Old Wine In New Bottles: A Three-Part Evaluation of the “Above the Influence” Campaign - Morgan Coe

1. Critique

Although it frames itself as a general resource for young people coping with peer pressure and stress,1 the Office of National Drug Control Policy’s “Above the Influence” (AtI) campaign appears to be aimed primarily at teens who are using or at risk for trying marijuana. While AtI’s print materials, advertising spots, and website design clearly attempt to reflect the visual, musical, and conversational language of its target demographic, it would appear that these efforts are largely focused on surface elements and fail to take advantage of behavioral theories that would make their content as audience-appropriate as their style. I will examine the AtI campaign’s 20 video spots2 through the lens of three such theories: Social Network Theory, Labeling Theory, and Marketing Theory.

Social Network Theory
Because adolescent life often focuses strongly on social group membership, the Social Network Theory (SNT) perspective is a useful lens for examining a youth-oriented advertising campaign like AtI. SNT defines a “social network” as a cluster of one-to-one connections between individuals. These connections are based on family, friendship, employment, or geography (to name some of the more common ties) and can operate in one or two directions.3 SNT research has found that individuals’ choices are strongly influenced by the people they are connected to, and that tightly-clustered social networks tend to make changes at the same time. Conversely, individuals who do not belong to such tight networks or who are socially situated on their periphery, are less likely to go along with such group-wide changes.4, 5
Whether they are reinforcing abstinence or urging marijuana users to quit, with very few exceptions AtI’s video spots do not frame such decisions as a social process; rather they rely implicitly on an individual-level model of behavior change. Furthermore, they generally feature individuals whose drug-related decisions have alienated them from peers, or are in the process of doing so. For instance, in both “Anne Marie” and “I Feel Bad” we see marijuana users who are pathetically unable to convince peers to smoke with them. Conversely, “Pony” and “Pete’s Couch” depict young people whose choices not to use marijuana are causing them to pull away from their peer group. Although these four ads vary considerably in style and tone, their messages follow one of two formats: on the one hand, marijuana users are marginalized by their predominantly non-using peers, while on the other hand it is non-users who choose to distance themselves from their marijuana-using peers. While both of these scenarios superficially capture one feature of social networks (namely, that individuals who choose differently from the networks they belong to tend to be found on the edges of those networks), they do so with causality inverted. According to SNT research and modeling these individuals are not marginalized from their networks because they have chosen differently, but quite the opposite – they have chosen differently because they were marginally placed in their networks to begin with.4, 5 The distinction is a crucial one, because it means that the behaviors being modeled in these spots (where non-users push users to the margins of their peer group, or actively remove themselves from user-dominated peer groups) are unlikely to take place. In other words, while the message might resonate with youth who are already at lower risk (non-users in predominantly non-using peer groups, or non-users who are only peripherally attached to user-dominated peer groups), they are unlikely to have much direct or indirect effect on those who are at higher risk according to SNT (non-users with significant ties to predominantly using peer networks).
However, a small minority of AtI’s spots are more in line with SNT. “Little Brother” (with its tagline of “Big brothers who live above the influence have little brothers who live above the influence”) illustrates the fact that sibling relationships are a component of social networks, and that they are likely to transmit influence from older sibling to younger. However, research suggests that these relationships are less significant than friendships, which are also likely to be more numerous.4, 5 On the other hand, “Whatever” posits a non-user who appears to be an influential member of a user-dominated social network – he drives three friends around while narrating his role in dictating their collective social activities (they do not speak, and their faces are not shown), and finally states that he intends to attend college and leave them behind. Since higher educational status gives someone more influence over fellow social network members while at the same time making him or her less likely to be influenced by them in turn, and since geographic distance does not greatly reduce the influence of close social network ties,4, 5 it is possible that an individual like the narrator of “Whatever” could help to tip the balance of his social network towards non-use; however the spot does not explore or attempt to encourage that outcome. Ultimately, “Little Brother” and especially “Whatever” are promising from a SNT standpoint but fall short in execution: while they effectively illustrate social network truisms, they take a largely passive approach and do not use these truisms to develop an active approach to influencing at-risk youth.

Labeling Theory
Because their identity and self-image are in a state of flux, adolescents tend to be extremely preoccupied with labels, either self-defined or applied by others. Labeling Theory (LT) provides us with some important insights into the process and effects of labeling, especially as it applies to mental illness and “deviant” behavior. LT sees “deviant” as a label that is applied sometimes arbitrarily to those who are perceived (rightly or wrongly) to have behaved in certain ways. The theory is not generally concerned with the reasons for an individual’s initial act of deviance, but rather with the ways that the “deviant” label subsequently changes their behavior and supersedes other identity categories.6, 7, 8 Since AtI portrays marijuana use as a deviant behavior that is often tied to mental illness and even suicidal ideation,9 the LT perspective is especially relevant here.
The labels applied by AtI’s video spots generally fall along one of two lines, either attributing negative characteristics to teen marijuana users directly, or negatively portraying non-users’ reactions to them. Since these approaches function differently under LT, we will consider them separately. “Fire,” “The Cocoon,” “Huggin’ The Block,” “S.L.O.M.,” “Pete’s Couch,” and “Whatever” all fall into the first category, depicting users as passive, unable to achieve, pathetic, and in one case physically repulsive. Although milder than might be the case with other forms of deviance, these images are analogous to LT’s “degradation ceremony,” after which an individual labeled as deviant begins to adopt a deviant lifestyle and incorporate the labels assigned her into her self-image.8 In addition to encouraging this kind of negative change by directly labeling marijuana users, these spots are likely to contribute indirectly to the same users’ being similarly labeled by peers, family, and society at large because the expectations and stereotypes that lead to labeling are often influenced by media depictions of the groups being labeled.7
On the other hand “Sent,” “The Conversation,” “Not Again,” “Try Football,” “Dog,” “Stop Looking At Me,” “I Feel Bad,” and “Walk Yourself” go beyond labeling marijuana-using teens (although they do this as well, in a similar manner to the spots mentioned above) to focus on their relationship with peers. The main themes here are rejection, disappointment, and humiliation at the hands of friends, romantic partners, and pets. According to LT, individuals who have been labeled or fear being labeled as deviant are significantly affected not only by their own experience of being labeled, but the ways they expect other people to react to their being labeled. Because of this they may withdraw from contact, perform poorly in challenging situations, or engage in negative behaviors as a preemptive defense.10 In other words, while a few of the other AtI spots do encourage non-users to reach out to their at-risk peers (“Mirror,” “Little Brother”), many are instead encouraging further isolation and withdrawal.
The concept of self-fulfilling prophecy is critical in LT 6, 8 and should raise serious concerns about a campaign that relies so heavily on negative labeling. While AtI nominally acknowledges that many teens use drugs in order to cope with stress and social anxiety,9 it labels them as social misfits who lack coping skills, thereby perpetuating the cycle. While it nominally urges non-using teens to reach out to their peers at risk,11 it builds expectations that will sabotage such outreach efforts with isolation and defensiveness. At best, this may result in a campaign whose mixed messages render it ineffective; at worst the harm done by AtI’s labeling may outweigh whatever good it accomplishes through other channels. This internal incoherence could have been avoided by applying principles of LT to the campaign’s design from the start in order to avoid undermining its good intentions with negative labeling, and perhaps even to use positive labeling to reinforce its message by fostering communication, self-efficacy, and coping skills.

Marketing Theory
As adolescence in the United States has become more commercialized and adolescent consumers have become prime targets for advertisers, it is not surprising that anti-drug campaigns like AtI have drawn more and more influence from the world of marketing. Marketing Theory (MT) is a broad and multi-faceted discipline and while many of its principles are specific to commercial enterprises, others can be applied just as well to organizations that are “selling” ideas or behaviors. For instance the process of symbolization, whereby the marketer creates value for their product by making it more culturally and symbolically attractive to potential consumers, could be used to encourage drug abstinence as well as to sell cigarettes.12 However, MT recognizes that a product’s symbolic appeal is a product of several layers of influence: a culture or sub-culture’s climate of valuation can affect consumers’ relationship towards a general type of product, while individual marketers’ efforts at branding can encourage consumers to choose one particular product over another. Since it is easier to successfully brand a product than to change the way a culture values it, branding has become one of the foundations of modern marketing.13
Clearly, AtI’s video spots were designed with an eye towards establishing “Above the Influence” and its circled arrow logo as a recognizable brand – nearly every ad prominently displays this phrase and logo, and all of them feature at least one of the two. Furthermore, five of the animated ads (“Not Again,” “Stop Looking At Me,” “I Feel Bad,” “Walk Yourself,” “Try Football”) are dominated by a shared visual and musical aesthetic that makes their common branding even more distinctive. MT’s familiarity principle states that even in the absence of effective advertising messages consumers will tend to choose a familiar product over an unfamiliar one. While this principle is sometimes employed in situations where many similar products are competing with few substantive ways to distinguish them,14 this is not the case with AtI – their ads should therefore go beyond simply establishing a presence, and must actively work to build a brand that holds value for their target demographic.
Unfortunately, most of the emotions and situations that AtI has associated with their brand are unlikely to build positive value for their viewers: feelings of guilt, regret, and shame; situations in which teens disappoint others, or are judged and rejected by friends and romantic partners. Although these negatives could be said to “make sense” because they are presented as consequences of marijuana use, MT research has found that while viewers sometimes respond cognitively to the rational content of an advertisement, they are just as likely to be primarily affected by its emotional content.15 Regardless of how valid these ads’ rational context may be, they will function as negative branding for AtI simply because the situations and emotions they depict are overwhelmingly negative. In other words with a few exceptions (“T Shirts,” “Little Brother,” and “Whatever” maintain a generally positive tone, and associate their brand with generally desirable situations and outcomes), these ads will tend to cause their viewers to associate the “Above the Influence” brand with guilt, regret, shame, abandonment, lost opportunity – precisely the negative values that they are meant to see as consequences of marijuana use.
MT states that in order to market a product effectively, it is critical that marketers be able to measure the success of their efforts.12 In the absence of such (admittedly difficult to implement in many cases) measures, there is a danger that they will substitute other more accessible ones such as “cleverness” or the aesthetic approval of their peers – while these are appealing, they do not necessarily contribute to the effectiveness of a campaign and may even hinder the process of building and communicating brand and product value to consumers.14 AtI’s video spots bear signs of having fallen into this trap, because in spite of their often clever and well-crafted dialogue, soundtracks, and visual imagery they are incoherent at best and self-sabotaging at worst. Most troublingly, they consistently expose their viewers to the “Above the Influence” brand and signature imagery in juxtaposition with the negative values that they want their viewers to associate not with abstinence but with marijuana use. MT suggests that if AtI’s intent is to create a negative brand around marijuana use they should employ a different strategy, perhaps creating a “fake” marijuana-related brand to use in place of their current “real” one. On the other hand, if these videos are meant to build up the “Above the Influence” brand, their primary role is not to be clever or amusing but to connect that brand with the promise of “implied experiences” that will be attractive to at-risk youth. According to MT, “[i]f this aspect of the function of advertising is recognized, much of the arrogant and sanctimonious tone in some advertising can be relieved and a positive program of distinctive image development put in its place.” 13

Conclusion
While a great deal of effort has clearly been put into making AtI look and sound appealing to its youthful target demographic, it appears that no similar attempt has been made to ensure that its underlying message will come across as intended. The result is superficially polished and “hip,” but substantively similar to the generations of largely ineffective anti-substance campaigns that have preceded it (“Just say no”,16 “This is your brain on drugs”17). There is no reason to expect that contemporizing this outdated model will make it work, and some theories suggest the opposite: that novelty and gloss often take the place of clear and effective persuasion.14 Rather than seeking new ways to convey old messages, AtI would do well to explore new models on which to base new messages that may be far more effective in reaching at-risk youth. Social Network Theory, Labeling Theory, and Marketing Theory are three such models.

2. Recommendations

We have seen that AtI’s approach to teen marijuana use is likely to be ineffective at best, and harmful at worst. Furthermore, it would appear that the program’s ads were designed with little or no regard for recent theories of decision-making – or perhaps for any such theories at all. This leaves us with a (nearly) blank slate upon which to use contemporary socio-behavioral theories to more effectively formulate and communicate an anti-drug message.

Social Network Theory
From the perspective of Social Network Theory (SNT), AtI’s most significant flaw is that it views the decision making process solely as an individual one, ignoring the reciprocal relationships between members of a social network. As discussed above, this results in ads that predominantly depict non-using youth distancing themselves from their marijuana-smoking peers, or marijuana-smoking youth rejected by their non-using peers. To the extent that they are simply descriptive, these ads are essentially tautological (because they depict the end result of processes that occur naturally between individuals in social networks); to the extent that they expect their viewers to emulate the examples they model, they are almost certainly doomed to failure (because the processes they depict are not fundamentally based on conscious individual action of the kind that would be influenced by AtI’s literal messages).4, 5
In many respects, SNT is more suited to observation and description than to direct intervention. However, it does suggest several ways that AtI’s ad campaign could be improved. Most obviously, the ads could be reworked to target whole social groups instead of individuals. For instance, “Pete’s Couch” currently focuses on a single narrator who is presented as articulate, irreverent (his first spoken line is “I smoked weed, and [dramatic pause] nobody died”), and ultimately willing to venture, drug-free, into the world outside the Couch. His friends, including the titular “Pete,” do not speak or move, are never shown in close-up, and are eventually abandoned to their marijuana-using ways. Clearly viewers are meant to identify with the narrator, and emulate his journey from wry drug use to newfound sobriety. However, SNT would predict that only viewers who were marginally attached to drug-using social networks to begin with (and therefore already less likely to be influenced by their peers’ use) were likely to respond to this ad. Although these viewers should not be ignored, they represent low-hanging fruit compared to the couch-dwelling Petes of our target population. Redesigning “Pete’s Couch” to work as a population-level intervention would mean altering its basic storyline: instead of telling the story of one marijuana user who rebelled against his faceless friends’ apathy to embrace an exciting and sober lifestyle, it would show an entire peer group making the same change together (as SNT says is most often the case).
In addition to group-level interventions, SNT also leaves room for interventions aimed at individuals based on the places they occupy in their social networks – older siblings, more educated peers, and centrally placed individuals are likely to exert a powerful influence on those around them,4, 5 and could help to tilt the balance in their social networks. Although it was not done from a strictly SNT perspective, some research has shown the effectiveness of targeting such “opinion leaders” to spread public health messages.18, 19 In fact, two of AtI’s ads do just this in a rudimentary (and perhaps unintentional) way: “Little Brother” and “Whatever.” Although both of these ads depict individuals who would very likely be opinion leaders within their social networks (an older brother and a socially dominant non-user, respectively), neither one reaches significantly beyond the descriptive level.
In “Little Brother” a final voiceover does remind us that “big brothers who live above the influence have little brothers who live above the influence,” but this earnest truism feels like an afterthought, totally disconnected from the ad’s naturalistic camera work and dialogue. According to SNT this message is the very core of the ad, and should be illustrated at least as vividly and engagingly as the interactions leading up to it. Similarly, “Whatever” fails to resolve the situation it has established, instead falling back on a story arc similar to the one we saw in “Pete’s Couch”: once the non-using protagonist has gone to college, the marijuana-smoking friends he leaves behind will be “somebody else’s problem.” It would be far more constructive for this ad to emphasize the positive changes that such an opinion leader could have on his social network, for instance by showing one or more of the narrator’s friends following his example (as SNT predicts might be the case). In both cases our intent would be to move beyond an essentially neutral depiction of the opinion leader in order to present a more active image that will empower such leaders to continue exerting a positive influence on those around them, while at the same time remaining faithful to what can be reasonably expected under SNT.

Labeling Theory
Labeling Theory (LT) would tend to see many of AtI’s ads as extremely problematic, because a majority of them label marijuana-using teens in ways that are likely to alienate them from non-using peers and family, make them even less able to cope with the stress and/or mental health issues that may have pushed them to use drugs in the first place, and ultimately create a self-fulfilling prophecy that pushes them further into a “deviant” lifestyle.6, 7, 8 The challenge for LT would be to rework this labeling process in order to make it more conducive to positive outcomes, while at the same time taking care not to be perceived as positively labeling marijuana use itself.
Several of the AtI ads directly apply a variety of negative labels to young drug users. In fact, some of them appear to be primarily intended to foster feelings of shame, regret, and revulsion towards marijuana and those who use it – “Fire,” “The Cocoon,” “Huggin’ The Block,” and “S.L.O.M.” fall into this category. These ads would be difficult to re-work along LT lines without rebuilding them from the ground up. For instance, “Fire” shows three teens symbolically burning things they once valued (car, guitar, athletic trophies) while wistful music plays in the background – although the connection is not depicted literally, a voiceover solemnly reminds us that “marijuana costs you more than you think.” Similarly, in “Cocoon” a young boy gradually seals himself into a literal cocoon made of marijuana leaves and rolling papers; when he emerges under the caption “what you choose today affects who you are tomorrow,” he has become a pathetic middle-aged man who still lives with his parents. LT suggests that we should re-work ads like these to invert them, almost like running a movie reel in reverse – instead of threatening that marijuana will take away users’ hobbies, interests, and youth (thereby labeling users as having forfeited all of those things), we should show users reclaiming those very things as they give up marijuana (thereby labeling them as being capable of making such a change).
A second type of AtI ad focuses on marijuana-using teens’ relationship to their non-using peers. The dominant themes are rejection, disappointment, and alienation, and again LT would lead us to worry that they could create a self-fulfilling prophecy. For instance, in “Dog” the family pet expresses his concern over a girl’s marijuana use while she listens in silence, her body language guilty and evasive. He then goes outside, leaving her alone and visibly sad in a final close-up. We are presumably meant to perceive her as being deep in thought, perhaps steeling herself for a difficult decision as a result of the dog’s “tough love” (in the parlance of many an anti-drug intervention), but according to LT these perceptions are secondary to the labels that are being applied to their relationship here: abandonment and disappointment. Not only are these labels likely to discourage viewers from connecting with the metaphorical “dogs” who might try to reach out to them, it is quite likely that they could feed into the very same fears and anxieties that push them to use marijuana in the first place. In order to rework this ad along more constructive lines we would have to invert the relationship dynamic and emotional tone it depicts, to show a dog who accepts his owner without judgment in her time of need, and an owner who is able to make a positive choice to reconnect with her pet instead of using drugs. In other words, we would label users as being self-efficient and capable of making positive life choices, while simultaneously labeling their peers as accepting, non-judgmental, and able to support them in making those same choices.

Marketing Theory
The concept of “branding” is central to Marketing Theory (MT), and gets to the core of one of the significant weaknesses of the AtI campaign. AtI has gone through the motions of establishing an “Above the Influence” brand through their logo, visual design, and even music choices but instead of making their brand culturally and symbolically attractive to potential “consumers” of marijuana abstinence, they have consistently connected it with negative values such as guilt, shame, rejection, revulsion, and loneliness – the very negative values they would like their viewers to associate with drug use. In other words, while AtI’s design choices are set up to build a positive brand around their abstinence program, their content appears to be aimed at establishing a negative brand around marijuana use. This kind of negative branding has been used effectively against the tobacco industry (by the Truth campaign, for instance 20) but would be more problematic here in the absence of a well-defined “marijuana industry” upon which to hang the negative brand. In addition, there is a very real risk that explicitly marijuana-focused ads – no matter how negative – might be perceived as glamorizing drug use, and indeed might function as such due to design flaws or chance fluctuations in the pop-cultural context that surrounds them. From an MT perspective, it would be more advisable for AtI to abandon its attempts at negatively branding drug use in order to re-focus on establishing a positive brand around their own program.
In order to make their program (and by extension, drug abstinence) more attractive to at-risk teens, AtI would first need to conduct in-person qualitative research in order to ascertain what sorts of symbols or “implied experiences” hold a positive value for them. Such research is unfortunately beyond the scope of this discussion; however the Truth campaign’s findings might be of some relevance here. Their researchers found that teens smoked cigarettes in order to “signal that they were in control” and to “[send] a signal to the world that the user made decisions for themselves.” 20 While the act of smoking cigarettes has a dimension of public performance that is absent (or substantially altered) when we look at marijuana use, it is reasonable to suppose that it may touch on some of the same deeper motivations. Accordingly, we would need to reverse the symbolism present in most of AtI’s ads: instead of associating the “Above the Influence” brand with images of youth who have given up control of their life to their peers or to the drugs they use, we should associate it with symbols of control and independence.
Several of AtI’s ads could be re-worked in order to operate more effectively as positive branding. For instance, both “Pete’s Couch,” “Whatever,” and “Pony” feature protagonists who embody the values above – unfortunately, these values are presented as a zero-sum game in which the protagonists’ independence can only be asserted at the expense of a group of non-independent peers. From an MT perspective the overall effect is, unfortunately, negative. Altering this dynamic could be done in two ways: either by removing the protagonists’ peers from the equation and presenting the situation as one in which a single independent individual asserts self-control against marijuana, or by incorporating the peers into the positive branding, depicting not just one but several teens who get off the couch, go to college, and give up drugs. Among AtI’s ads only “T-Shirts” (in which an initially apathetic teen sheds layer after layer of bulky clothing, representing his gradual disenchantment with marijuana, and eventually goes outside into the sunshine) takes the former approach; there are no examples of the latter.

Conclusion
It is striking that Social Network Theory, Labeling Theory, and Marketing Theory often suggest that nearly identical changes be made to AtI’s ads. For instance, where AtI depicts one individual leaving her drug-using peers behind, SNT, LT, and MT would have us substitute a peer group who all exert their independence against drugs – SNT because this scenario better reflects reality, LT because it avoids applying harmful labels to the protagonist’s peers, and MT because it puts a much stronger emphasis on the kinds of positive values that should be attached to the “Above the Influence” brand. For the same reasons, where AtI portrays non-users as judgmental and inclined to reject users, SNT, LT, and MT would portray them as support sources who remain connected to those who need them. In other words, while these three theories were developed in different settings, at different times, and for different purposes, there is tremendous potential for applying them in synergy here.
REFERENCES
1. “Above the Influence” [http://www.abovetheinfluence.com/about.aspx] 11/17/08
2. “Above the Influence” [http://www.abovetheinfluence.com/the-ads/default.aspx?path=nav] 11/17/08
3. Barnes, J. (1954) “Class and Committees in a Norwegian Island Parish” Human Relations; 7:39-58.
4. Christakis, N. and Fowler, J. (2007) “The Spread of Obesity in a Large Social Network Over 32 Years” New England Journal of Medicine; 357:370-9.
5. Christakis, N. and Fowler, J. (2008)“The Collective Dynamics of Smoking in a Large Social Network” New England Journal of Medicine; 358:2249-2258.
6. Becker, H. (1963) Outsiders: Studies in the Sociology of Deviance New York: The Free Press.
7. Link B., Cullen F., Struening E., Shrout P., Dohrenwend B. (1989)“A Modified Labeling Theory Approach to Mental Disorders: An Empirical Assessment” American Sociological Review; 54(3):400-423.
8. Gove, W. (1970 ) “Societal Reaction As An Explanation Of Mental Illness: An Evaluation” American Sociological Review; 35(5):873-84.
9. “Above the Influence” [http://www.abovetheinfluence.com/help/ask-the-doc/is-it-really-bad-to-use-drugs.aspx] 11/18/08
10. Link, B. (1987) “Understanding Labeling Effects in the Area of Mental Disorders: An Assessment of the Effects of Expectations of Rejection” American Sociological Review; 52(1):96-112.
11. “Above the Influence” [http://www.abovetheinfluence.com/help/help-a-friend.aspx, http://www.abovetheinfluence.com/help/help-a-friend-2.aspx] 11/18/08
12. Kotler, P. (1972) “A Generic Concept of Marketing” The Journal of Marketing; 36(2):46-54.
13. White, I. (1959) “The Functions of Advertising in Our Culture” The Journal of Marketing; 24(1):8-14.
14. Politz, A. (1960) “The Dilemma of Creative Advertising” The Journal of Marketing; 25(2)1-6.
15. Vakratsas, D. and Ambler, T. (1999) “How Advertising Works: What Do We Really Know?” The Journal of Marketing; 63(1):26-43.
16. Ronald Reagan Presidential Foundation & Library “Mrs. Reagan’s Crusade” [http://www.reaganfoundation.org/reagan/nancy/just_say_no.asp] 11/19/08
17. Partnership for a Drug-Free America “The Partnership’s ‘Fried Egg’ TV Message” [http://www.drugfree.org/Portal/About/NewsReleases/Fried_Egg_Message] 11/19/08
18. Kelly J., St Lawrence J., Diaz Y., et al. (1991) “HIV Risk Behavior Reduction Following Intervention With Key Opinion Leaders Of Population: An Experimental Analysis” American Journal of Public Health; 81:168-71.
19. Kelly J., St Lawrence J., Stevenson L., et al. (1992) “Community AIDS/HIV Risk Reduction: The Effects Of Endorsements By Popular People In Three Cities” American Journal of Public Health; 82:1483-9.
20. Truth [http://www.thetruth.com] 12/10/08