name is supplied about 25% of the time. Furthermore, we found that brand-name references were commonly associated with a lifestyle often characterised by degrading sexual activity, wealth, partying, and violence. Because both brand recognition and positively-portrayed substance use are powerful predictors of adolescent substance use, it may be important to assess the associations between brand-name alcohol references in popular music with clinically-relevant alcohol use outcomes.
When alcohol is referenced in U.S. popular music, a brand name is supplied about 25% of the time. Based on these estimates and published data, the average U.S. adolescent is heavily exposed to branded alcohol in popular music. Of songs with alcohol and an alcohol brand appearance, the vast majority are categorised as Rap (~60%) or R&B/Hip-Hop (~25%). Brand-name references to alcohol are commonly associated with a lifestyle characterised by degrading sexual activity, wealth, partying, and violence.
DECLARATION OF INTEREST
This work was supported by the National Cancer Institute (K07-CA114315 to B.P.). Although they provided financial support, they were not involved in the design and conduct of the study; the collection, management, analysis, and interpretation of the data; or the preparation, review, or approval of the manuscript. The authors have no connection with and do not receive any funding from the tobacco, alcohol, pharmaceutical, or gaming industries.
Brian A. Primack, Departments of Medicine and Pediatrics, University of Pittsburgh School of Medicine, Pittsburgh, PA.
Erin Nuzzo, University of Pittsburgh School of Medicine, Pittsburgh, PA.
Kristen R. Rice, Department of Medicine, University of Pittsburgh School of Medicine, Pittsburgh, PA.
James D. Sargent, Department of Pediatrics, Dartmouth University School of Medicine, Hanover, NH.