PAPER CONTRIBUTION K
The Need for, and Value of, a Multi-Level Approach to Disease Prevention: The Case of Tobacco Control
NATURE AND MAGNITUDE OF THE HEALTH CONSEQUENCES OF TOBACCO CONSUMPTION 2
In a prominent article published in 1993, McGinnis and Foege (1993) compared the leading disease causes of death with “the actual causes of death.” Accounting for 1.1 million annual deaths among Americans, half of all mortality, these “actual causes” —unhealthy behaviors and environmental exposures—represent the principal challenge to public health in this new century. Topping the list was tobacco consumption, the source of nearly a fifth of all deaths among Americans.
Almost single-handedly, smoking has transformed lung cancer from a virtually unknown disease at the turn of the twentieth century to the leading cause of
Dr. Warner is Richard D.Remington Collegiate Professor of Public Health and director, University of Michigan Tobacco Research Network, Department of Health Management and Policy, School of Public Health, University of Michigan. This paper was prepared for the symposium “Capitalizing on Social Science and Behavioral Research to Improve the Public's Health,” the Institute of Medicine and the Commission on Behavioral and Social Sciences and Education of the National Research Council, Atlanta, Georgia, February 2–3, 2000.
Throughout this paper, phrases such as “cigarette smoking” and “tobacco consumption” are used reasonably interchangeably. Unless the context dictates otherwise, the reader should interpret references to smoking as applying generally to other forms of tobacco consumption as well. (An example of a context that precludes generalization is the discussion of laws that prohibit smoking in public places. Clearly this would not apply to use of smokeless tobacco as well.) Cigarette smoking is referred to specifically in many instances primarily because it is far and away the most important source of tobacco-produced disease.
cancer death at its conclusion. Including mortality associated with environmental tobacco smoke and with interactions with other exposures (especially radon), smoking is responsible for more than 90% of the lung cancer deaths that fell Americans each year. Smoking is also the leading cause of chronic obstructive pulmonary disease mortality, accounting for at least 85% of deaths attributable to emphysema and chronic bronchitis. Although smoking accounts for only 17 to 30% of cardiovascular disease deaths, the dominance of heart disease as the leading disease cause of death accords this illness its dubious status as a contender, with lung cancer, as the chief tobacco-produced source of mortality. In addition to the “big three,” smoking contributes to a host of less common causes of death, and it creates an enormous burden of preventable morbidity and disability as well (U.S. DHHS, 1989; Napier, 1996).
The enormity of the toll of smoking is due to a combination of the widespread prevalence of this behavior (more than 45 million Americans smoke), its intensity (smokers take 10–12 puffs per cigarette on an average of more than 25 cigarettes per day), and the chemical composition of the smoke inhaled (more than 4,000 chemicals, including tobacco-specific nitrosamines, ammonia, for-maldehyde, naphthalene, carbon monoxide, hydrogen cyanide, arsenic, benzo[a]pyrene, and polonium-210; 43 of the chemicals are known human carcinogens (U.S. DHHS, 1989). Over a typical smoking “career” of 50 years, a lifelong smoker inhales these 4,000 chemicals 4 million to 5 million times. Given such exposure, it is perhaps nothing short of remarkable that an estimated one-half of lifelong smokers do not die as a consequence of the behavior. Indeed, there may be no more impressive testimony to the strength of the human organism.
The fact remains, however, that reflecting its burden on the other half of the smoking population, this chemical onslaught takes a toll unparalleled in the course of human history. In the United States, the toll will subside over time, a reflection of a gradual and continuing decline in smoking rates dating from the 1960s. The prevalence of smoking has dropped from approximately 45% in 1963, the year prior to publication of the first Surgeon General's report on smoking and health (U.S. DHEW, 1964), to 25% in 1997 (CDC, 1999b). Among men, prevalence has halved. Despite a rising population, total cigarette consumption in the United States has fallen from 633 billion cigarettes in 1981 to 479 billion in 1997. Adult per capita cigarette consumption —a common measure of consumption that adjusts for population growth —has fallen almost annually since 1973 (Tobacco Institute, 1998). Based on projections of the demographics of smoking, even in the absence of stronger tobacco control education and policy than exist at present, and assuming no change in youth initiation of smoking, prevalence should continue to fall in the United States over the next two decades or so, “bottoming out” at around 18% of adults (Mendez and Warner, 1998).
Even at these diminished levels of smoking, however, the toll of smoking will remain substantial throughout at least the first half of the new century. Still, the achievements of America's “antismoking campaign,” dating from publication of the Surgeon General's report, rank tobacco control among, and perhaps at the
top of, the major public health success stories of the second half of the twentieth century (CDC, 1999a).3 The consumption declines described above suggest the magnitude of the shift in smoking behavior since the early 1960s, but in fact they considerably understate the extent of the accomplishments. In the absence of the then-new knowledge about the dangers of smoking, and the publicizing of this knowledge that constituted the heart of the early antismoking campaign, smoking prevalence almost certainly would have continued to climb, reflecting the rapid rise of smoking rates among women. It is certainly plausible, even probable, that total smoking prevalence would have exceeded 50% by the end of the 1960s or early 1970s. Adult per capita cigarette consumption would have exceeded 6,000 per year, compared to 4,345 in 1963—the highest level ever attained in the United States—and 2,333, today's figure. According to one analysis, the first two decades of the antismoking campaign should be credited with avoiding nearly 3 million premature deaths through the year 2000 (Warner, 1989).
Reflecting on America's tobacco control victories, juxtaposed against the enormity of the continuing problem, one can consider America's situation as either a cup half full or a cup half empty. The picture for the rest of the world is far bleaker. The World Health Organization predicts that today's global smoking population of 1.1 billion will mushroom to 1.6 billion by the year 2030. By then, tobacco will have become the leading cause of death in developing countries as it is today in developed countries. Currently the cause of 4 million deaths worldwide, tobacco will kill 10 million of the globe's citizens annually beginning near the end of the third decade of the twenty-first century (World Bank, 1999). No other behavioral, environmental, or biological cause of death will come close (Murray and Lopez, 1996).
The magnitude of America's continuing tobacco-produced death toll, combined with the success of the nation's multifaceted smoking control endeavors, make tobacco control an excellent candidate for examining why a multilevel approach to disease intervention is essential. Lessons from the tobacco story certainly have relevance to other health behavior dilemmas within the United States, and they generalize to other countries' efforts to come to grips with their own emerging and existing epidemics of tobacco-produced disease.
DEVELOPMENT OF SMOKING AS A NORMATIVE BEHAVIOR
While American public health leaders like to fantasize about a “tobacco-free” country, the prospects for eliminating tobacco from the McGinnis-Foege list in the foreseeable future verge on the nonexistent. The fascinating history of tobacco use, eloquently related by Goodman (1994) and others (Wagner, 1971; U.S. DHHS, 1992), demonstrates the vise-like grip that tobacco, and its principal dependency-forming constituent, nicotine, have long held on the members of
The nature of the campaign is discussed later in this paper.
all societies exposed to “the golden leaf.” The earliest recorded evidence of tobacco use in the Americas dates from the ninth century, and in the intervening 1,000-plus years, people have developed an extraordinary array of methods of consuming tobacco, including as a suppository for medicinal purposes. In India at present, tobacco is consumed in about a dozen distinct forms by millions of people (Bhonsle et al., 1992). Although the principal early purposes of tobacco use were medicinal and religious, history reveals evidence of social use, multiple times per day, in several North American tribes in the early years of the second millennium.
Tobacco use spread rapidly throughout Europe and Asia beginning in the sixteenth century, when explorers of “the New World” first brought it back. So potent was the hold of nicotine that smokers defied official prohibitions of its use. Most notably, in the late sixteenth century, Sultan Murad IV of Turkey declared smoking punishable by beheading or being drawn and quartered; yet thousands of Turks persisted in inhaling the intoxicating fumes and, in many cases, suffering the consequences. This experience serves not merely as an interesting historical footnote. It demonstrates that official tobacco control policies have existed for centuries and that tobacco smoking has persisted in the face of far more draconian penalties than any contemplated today.4
Suspicions about the dangers of smoking have existed for at least four centuries. In 1604, writing in a document entitled Counterblaste to Tobacco, King James I of England called smoking “a custom loathsome to the eye, hateful to the nose, harmful to the brain, dangerous to the lungs and in the black stinking fume thereof, nearest resembling the horrible Stygian smoke of the pit that is bottomless” (as quoted in Wagner, 1971, p. 11).
Despite this early insight, true scientific understanding of the hazards is a strictly twentieth century phenomenon, dating primarily from the second half of the century (U.S. DHHS, 1989). In large part, this reflects the fact that smoking became a widespread threat to health beginning only in the second decade of the twentieth century, with the refinement of the easy-to-inhale cigarette. Prior to 1913, the harsh tobaccos used in cigarettes made cigarette smoking a minor form of tobacco consumption. Cigars, pipes, snuff, and chewing tobacco dominated the tobacco market. Rarely inhaled deeply and frequently, these forms of tobacco consumption, although hazardous to health, posed only a minor risk compared to that which would become associated with cigarette smoking.
In 1913, Camel cigarettes introduced the “American blend” of tobaccos, a combination of flavorful tobaccos imported from Turkey and Egypt with milder American tobaccos that permitted deep inhalation for the first time. Camels also introduced what is widely regarded as the first modern advertising campaign. The pairing of flavor and ease of inhalation with creative advertising has been credited with inaugurating the modern era of the cigarette (Tilley, 1985). Other
One might credit Murad IV not only as the first government official to develop a tobacco control policy, but also, through its enforcement, as the first person to prove that smoking was hazardous to health!
factors contributed to the rapid emergence of cigarette smoking as well, however. They included the relatively low cost of the new cigarettes, the addictiveness of easily inhaled nicotine, and the convenience of packaged cigarettes: in an increasingly harried daily life, tobacco users appreciated the ease and brevity of the cigarette smoking experience. The latter constituted a principal reason that cigarettes were included in soldiers' rations in World War I and in every war thereafter until the Gulf War.
Previously considered effeminate, cigarette smoking was converted into a normative behavior among males by the return to America of tens of thousands of newly addicted soldiers. Through effective marketing, the cigarette manufacturers managed to associate smoking with athleticism and romance. A veritable who's who of baseball stars, such as Lou Gehrig and Joe DiMaggio, advertised cigarettes widely. And smoking became de rigeur in seduction scenes in the movies, epitomized by the cigarette dangling from Humphrey Bogart's lips whenever he was sweet-talking an attractive lady in one of his films. In the 1950s, America's most respected man, Edward R.Murrow, chain-smoked his way through his television news and commentary broadcasts5. When the cohort of men born in the 1911–1920 decade reached their age of peak smoking prevalence—their late 20s to mid-30s—fully 70% smoked cigarettes. As recently as the early 1960s, more than half of adult males smoked (Warner, 1986).
In contrast, at no time did a majority of women smoke. Smoking by women was considered completely socially unacceptable during the first two to three decades of the century, and “daring” thereafter until World War II opened the “man's world” to women (e.g., with so many men overseas fighting, women began working in factories in large numbers). The cigarette industry exploited the image of smoking as risqué in multiple ways. In early cigarette ads, women urged their male companions who were smoking to “blow some my way.” In a precursor to the “liberated woman” advertising campaign for Virginia Slims in the late 1960s, ad agencies staged marches through downtown New York with defiant women smoking. Through such techniques, the agencies tried to link smoking by women to the suffragette movement.
By the 1940s, women were beginning to smoke in large numbers. Indeed, it is striking to note that four 10-year birth cohorts of women, those born from 1901 to 1940, all reached their rates of peak smoking prevalence within the single five-year period 1958–1963. For the oldest of the four cohorts, born during the century's first decade, this meant that they peaked at an average age of 52.5 (Warner, 1986). (By comparison, women born from 1951 to 1960 achieved their peak prevalence in 1976, at an average age of 20.5.) From the 1930s through the early 1960s, the diffusion of smoking among women was paralleling that experienced among men approximately three decades earlier. Unlike men, however,
As a sad footnote, both Bogart and Murrow died of lung cancer in their 50s, along with a plethora of other prominent smokers from the era.
TABLE 1. Smoking Prevalence (%) by Gender and Racial or Ethnic Group
Race or Ethnicity
American Indian or Alaska Native
Asian or Pacific Islander
SOURCE: CDC, 1999b.
women's smoking prevalence peaked at about a third. The growth of smoking among women was interrupted by the advent of the national antismoking campaign, inaugurated in January 1964 with publication of the first Surgeon General's report (Warner and Murt, 1982).
The antismoking campaign has never been a single “campaign” in the conventional sense. Rather, it has consisted of an unorchestrated mix of varied private and public sector efforts, first to educate the public about the hazards of smoking and, subsequently, to protect nonsmokers from exposure to environmental tobacco smoke. The major component parts of the campaign, and their effects, constitute the focal point of this paper. As such, discussion of them is deferred to the last section. However, as one indication of the overall impact of the campaign, consider the subtle shift in the norms surrounding the act of smoking. In the middle of the century it was considered impolite to light a cigarette without offering one to one's companions. Beginning in the 1970s, however, the question asked by a smoker contemplating lighting up switched from, Would you like a cigarette? To do you mind if I smoke? By the 1990s, in many social circles, the latter question had been mooted by the expectation that the answer would be yes. Today, many smokers generally ask nothing and refrain from smoking in the presence of nonsmokers.
WHO SMOKES AND WHY
Of adult Americans, 24.7% were smokers in 1997 (CDC, 1999b). Although a greater percentage of men smoke than do women (27.6% and 22.1%, respectively), the gap between the two genders has declined gradually over time. Racial and ethnic differences in smoking prevalence are substantial, ranging from 16.9% for Asians and Pacific Islanders to twice as much, 34.1%, for American Indians and Alaskan Natives. Race or ethnicity and gender differences in smoking prevalence are presented in Table 1.
Race or ethnicity prevalence differences mask other differences in smoking behaviors that affect disease outcomes. For example, although more African-
American males smoke than do white males, African Americans smoke fewer cigarettes per day. Possibly mitigating the potential health advantage of lower daily consumption is African Americans' preference for mentholated cigarettes, believed to have an anesthetizing effect on the throat that may lead to deeper inhalation. Menthol may also contribute to a higher rate of addiction (Ramirez and Gallion, 1993).
Smoking rates vary substantially by age, with prevalence declining in the fourth and subsequent decades of life. In the older ages, differential death rates for smokers and nonsmokers account for a significant fraction of the prominent decrease in smoking prevalence (Harris, 1983). Smoking cessation, the principal determinant of the decline in prevalence with age, rises significantly with age. Cessation rates appear to have leveled off during the 1980s and early 1990s, with concern that they may actually have fallen in the late 1990s.
A real challenge to students of the demographics of smoking is assessing smoking initiation, a phenomenon that occurs almost exclusively during childhood. Much concern has been expressed about the documented increase in 30-day smoking prevalence among eighth, tenth, and twelfth graders during the first half of the 1990s, a trend that has, fortunately, reversed in the most recent years. In the 1999 Monitoring the Future Survey, 34.6% of high school seniors had smoked within the past 30 days. The comparable figures for tenth and eighth graders are 25.7 and 17.5% (Monitoring the Future, 1999). Troubling, however, is the question of how one should assess “smoking” by children, for while 30-day prevalence rates were rising during the 1990s, measures of regular and heavy smoking (e.g., half a pack or more per day) were not.
Data on youth smoking also raise perplexing questions about racial and ethnic differences. Most notably, the rates of smoking for African-American students were dramatically lower than those for whites, although the gap has narrowed in recent years. Yet smoking rates among young adult African Americans often exceed those of comparably aged whites. The difference is explained, in part, by lower quit rates among African Americans.
A large majority of children experiment with smoking, yet fewer than half go on to become regular smokers. Sociologists attribute much of the propensity to experiment with tobacco, as well as the propensity to become a regular smoker, to the influence of peer and parental behavior. It is widely believed, for example, that the children of smokers are twice as likely to smoke as the children of nonsmokers (although the data from multiple studies are not uniformly consistent in finding a significant association [U.S. DHHS, 1994]). There are important socioeconomic and educational links as well: in the United States, as in most developed nations today, smoking is increasingly becoming a marker for lower socioeconomic status (U.S. DHHS, 1989).
Clearly, the preeminent determinant of smoking dependency in a given individual is addiction to nicotine. A wealth of evidence from biology, brain chemistry, and sociology indicts nicotine as a classic addictive substance (U.S. DHHS, 1988; Henningfield et al., 1993). Ironically, when in 1988 the Surgeon General observed that nicotine was as addictive as heroin and cocaine, he may
have been understating the case in one important respect: of all the dependency-forming substances of abuse, nicotine likely addicts the greatest proportion of its users. But the question remains as to why some people can take it or leave it, while others find themselves incapable of renouncing its use. Intriguing clues are emerging from the rapidly developing field of genetic science. Recent research offers provocative evidence of a genetic explanation for as much as half of the propensity to become a smoker and half of the apparent inability of some smokers to quit (Pomerleau, 1995).
That the social context in which smoking occurs affects the amount and nature of smoking is evident in the literature on children's role modeling (Lynch and Bonnie, 1994; U.S. DHHS, 1994). Intriguing new evidence on contemporary patterns of adult smoking illustrates just how influential societal norms can be. A recent (1992) change in the questions asked by the National Health Interview Survey (NHIS) to determine smoking status permitted analysts to assess how many people smoke cigarettes on a nondaily basis. The conventional wisdom among experts on smoking had always been that only 5% or so of smokers were “chippers” —recreational smokers who could smoke only a few cigarettes per day or even smoke on a nondaily basis. The NHIS data indicate that close to a fifth of all smokers do not smoke every day. Although multiple possible explanations come to mind, there is widespread agreement that the “clean indoor air” movement, prohibiting smoking in many public places and workplaces, likely has redefined smoking for a subset of smokers. These smokers have learned how to “survive” in a smoking-hostile environment by restricting their smoking to locations, and days, in which it is acceptable.
This shift in the social environment in which smoking takes place is but one example of the determinants of smoking that lend themselves to purposeful collective intervention. Another, of great concern within the tobacco control community, relates to the marketing of cigarettes and other tobacco products. It is an article of faith within the tobacco control community that advertising and, increasingly, other forms of marketing (ranging from sports sponsorship to distribution of cigarette brand-related paraphernalia) seduce youngsters into experimenting with cigarettes and keep adults hooked who otherwise would quit. Although the weight of the evidence supports this view, there is no “smoking gun” that demonstrates it conclusively (U.S. DHHS, 1989). The empirical evidence on the issue is mixed (Chaloupka and Warner, in press).
Nevertheless, it is clear that the marketing of tobacco products is a major front in the war on smoking, and notable tobacco control victories have been realized within the past few years, most recently the result of the multistate settlement concluded between the state attorneys general and the tobacco industry (National Association of Attorneys General, 2000). As a consequence of that settlement, tobacco billboards have disappeared and human subjects and cartoon characters will no longer grace the pages of cigarette ads. Product sampling is banned, as is product placement in movies and other media. Brand names are prohibited on merchandise (e.g., clothing and caps). Team sports sponsorship is prohibited and brand-name sponsorship is limited to one per year. In addition,
the settlement provided for the establishment of a foundation (the American Legacy Foundation) that will devote substantial resources to marketing a strong antitobacco message to youth. Combined, the restrictions on tobacco marketing and the introduction of a well-funded counteradvertising campaign may have a significant influence on youth smoking by altering the social environment in which it has been glamorized so effectively for decades.
THE OBJECTIVES OF TOBACCO CONTROL AND BROAD STRATEGIES TO ATTAIN THEM
Tobacco control has three principal objectives, all directed toward avoiding the enormous burden of disease wrought by tobacco use. Although the three objectives are distinct in concept, realization of any one of them (partial or complete) will often affect one or both of the other objectives, as considered in the next section. The three objectives are the following:
preventing the initiation of tobacco use by young people,
helping adult smokers to quit (or at least to reduce their risk), and
protecting nonsmokers from the annoyance and risk posed by environmental tobacco smoke.
In principle, proponents of tobacco control should and do support all three objectives. In practice, however, a vigorous national debate has emerged, pitting advocates for a youth prevention focus (the first objective) against those who insist upon a more comprehensive strategy, one devoted more centrally to preservation of clean indoor air (the third objective) and more explicitly to helping adult smokers to quit (the second objective).
The youth prevention orientation has evolved in large part as a political strategy: preventing smoking by kids is a nearly universally lauded objective; even the tobacco industry pays lip service to this goal (Brown and Williamson, 2000; Philip Morris, 2000; R.J.Reynolds, 2000). This recognition drove the Food and Drug Administration (FDA) to adopt an exclusively youth-oriented set of policies when it courageously ventured into the realm of nicotine and tobacco product regulation (FDA, 1996). Similarly, state Medicaid lawsuits against the industry emphasized the prevention of youth smoking (Jacobson and Warner, 1999), and the multistate settlement's public health provisions all concentrated on that same goal (National Association of Attorneys General, 2000). In a congressional debate on policy measures designed to reduce smoking by kids, no representative or senator could afford to be depicted as opposing the legislation's purpose (although they could oppose its approaches, in part because they might impinge on the liberties of responsible adults).
Given these apple pie and motherhood characteristics, the youth prevention orientation pays homage to the art of the possible. In addition, from the perspective of many tobacco control advocates it permits a foot in the legislative door to achieve broader objectives by dressing them in prevention clothing. For exam-
ple, the call for a large tax increase to discourage youth smoking would, if adopted, also decrease smoking by adults. A media antismoking campaign would reach adults both directly and indirectly, the latter by adults' awareness of the changing attitudes and behaviors of young people with regard to smoking. It is certainly reasonable to call for smoke-free indoor environments in locations frequented by young people, such as schools, malls, and sporting events. Adults work and play in those settings as well. Banning tobacco billboards has an effect on its ostensible target, youth, and on adults too.
Proponents of a more comprehensive strategy perceive the youth prevention approach as devious and, ultimately, counterproductive. They are not so sanguine about the ability to achieve broader tobacco control objectives when policies and programs are constrained to explicitly target youth. Even if this would work at least partially, in their judgment it would necessarily represent an inefficient attack on issues such as adult quitting and protection of nonsmokers' rights to clean air. Confronted with the massive war chests of the tobacco industry, their top-quality legal and public relations talent, and the inherent allure of nicotine, tobacco control can ill afford less than the best-possible approaches to achieving the three objectives. Further, many proponents of a frankly adult-oriented comprehensive strategy believe that any campaign explicitly focused on discouraging children from smoking will have exactly the opposite effect: rebellious kids often take their cues as to what to do based on what the adult world tells them not to do (Glantz, 1996).
Proponents of the comprehensive strategy, which targets smoking by adults quite explicitly, risk the charge that their objective smacks of paternalism. Lacking the easy political constituency of the youth orientation, the comprehensive tobacco control strategy creates an instant opposition in the form of the individual liberty and responsibility lobby. This is especially troublesome at the level of federal policy, which the comprehensive approach tacticians denounce as hopeless. At the national level the tobacco industry has repeatedly demonstrated its unrivaled strength to mobilize the troops to fight adult-affecting policy measures in Congress. Federal cigarette excise tax increase proposals are labeled, successfully, as ruinous and unfair “tax-and-spend ” government “business as usual.” Restrictions on cigarette advertising are an affront to First Amendment protections of commercial speech. Clean indoor air laws assault Americans ' basic liberty rights, and the right and obligation to give “common courtesy” a chance to work first (Advocacy Institute, 1998).
Given the history of very timid tobacco control legislation in Congress, and the related ability of the tobacco industry to marshal its forces in successful opposition, the comprehensive tobacco control school emphasizes grass roots action, attacking tobacco at the local level (e.g., municipalities) and, when necessary, at the state level as well. A divide-and-conquer strategy—spreading the resources of the tobacco industry thinly across hundreds of local jurisdictions—has produced some impressive tobacco control victories, particularly in the domain of clean indoor air ordinances (Samuels and Glantz, 1991).
The division of the tobacco control forces into the two strategic camps reflects their terrain: the leadership of the youth prevention strategy comes from “establishment” national organizations, including the American Cancer Society, the American Heart Association, the American Medical Association, the National Center for Tobacco-Free Kids, and the new American Legacy Foundation. Savvy in the ways of Washington, replete with financial resources and congressional lobbying experience, these organizations believe that the greatest bang-for-the-buck ultimately must be achieved with a national strategy; for the reasons given, they believe that that strategy must necessarily be youth oriented.
In contrast, and not surprisingly, the comprehensive tobacco control school derives its support from state and local organizations, including local chapters of the major health voluntaries, tobacco control divisions of state and local units of government, and tobacco control grass roots organizations. The leadership of this school consists of individuals who have fought the tobacco wars in the trenches of state and local politics, directing state ballot initiatives to raise cigarette excise taxes and dedicate revenues to tobacco control, and heading campaigns to ban smoking in restaurants and other public places within cities and counties.
While the battle between the two schools of thought rages on, it is imperative to emphasize that virtually all tobacco control proponents concur that all three of the major objectives of tobacco control —youth smoking prevention, adult cessation, and protection of nonsmokers —are meritorious. The issues that divide the two camps are principally strategic, essentially pragmatic in nature. In truth, of course, as is invariably the case, they also reflect concerns about “turf (Advocacy Institute, 1999).
Yet another strategic division appears poised on the horizon to dominate many coming debates about tobacco control, centered in, but certain to erupt outside, the community of smoking cessation experts. These experts are pondering the possibility of finding effective harm reduction strategies that fall short of smokers' completely renouncing their dependence on nicotine. The interest in harm reduction derives from frustration with the slow pace of smoking cessation—only 3% of smokers quit each year—and the emergence of a plethora of new nicotine delivery technologies, produced by both the tobacco and the pharmaceutical industries. The notion is that many smokers who find themselves unable (or unwilling) to give up nicotine might find lower-risk nicotine delivery devices acceptable substitutes for conventional cigarettes (Warner et al., 1997; Warner, in press).
Fraught with perils, the harm reduction concept faces considerable opposition among public health professionals, many of whom adhere strictly to a just-say-no philosophy. To many experts, however, harm reduction is the crucial new frontier in America's ongoing battle against tobacco-produced disease. This complicated and fascinating debate is certain to capture the attention of much of the tobacco control community as technology evolves and appropriate policy responses are contemplated. It explains why the second general objective of tobacco control—helping adult smokers to quit—was modified by “or at least to
reduce their risk.” For the purposes of the present paper, however, with few interventions having been studied to date, harm reduction will not be considered further. Interested readers should consult the provocative literature emerging on the subject (Tobacco Dependence, 1998; Ferrence et al., in press).6
APPROACHES TO TOBACCO CONTROL: WHAT ARE THEY AND HOW WELL DO THEY WORK?7
One of the defining characteristics of tobacco control—regardless of whether or not one adopts a youth-oriented focus—is recognition that a successful assault on the disease burden created by smoking necessarily must be multidimensional (CDC, 2000; National Cancer Policy Board, 2000). Individual interventions do work on their own. For example, smoking cessation treatments do help a subset of smokers to quit. Tax increases clearly discourage children from smoking and reduce smoking by adults as well. Prohibitions on smoking in public places clean the air for nonsmokers, decrease smoking prevalence and daily consumption among ongoing smokers, and help to establish and reinforce a non-smoking social norm. Yet, alone, each of these interventions succeeds by tinkering on the fringe. Collectively, these and other interventions may exhibit powerful synergism. The smoker committed to participating in a cessation treatment, for example, stands a better chance of quitting if the external environment discourages smoking, as it does when smoking is prohibited in public places. That same smoker is more likely to succeed in quitting if the price of cigarettes rises due to a tax increase.
In the remainder of this paper, attention focuses on understanding the varied dimensions of tobacco control intervention, briefly describing knowledge of their impacts and how they interact. First, however, a digression concerning the title and purpose of the paper seems warranted to set the stage for this section's discussion. The purpose is to examine a “multilevel” approach to improving the public's health, in this case by decreasing the use of tobacco products. The term “multilevel” is ambiguous, yet usefully so. One can envision a sizable number of alternative constructions of “level, ” each of which would permit the development of a different set of insights about public health intervention. By way of summary, Box 1 lists several obvious candidates for the meaning of “level.”
As this paper is being written, an Institute of Medicine committee is embarking on a study of the science base for evaluating the harm reduction potential of new nicotine delivery technologies.
Given the brevity of this paper, this section can merely summarize important findings pertaining to this large subject. The most comprehensive review of tobacco control interventions, although now dated, is found in Chapters 6 and 7 of the 1989 Surgeon General's report (U.S. DHHS, 1989; see also U.S. DHHS, 1991). More recent references, discussing specific interventions, are cited in the text.
BOX 1. Alternative Constructions of “Level” in “Multilevel Approaches to Public Health”
For the purposes of this paper, attention focuses on the last conception of level in Box 1—the intervention function. The section opens with a brief discussion of an intervention function typology that is applied in subsequent subsections. The section then turns to the application of this typology to each of the three major objectives of tobacco control.
A Typology of Intervention Functions
All tobacco control (and indeed other public health) interventions can be classified as one of three broad categorical types of interventions, here organized in increasing order of coerciveness (i.e., the degree to which they force behavior change): education and information interventions, incentives, and laws and regulations. The first of these encompasses all activities designed to inform the public about hazards or benefits to health and/or to persuade people to take health-enhancing behavioral action. In the case of tobacco control, dissemination of the findings published in the Surgeon General's reports on smoking and health serves to educate the public about the dangers of smoking (and of passive smoking) or the health benefits of quitting. Media “counteradvertising ” campaigns attempt to persuade young people or adults to avoid tobacco use. Warning labels on cigarette packs and ads are intended to inform about dangers and, implicitly, to discourage use.
The second category, incentives, refers primarily to economic inducements to avoid tobacco. The most obvious and important example is an increase in a cigarette excise tax, driving up the price of cigarettes and thereby discouraging cigarette purchases by individuals who “feel the bite” of the higher price in their wallets. Other examples include differential life insurance rates (you pay more for given coverage if you smoke than if you do not) and explicit smoking cessation incentives, such as employers' rewarding workers who do not smoke with pay bonuses (Warner and Murt, 1984; U.S. DHHS, 1989).
The final category, laws and regulations, refers to explicit, legally binding requirements to do, or not do, something pertaining to tobacco consumption. Most notable here are clean indoor air laws, prohibiting smoking in public places, and minimum age of purchase laws that forbid vendors from selling cigarettes to minors and minors from buying them.
Walsh and Gordon (1986) described this method of classifying tobacco control interventions in an article written for the Annual Review of Public Health. Warner et al. (1990) developed a more general drug policy typology that elaborates on this classification, distinguishing the intervention type by point of intervention and end user. They observe that an intervention can represent a legal requirement of one party that serves to educate another. For example, cigarette and smokeless tobacco manufacturers are required to place warning labels on all packs of cigarettes and ads. This legal obligation of a member of the production and supply chain is intended to educate the end user, the purchaser of the tobacco product.
For the remainder of this section, intervention types will be judged by their relationship to the end user, the smoker or potential smoker. It is useful to recognize, nevertheless, that most education and incentive interventions, as viewed by consumers, represent legal or regulatory obligations imposed on either members of the production and supply chain or other “intermediaries” (e.g., schools required to present tobacco and health education to their students).
Preventing Initiation of Smoking8
Each of the major tobacco control objectives draws on interventions in all three categories of intervention types. Only in the case of preventing youth initiation of tobacco use, however, do interventions in all three categories receive roughly comparable emphasis. Education of children about the dangers of tobacco products, intended to discourage their use, was the hallmark of early youth prevention efforts (U.S. DHHS, 1994). This included both formal inschool education and media counteradvertising campaigns oriented toward children. Education and persuasion remain a core approach in youth prevention, as reflected in the new multimillion-dollar media campaign launched by the American Legacy Foundation in January 2000. Raising cigarette excise taxes,
Good general references on this subject are the IOM report Growing Up Tobacco Free (Lynch and Bonnie, 1994) and Lantz et al. (in press).
the principal intervention from the incentive category, has become a central feature of nearly all comprehensive efforts to discourage youth smoking (Grossman and Chaloupka, 1997; National Cancer Policy Board, 2000). Recently, a great deal of emphasis has been placed on a legal strategy: strictly enforcing laws prohibiting sales of tobacco products to minors, generally defined as under age 18 (Forster and Wolfson, 1998; DiFranza, 1999). Indeed, strict enforcement is one of the main provisions of the Food and Drug Administration's 1996 regulatory policy designed to reduce youth smoking (FDA, 1996).
Preventing youth initiation of tobacco use logically appears to be critical to the overriding objective of breaking the seemingly endless cycle of youthful experimentation, addiction, subsequent long-term use, disease, and death. It is therefore exceedingly frustrating for the tobacco control community, and for public health more generally, that so little is understood about how to intervene effectively to reduce youth tobacco use. As it is implemented in practice, school health education, the quintessential ingredient of youth tobacco control, has contributed little to discouraging future experimentation or addiction. There is evidence that state-of-the-art education programs can reduce tobacco use (U.S. DHHS, 1994). But the reality, thus far, is that school systems lack the resources and the will to implement state-of-the-art programs effectively and on a sustained basis.
More encouraging, at least in the short run, have been media counteradvertising campaigns. In California, Massachusetts, Arizona, and Florida, well-funded media campaigns have caught the attention of young people and been followed by youth smoking rates that either declined or failed to rise at the rate experienced nationally through much of the 1990s (Popham et al., 1994; Florida Dept. of Health, 1999; Gregory Connolly, Massachusetts Dept. of Health, personal communication, 1999). In particular, advocates of media campaigns cite the apparent recent success of the Florida campaign, funded by resources from that state 's settlement of its Medicaid lawsuit with the tobacco industry. Thirty-day smoking prevalence among teens fell from 23.3% to 20.9% from 1998 to 1999.
How long a media campaign can produce sustained decreases in youth tobacco use is unknown. Both logic and experience with cigarette advertising and counteradvertising suggest that as the novelty of an advertising or counteradvertising campaign recedes, so too will its effectiveness (Saffer and Chaloupka, 1999). Their novelty was likely one of the principal reasons that the broadcast media Fairness Doctrine antismoking ads of the late 1960s were so effective in discouraging smoking in the face of broadcast cigarette ads that outnumbered them by from 3:1 to 12:1 (Warner, 1979). Nevertheless, the apparent success of the state-funded campaigns has encouraged the American Legacy Foundation to devote a substantial proportion of its nearly $200 million in annual expenditures to a professionally designed media campaign.
The effect of price increases on discouraging youth smoking is clearly one of the best-documented and most encouraging stories in all of tobacco control. Although the evidence is not unequivocal, there is a consensus among economists who have studied the relationship that higher prices do decrease smoking
by young people (National Cancer Institute, 1993; Warner et al., 1995; Chaloupka and Warner, in press). While the overall, primarily adult demand for cigarettes is only modestly responsive to price changes—a 10% increase in price decreases cigarettes demanded by approximately 4%—analysts believe that smoking by kids is roughly twice as price responsive as that of adults. As such, since tax constitutes only a fraction of total cigarette price, even a modest increase in a tax rate can reduce youth smoking significantly. To achieve maximal impact, however, such tax increases have to be sustained in real value over time. Since most taxes are set in nominal amounts, inflation gradually erodes their consumption-discouraging impact, unless further tax increases keep pace with inflation (Grossman and Chaloupka, 1997).
In contrast with most of the literature, a few studies by economists have not identified greater price responsiveness among children than among adults (Chaloupka, 1991; Wasserman et al., 1991). And a recent debate, inaugurated by conflicting interpretations of the same empirical data set (DeCicca et al., 1998; Dee and Evans, 1998), has broken out among health economists as to whether higher prices discourage youth smoking initiation per se. Despite these uncertainties, the overall strength of the evidence supporting the proposition that higher prices reduce youth smoking led a group of economists who had substantial experience working on tobacco economic issues to call for increased cigarette taxation primarily to reduce youth smoking (Warner et al., 1995). 9
Interest in taxation as a means of combating youth tobacco use followed the early emphasis on education. The latest addition to the youth prevention armamentarium is policy oriented toward reducing youth access to tobacco products at retail outlets. A series of studies over several years established that minimum-age-of-purchase laws were enforced rarely; in monitored experiments, teenagers found it easy to buy cigarettes from stores in multiple states (Rigotti et al., 1997; Forster and Wolfson, 1998; Lancaster and Stead, 1999). Offended by vendors' utter disregard for the law and hopeful that enforcement might stem the tide of smoking by young people, a number of advocates pushed for legislative action. Federal and state laws were tightened, and most recently, the FDA's regulations required vendors to request photo identification from anyone appearing to be less than 27 years old (FDA, 1996). As a consequence, in some jurisdictions,
Recent work by economists has found variations in the responsiveness of youth and young adults to price changes associated with race, gender, and socioeconomic status (Centers for Disease Control and Prevention, 1998; Chaloupka and Pacula, 1999). This new work introduces as many questions as it answers, since it does not explain why these differences exist. Still, the attempt to differentiate the effectiveness of policy interventions by race or ethnicity and gender, as well as age and income, seems an important step in refining policy approaches to tobacco control. To date, the literature on sociodemographic differences in smoking is related primarily to smoking prevalence, rather than to responses to educational and policy interventions (U.S. DHHS, 1998), although there are important exceptions (Townsend et al., 1994). Clearly, more work is needed on the latter.
compliance with the law has increased considerably, although not all states are in conformance with federal mandates (DiFranza, 1999).
Whether stricter enforcement of youth access laws will make a substantial difference in youth tobacco use remains to be seen. A study of what is likely the gold standard of enforcement, in Woodbridge, Illinois, produced very encouraging results: merchant sales to minors fell from 70% to 5% over one and a half years of compliance checks, following implementation of a youth access ordinance, and experimentation with and regular use of cigarettes by adolescents fell by more than 50% (Jason et al., 1991). In another study, however, while retail outlets reduced their cigarette sales to minors significantly (compliance with the law in three intervention communities rose to 82%, compared to 45% in three control communities), analysts found no evidence of a decline in smoking by youth (Rigotti et al., 1997). A recent analysis demonstrates convincingly why anything short of near-complete decreases in illegal retail sales is unlikely to substantially impact youth smoking. Young people will learn where cigarettes can be purchased and they have multiple other means of acquiring cigarettes as well (e.g., older siblings, friends, and black markets on school grounds) (Levy et al., 1998).
A more novel approach to youth smoking consists of directly penalizing youth caught smoking (or in possession of cigarettes). Approaches vary across the states, including requiring offenders to attend youth “tobacco courts,” restricting law violators' driving privileges, and fining offenders. These demand-oriented approaches have drawn criticism from some members of the tobacco control community, who prefer to emphasize legal penalties against adult suppliers of cigarettes. To date, there is little evidence on the effectiveness of any of these new strategies (Lantz et al., 2000).
Analyses of the effects of comprehensive youth prevention programs are few in number and thin on empirical evidence. The state programs in California and Massachusetts consisted primarily of a tax increase and a tax revenue-funded media counteradvertising campaign. (Both states put additional resources into other activities, but these have received only modest funding and little if any analytical attention.) It is clear that these states' programs have reduced tobacco use by youth (as well as adults, as discussed in the following section) (National Cancer Policy Board, 2000). Given the paucity of evidence on the effects of media campaigns by themselves, however, it is impossible as yet to determine whether the tax increase and the media campaign together have decreased youth smoking by more than one might expect by adding the independent effects.
In concluding this examination of youth prevention efforts, and as a natural segue into the next section's discussion of adult smoking cessation, it is important to recognize a new frontier in tobacco control, one so new that there is insufficient literature to review: youth smoking cessation. Traditionally, youth and adult smoking concerns have been neatly distinguished as prevention and cessation, respectively (excluding, of course, exposure to environmental tobacco smoke—an issue for all age groups). Recently, however, recognition has
dawned that there are likely hundreds of thousands of young people who, by virtue of early experimentation, are beyond the prevention stage; they, like most adult smokers, are addicted. Providing cessation assistance for youth raises a series of different issues than those confronted by professionals attempting to help adults to quit. For example, given strict prohibitions against being a young smoker in many families and some schools, many addicted children are unwilling to reveal their smoking status to professionals. Also, adults can walk into a pharmacy and buy nicotine replacement products over the counter. Underage minors cannot. Clearly, this is a problem crying out for attention.
Adult Smoking Cessation
If preventing youth smoking is necessary to breaking the generational transmission of nicotine addiction, speeding up the pace of adult smoking cessation is vital to reducing the disease toll of smoking in the foreseeable future. Even if prevention of youth smoking succeeded completely, in all countries of the world, starting today—that is, not another child ever started to smoke—the tragic milestone of 10 million annual global tobacco-produced deaths three decades from now will still be realized. Reflecting the long lag between initiation of smoking and its disease sequelae, those deaths will occur in people smoking today, young to early middle-age adults.
As with youth prevention, early efforts to foster adult smoking cessation all centered on education and related activities. When the first Surgeon General's report was released on January 11, 1964 (U.S. DHEW, 1964), the optimistic expectation among government public health officials was that it would lead to widespread quitting. (The report ranked as one of the major news stories of the year [U.S. DHHS, 1989].) Certainly the early response was encouraging: during the first three months of 1964, cigarette sales fell by 15%. By the end of the year, however, they were off by just 5%. Nicotine (and smoking) addiction was offering early evidence of its tenacity in the face of overwhelming evidence of its cost.
The Fairness Doctrine antismoking messages on the broadcast media from 1967 to 1970 produced the first four-year decline in per capita cigarette smoking in the century (Warner, 1979). More recent evidence from California and Massachusetts indicates that those states' media campaigns have worked to decrease adult smoking as well (Harris et al., 1996; Goldman and Glantz, 1998). Since very few people start to smoke as adults, virtually all declines in adult smoking prevalence reflect quitting.
If antismoking media campaigns affect smoking by adults, less clear is the impact of restrictions on cigarette advertising. As industry critics have observed, with an annual budget of $6 billion devoted to advertising and other methods of promoting cigarettes in the United States, the tobacco companies are the nation's principal source of tobacco and health “education.” The collective resources of all the health voluntary organizations devoted to antismoking education, com-
bined with governmental monies dedicated to the same purpose, fall well short of a tenth of the industry's effort.
The tobacco control community believes fervently that cigarette advertising and marketing increase smoking, not only among children, who may be enticed to experiment by the seductive marketing campaigns, but also among adults. For adults, the effects of advertising are more subtle. Advertising works on adult smoking by creating an environment in which constant cues to smoke likely prompt subconscious desires to light up. Smokers contemplating quitting may find it harder to do so in such an environment, and recent quitters may find it more difficult to sustain their commitment (Warner, 1986). The ubiquity of advertising also sends a message that smoking cannnot really be as hazardous as “they” say it is; otherwise the government would not permit its presence everywhere (Marsh and Matheson, 1983). Finally, the media's dependence on cigarette ad revenues may restrict free and open discussion of the dangers of smoking. More than one empirical study has demonstrated that dependence on cigarette advertising reduces magazines' coverage of the hazards (Whelan et al., 1981; Warner et al., 1992).
Finding the “smoking gun” that demonstrates that cigarette advertising and promotion increase adult smoking has proven to be an elusive challenge. A decade ago, the Surgeon General concluded that the collective evidence supported the conclusion that advertising does increase smoking (U.S. DHHS, 1989). However, findings in the empirical literature have been mixed (Chaloupka and Warner, in press). A thorough new review of both theory and the empirical evidence presents a persuasive case that cigarette marketing does increase smoking and that comprehensive bans would decrease it (Saffer and Chaloupka, 1999). To date in the United States, however, restrictions on advertising and promotion have been partial. In 1971, cigarette ads went off the broadcast airways. In 1999, tobacco billboards came down and sports sponsorship by tobacco companies was limited (NAAG, 2000). The impacts of the latest restrictions on adult smoking, if any, remain to be established.
The other major government-produced education/information effort—placing warning labels on cigarettes and cigarette ads—has never been demonstrated to have had a significant impact on smoking. In part, this likely reflects the modest size and “drab” character of the warnings (U.S. DHHS, 1989). Dramatically larger labels, with graphic evidence of the harms caused by smoking, might be expected to command the attention of anyone who sees them (Lynch and Bonnie, 1994). That is the hope and expectation of Canadian tobacco control activists who have worked long and hard for plain packaging and large warning labels (Kennedy, 2000).
Judging intervention effectiveness by the percentage of those exposed who succeed in quitting, one must rank many formal smoking cessation programs, including medically directed treatments, as among the most effective weapons in the cessation arsenal (U.S. DHHS, 1996). Cessation programs are placed under the “education/information” category for lack of a better option. Not generally a policy measure (although there are exceptions, especially if one considers pri-
vate business policies), participation in smoking cessation programs represents an individual smoker's choice, one adopted with the expectation that the smoker will learn methods to quit (hence the categorization as education/information).
While quit rates vary dramatically across participant and program types, formal programs commonly succeed in helping 15–25% of participants to quit (U.S. DHHS, 1996; Warner, in press). This is a dramatically higher figure than those associated with virtually all other tobacco control interventions. Nevertheless, one must acknowledge the high level of motivation of participants and the very modest reach of such programs. Relatively few smokers participate in them, and it remains true that the vast majority of smokers who quit do so without the aid of a formal program. Indeed, despite their high rate of effectiveness, individual cessation treatments have been criticized by Chapman (1985) as being socially undesirable. Given their low reach and high cost per participant, they are less cost-effective than a comparable level of resources devoted to a less effective, but high-reach, intervention such as a media quit campaign. This perspective has been challenged as being too narrow: specific individuals may need the individual attention accorded program participants; for them, formal programs may be the only effective option. And even if cessation programs cost more per quitter than public health interventions, as Chapman observed, they remain a highly cost-effective health care intervention (Warner, 1997).
Smoking cessation rates respond to interventions from the incentive and regulation categories, as well as education/information. Unlike the case for youth prevention, however, for which all categories of interventions are intended to reduce youth smoking, neither of the principal incentive and regulation interventions is explicitly intended to affect smoking rates among adults. Or at least proponents of these measures do not explicitly acknowledge reduction in adult smoking as their objective, although one suspects that many proponents harbor the hope and expectation that smoking prevalence will fall when the interventions are adopted.
As with youth smoking, the main incentive intervention is taxation. Although the impact of rising prices on adult smoking falls short of that for kids, adult smoking does respond to price. The empirical evidence suggests that the degree of price responsiveness is inversely related to age (Chaloupka and Warner, in press). That is, young adults are more price responsive than middle-aged adults, who in turn respond more to price than older smokers. Research also indicates that price responsiveness is inversely related to socioeconomic status (SES). Analysis of data in the United Kingdom demonstrates that in the highest social class, smoking is virtually unrelated to price changes. In striking contrast, in the lowest social class, the price elasticity of demand (the economist's measure of price responsiveness) is close to –1.0, meaning that a given percentage increase in price will reduce low-income smokers' demand for cigarettes by a comparable percentage (Townsend et al., 1994). In the United States, Farrelly and colleagues recently found that lower-income persons had a price responsiveness 70% greater than those with higher incomes (Farrelly et al., 1998).
Chaloupka (1991) earlier had found a qualitatively similar pattern comparing price responsiveness and education, which is highly correlated with income.
Collectively, as noted above, the adult price elasticity of demand is on the order of –0.4, meaning that a 10% increase in price will reduce the quantity of cigarettes demanded by 4%. As such, the tobacco control community recognizes that, translated into a price increase, a sizable tax increase can have a significant impact on adult smoking rates. In particular, it will reduce smoking more among young and poorer adults. The latter is particularly important since smoking and its disease sequelae are most common in groups of lower socioeconomic status.
From a political point of view, a cigarette tax increase has the added virtue that it will increase total tax revenues at the same time that consumption falls, simply because the tax rate rises much more than consumption declines. However, to avoid charges of social engineering, health professionals rarely advocate tax increases with the explicit objective of decreasing smoking by adults. Rather, they appeal to the impact of taxes on smoking by youth.
In a similar vein, the principal regulatory intervention that affects smoking by adults—prohibition against smoking in public places—is never advocated for that purpose. Rather, it is sought with the third major tobacco control objective in mind: protecting nonsmokers from the dangers posed by environmental tobacco smoke (ETS). In truth, clean indoor air laws were originally, and remain, motivated by the desire to protect nonsmokers. But as the research literature demonstrates, clean indoor air laws do decrease smoking in the aggregate, increasing the cessation rate modestly and causing continuing smokers to reduce their daily consumption (Evans et al., 1996; Brownson et al., 1997; Ohsfeldt et al., 1998). In addition, they help to establish and reinforce a nonsmoking social norm, which may help to discourage youth initiation (Lynch and Bonnie, 1994).
Evidence on the impacts of comprehensive tobacco control interventions on adult smoking prevalence is modest in quantity. Again, a major source of insight is provided by analyses of the experiences in California and Massachusetts, each of which has documented significant declines in smoking prevalence attributable to these states' campaigns (Harris et al., 1996; Goldman and Glantz, 1998). One major experimental effort to establish the effect of a multichannel community-based intervention, the COMMIT program, failed in its designated objective of reducing smoking among heavy smokers. However, researchers did find a statistically significant decline in smoking by light smokers (COMMIT Research Group, 1995).
The paucity of this empirical evidence notwithstanding, there is substantial reason to believe that multiple interventions produce at least an additive impact on smoking, and possibly a whole that is greater than the sum of its parts. The federal Office on Smoking and Health argues that comprehensive programs are vital to realizing substantial tobacco control success (CDC, 2000), and the Institute of Medicine has recently concurred (National Cancer Policy Board, 2000). What is unequivocally clear is that the modest resources devoted to all tobacco control programs to date require only modest success in controlling tobacco use
to justify their existence. The prospect that they may yield more than modest success makes them a very attractive investment of public health resources.
In concluding consideration of the objective of adult smoking cessation, it seems critical to comment on the evidence on the changing demographic patterns of smoking since the inception of the antismoking campaign in 1964, because the numbers tell a compelling tale. Most notable is the change in smoking prevalence by education class. In 1965, the year following the first Surgeon General's report, less than 3 percentage points separated the prevalence of smoking among college graduates (33.7%) from that of Americans who did not graduate from high school (36.5%) (U.S. DHHS, 1989). By 1997, prevalence among college graduates had fallen by nearly two-thirds, to 11.6%. Among people without a high school diploma, in contrast, prevalence had fallen by only one-sixth (to 30.4%) (CDC, 1999b). The gap had grown to almost 19 points.
These numbers mirror others relating to income and occupation, which, of course, are closely related to education. Perhaps most striking is evidence that smoking prevalence among American physicians—among the nation's most highly educated persons—may now be less than 4% (Nelson et al., 1994; Frank et al., 1998), while it remains around 40% for certain blue-collar occupations. In short, smoking has moved from an “egalitarian” burden in the mid-1960s to a heavy weight today on those of low socioeconomic status.
Particularly given that high-income and highly educated smokers are less responsive to price increases than are low-income and education smokers (Chaloupka, 1991; Townsend et al., 1994; Farrelly et al., 1998), the socioeconomic evidence strongly suggests that the educational component of the antismoking campaign has had a far more substantial impact on high-income and education groups than on those less fortunate. To be sure, the dramatic changes in smoking prevalence among the more fortunate members of society likely reflect the interaction of effective information dissemination and social pressure. That is, as a portion of high-SES smokers quit in response to the newly emerging evidence on the dangers of smoking, smoking lost its social cache in this class quickly; indeed, it became socially less acceptable, and eventually unacceptable. But the conclusion that the lower-SES groups did not respond comparably to information dissemination and likely did not experience the same social disapprobation of their smoking, seems inescapable.
There are numerous possible explanations for the divergence in smoking trends by SES. One is that health educators communicated effectively with people like themselves—highly educated persons—and failed to deliver the tobacco-and-health message to less educated people in a manner they would comprehend and find compelling. Recent state-based campaigns devote specific resources to developing culturally and educationally relevant health education messages. Another possible explanation is that the affluent, well-educated population finds the future, especially the prospect of retirement, far more attractive than the lower-SES population for whom the future bodes only continuing economic worries. As a consequence, the low-SES population heavily discounts the future promised by not smoking.
Whatever the reason, the widening gap between high- and low-SES smoking suggests a dimension of tobacco control that, to date, has received very limited attention and clearly deserves more.
Protecting Nonsmokers from Environmental Tobacco Smoke
In many ways, the third major objective—protecting nonsmokers from environmental tobacco smoke—has proved to be the domain of the most interesting and politically contentious tobacco control activity. More than two decades ago, the tobacco industry was warned by its polling consultant, the Roper organization, that the issue of ETS, and the nonsmokers' rights movement it had spawned, posed the single greatest threat to the economic vitality of the industry (Roper, 1978). In subsequent years, grassroots activists succeeded in getting numerous and increasingly strong clean indoor air policies adopted, initially at the state level and then, increasingly, at the level of local government. The latter occurred first (and thus far most) in California, the clear leader in tobacco control policy innovation.
Californians' success in adopting scores of local ordinances demonstrated that even the financially muscular tobacco industry could be tackled with a divide-and-conquer strategy. Arguably, the industry has controlled the U.S. Congress for years (Taylor, 1984; Advocacy Institute, 1998). It has also carried enormous weight in the deliberations of state legislatures; the industry has impressive lobbying resources in all 50 states and it makes effective use of campaign contributions at the state level, as it long has at the federal level (Advocacy Institute, 1998; Monardi and Glantz, 1998). But its ability to curry favor with literally thousands of city council members is clearly limited. As one consequence, the industry adopted a strategy of working behind the scenes at the state level to promote weak state clean indoor air laws that preempted stronger local action. Initially, by carefully garbing such laws in seemingly attractive public health language, the industry often succeeded in garnering the support of local antitobacco advocates. Once it became clear, however, that the purpose of such laws was to preempt stronger measures, the tobacco control community became increasingly informed and vigilant (Siegel et al., 1997).
As with youth prevention and adult cessation, tobacco control efforts to protect the health of nonsmokers began with an educational effort to inform the public about the dangers of ETS. However, with the issue of tobacco control further along by the time ETS became a source of public concern (roughly the late 1970s), the public education effort had a more explicit policy orientation to it: the “educators ” were attempting to mold public opinion, not merely inform it, with the objective of promoting legislated restrictions on smoking in public places. That this strategy has succeeded beyond their fondest hopes is demonstrated by the rapid diffusion of clean indoor air laws through the states (U.S. DHHS, 1989); the adoption of municipal ordinances in the heart of tobacco country, such as Winston-Salem, North Carolina; the fast transformation of the American
workplace into a largely smoke-free environment; and the banning of smoking not only in restaurants but also in bars in California. Bars and gambling establishments have long been viewed as the last bastions of unrestricted smoking. Until California proved them wrong, most tobacco control advocates likely believed that prohibiting smoking in bars verged on the impossible. The early evidence indicates that in large part, the ban is working (Glantz and Smith, 1997).
The objective of protecting nonsmokers from ETS has been pursued by this highly effective combination of education and regulation strategies. Economic incentives have proven far less important to this objective than to either of the other two. They are not nonexistent, however. Although clean indoor air laws are largely self-enforcing, fines have been levied in a few instances of flagrant violations. Clearly, economic incentives have been significant motivators for businesses to adopt smoke-free policies on their own (many having done so prior to state mandates requiring smoke-free work environments): the absence of smoking in the workplace reduces a series of business costs, including cleaning and equipment maintenance, fire insurance, and the risk of liability lawsuits from employees allegedly injured by ETS in the workplace.
Lessons from the Experience with Tobacco Control
If one wishes to learn by example how a society can address a complicated and tenacious behavior-related public health problem—and to what effect—tobacco control is an excellent choice of subject matter. Tobacco is unrivaled as a source of avoidable premature mortality; as such, it clearly ranks among the nation's, and indeed world's, most important public health problems. In the complex web of its causes, ranging from nicotine addiction to a rich and influential industrial disease vector, tobacco illustrates the full spectrum of challenges to effective public health intervention. Tobacco demonstrates vividly how health professionals, motivated by the purest of causes, must learn to wrestle with unseemly politics driven by anything but pure motives. With 35 years of history under its belt, tobacco control offers a treasure trove of the dos and donts of intervention. Tobacco control is at once a model for public health optimism and a sobering reminder of the limits of intervention success.
Likely the most important lessons deriving from the tobacco control experience are all the obvious ones. A first lesson is that whenever possible, behavioral interventions should be grounded in solid science, here the evaluation research and policy analysis that have informed tobacco control for more than two decades. As one notable example, research on the effects of cigarette taxation has quite literally transformed the practice of public health policy in the domain of tobacco control and affected thinking about policy across a wide range of public health problems, including alcohol abuse (raising beer taxes to reduce underage drinking and driving (Grossman et al., 1994), unhealthy diets (California's snack tax), and even control of gun violence (taxing guns and bullets, as proposed by Senator Moynihan [Tax Treatment, 1994]). Research on the effects of clean indoor air laws, smoking cessation interventions, and other measures
has increased their utilization and enhanced the efficiency of investments in tobacco control.
The qualification to the idea of grounding behavioral interventions in solid science—“whenever possible” —is itself important. When confronting large and urgent public health problems, the lack of definitive science should never paralyze action. As discussed throughout this paper, the evidence on the impacts of many individual tobacco control interventions and on comprehensive programs as well falls short of the kinds of definitive insights that constitute the lifeblood of science. But although the findings are imprecise, the evidence is good enough to warrant substantial investment in all but a handful of such interventions. Refinements in knowledge through further research promise the attainment of greater efficiency in mounting interventions in the future.
A leading lesson from tobacco control is a self-evident but important reminder: when one enters the fray, one should be prepared to stay for the long haul. After 35 years, tobacco control has achieved a great deal, enough that it is justifiably ranked among the great public health triumphs of the past 50 years (U.S. DHHS, 1989; CDC, 1999a). Yet despite this success, the toll of tobacco remains almost inconceivably high—higher in body count in the United States than the sum total of all other drugs, licit and illicit, all injuries, intentional and otherwise, and all the major infectious diseases, including AIDS. This is hardly the time to declare victory and go home, and few tobacco control advocates have done so.
Battles have been won, often small battles, and they should be savored. But they are never an excuse to lose sight of the bigger picture. Further, the small victories should never be considered immutable: just as the smoker who quits faces months and even years of risk of relapse, tobacco control successes are not necessarily ultimate victories. Witness, for example, how the industry converted federally mandated warning labels, which have never been demonstrated to discourage smoking, into successful product liability defenses and protection (preemption) against scores of more draconian and expensive state-based warning systems. Tobacco control—and, by extension, other behavioral health intervention issues—demands constant vigilance, evaluation and reevaluation, and creativity in looking forward to the next stages of intervention.
Yet another lesson is that innovation may prove crucial in finding new ways to attack old problems. Although not discussed much in this paper, the attorneys general's Medicaid tobacco lawsuits not only have broken new tobacco control ground with novel legal theories, but have made the judicial system an active participant in the Grafting of public health policy on tobacco. Comparable creativity at the political level might redirect tobacco control policy making to the legislative and regulatory institutions designed for that purpose (Jacobson and Warner, 1999).
Certainly a major lesson, if also obvious, is that a multidimensional problem demands a multilevel intervention. The greatest U.S. success stories in tobacco control during the past decade reflect multipronged strategies, most notably the cigarette tax-financed state programs in California and Massachusetts. The no-
tion that tobacco control programs must be comprehensive (be they youth oriented or not) is by now shared virtually universally within the tobacco control community (CDC, 2000; National Cancer Policy Board, 2000).
In fairness, as noted above, if one scrutinizes the evidence for this proposition in the harsh light of day, unassailable empirical support for the proposition is lacking. But this reflects the absence of experimental opportunities to evaluate the proposition, rather than any inherent weakness in the idea itself. The vast majority of multiply pronged tobacco control interventions are “natural experiments ” lacking controls. Their successes have, in general, exceeded those that analysis permits us to attribute to one or more of their individual components; often prior research does not permit assessment of the likely effects of each component individually. For example, as noted above, the California and Massachusetts programs have decreased smoking by more than their associated tax increases, acting alone, would have. But lacking a sound body of evidence producing a consensus estimate of the impact of media campaigns, we cannot know how much credit to accord the media campaigns alone (although valiant attempts to do so have been made (Harris et al., 1996; Goldman and Glantz, 1998). Nor can we determine whether or not interaction of the component parts produced a synergistic effect. The absence of a more complete science of tobacco control precludes determination as to whether the sum is greater than, less than, or equal to its parts. A source of frustration to contemporary analysts, this situation also poses wonderful opportunities for future research, including integrative multidisciplinary research (Chaloupka, in press).
The dearth of empirical evidence notwithstanding, specific examples present a compelling case for multilevel interventions. To take but one, adult smoking cessation has resulted from dissemination of information about the dangers of smoking; from increased excise taxes; from responses to media antismoking campaigns; and from participation in formal smoking cessation programs. Two important observations verge on truisms. First, no one of these interventions could have successfully prompted all of the quitting that has occurred. The diversity of interventions was essential to bring us to the point that half of all Americans who ever smoked have quit. In the absence of any one of these interventions, a significant number of now former smokers might have remained smokers and perished prematurely as a result. That we cannot attribute a specific fraction of the quitting to information dissemination and another to media campaigns does not diminish the validity or importance of this observation. (It does, however, mean that until the evaluation science improves, we must live with the possibility that one or more of the interventions has had little impact or perhaps a cost-ineffective impact.10)
With one notable exception—cessation—the literature on the cost-effectiveness of tobacco control interventions is sparse to nonexistent. Cessation lends itself readily to cost-effectiveness analysis (CEA) and has been the subject of many CEAs (Cromwell et al., 1997; Warner, 1997). Other types of intervention, such as youth access law enforcement, should lend themselves nicely to CEA, and a few analysts are attempting to perform such analysis at the time of this writing. Nevertheless, this intervention's susceptibility to useful CEA is limited by incomplete understanding of the effectiveness of the intervention (Forster and Wolfson, 1998; Lancaster and Stead, 1999). By definition, cost-effectiveness analysis presupposes the existence of good effectiveness knowledge. This problem plagues many areas of tobacco control, helping to explain why few CEAs have been completed outside the area of cessation. There is yet another important reason that many central tobacco control interventions have not been subjected to CEA: they represent the type of intervention for which costs are so hard to quantify that CEA may be meaningless, or at least controversial. For example, although taxation is documented to be a highly effective intervention, how does one value the cost of violating the economic neutrality of the fiscal system? How does one measure the cost of possible inequities associated with the burdens placed on poor smokers by tax increases?
Second, the interventions almost certainly reinforce each other. To contemplate how, consider further the example of smoking cessation raised at the beginning of this section. Information dissemination on the hazards of smoking may move smokers from the precontemplation stage to contemplation of quitting (Prochaska et al., 1993). A media campaign mocking smoking may move the now-ambivalent smoker toward action. A cigarette tax increase may push the smoker “over the top ” toward successful quitting. And the elimination of tobacco billboards in 1999 may aid that former smoker's resolve to stay “quit.” More generally, through specific interventions (such as clean indoor air laws) and the norms they help to create (a nonsmoking ethos), the social environment becomes increasingly conducive to quitting and to staying quit.
Throughout this paper, “multilevel” has referred to intervention functions. As observed in Box 1, however, other constructions of “multilevel” are equally valid. Although not discussed in this paper, a clear lesson of tobacco control is that any effective, comprehensive assault on a major behavior-related public health problem must avail itself of multiple “levels” along other dimensions. In tobacco control, for example, success has depended on the concerted efforts, sometimes coordinated, usually not, of professionals and volunteers from the full range of organizations within our society—government agencies, voluntary health societies, and private businesses (item 2 in Box 1). The need to rely on multiple intervention channels—the media, medical and public health societies, workplaces, schools, churches—verges on the self-evident (item 4).
Application of the lessons of tobacco control to other behavioral public health problems should proceed with caution. Some of the lessons will transfer quite directly; others likely will prove inappropriate in other substantive domains. The temptation is to think that problems most like tobacco will benefit most directly from the case study. Gun control and reduction of alcohol abuse come initially to mind as concerns that have borrowed intervention lessons directly from tobacco control (and vice versa; for example, the current youth access movement in tobacco control is modeled quite consciously on the experience with alcohol control). But the differences between any two behaviorrelated problems are invariably as profound as the similarities. Guns and alcohol
each produce health consequences with an immediacy not experienced in tobacco; guns and alcohol kill children as well as adults. Tobacco is dangerous when used as intended. Consumed in moderation, alcohol benefits health.11
While applications of lessons from tobacco control to other problems (and vice versa) should proceed with caution, they definitely should proceed. To date, although a few adventurous scholars and activists have crossed the boundaries of individual public health problems, their numbers are almost shockingly low. It is nothing short of remarkable, for example, how little interaction one observes between the tobacco and alcohol control communities. The two major licit drugs—the two leading sources of death and disability in our society—have much in common. Yet consider the warning label on alcoholic beverages, “won” after a bitter political struggle. In its wordiness, diminutive size, and obscure placement, it goes virtually unnoticed. One day it may protect the alcohol producers in court. The lessons of tobacco labeling—far bolder than the labels themselves—have been available, and understood, for at least two decades.
If translation of the lessons from tobacco control to other behavioral health problems has proved limited to date, it is at least encouraging to recognize and celebrate the sharing of tobacco control lessons across nations. Thanks in part to a sophisticated Internet operation known as GlobaLink, some 1,000 tobacco control activists from scores of countries are in daily communication. Lessons about the effectiveness of taxation in discouraging youth smoking, the potential of state-based lawsuits to generate tobacco control action, and the need for comprehensive multilevel tobacco control programs are shared, learned, and put into play with increasing regularity. The challenges posed by tobacco worldwide are immense and urgent. The resources available to the public health community to address them pale in comparison with those the tobacco industry devotes to increasing them. Still, the sharing of knowledge, and the sophistication with which it is utilized, offer hope that the tobacco beast will eventually be contained. At the beginning of a new century, it may not be unrealistic to envision the elusive tobacco-free world before the next century dawns.
Advocacy Institute. Smoke and Mirrors: How the Tobacco Industry Buys and Lies Its Way to Power and Profits. Washington, DC: Advocacy Institute, August 1998.
Advocacy Institute. A Movement Rising: a Strategic Analysis of U.S. Tobacco Control Advocacy. Washington, DC: Advocacy Institute, March 1999.
Bhonsle RB, Murti RP, Gupta PC. Tobacco habits in India. In: Gupta PC, Hamner JE III, Muri PR, eds. Control of Tobacco-Related Cancers and Other Diseases. Bombay: Oxford University Press, 1992.
Brown & Williamson, Inc. http://brownandwilliamson.com/8_yspc.html. 2000.
This is a source of consternation to many public health educators who struggle with how one can convey the health message about moderate alcohol consumption without encouraging deadly abuse among the most vulnerable members of society.
Brownson RC, Eriksen MP, Davis RM, et al. Environmental tobacco smoke: Health effects and policies to reduce exposures. Annual Review of Public Health 1997;18: 163–185.
Centers for Disease Control and Prevention. Response to increases in cigarette prices by race/ethnicity, income, and age groups—United States, 1976–1993. Morbidity and Mortality Weekly Report 1998;47:605–609.
Centers for Disease Control and Prevention. Achievements in public health, 1900–1999: tobacco use—United States, 1900–1999. Morbidity and Mortality Weekly Report 1999a;48:986–993.
Centers for Disease Control and Prevention. Cigarette smoking among adults—United States, 1997. Morbidity and Mortality Weekly Report 1999b;48:993–996.
Centers for Disease Control and Prevention, National Center for Health Promotion and Chronic Disease Prevention, Office on Smoking and Health. Best Practices for Comprehensive Tobacco Control Programs, http://www.cdc.gov/tobacco/bestprac. htm. 2000.
Chaloupka FJ. Rational addictive behavior and cigarette smoking. Journal of Political Economy, 1991;99:722–742.
Chaloupka FJ. Macro-social influences: the effects of prices and tobacco control policies on the demand for tobacco products. Nicotine and Tobacco Research. In press.
Chaloupka FJ, Pacula RL. Sex and race differences in young people's responsiveness to price and tobacco control policies. Tobacco Control 1999;8:373–377.
Chaloupka FJ, Warner KE. Economics of smoking. In: Culyer AJ, Newhouse JP, eds. Handbook of Health Economics. Amsterdam: Elsevier, in press.
Chapman S. Stop smoking clinics: a case for their abandonment. Lancet 1985; 1:918–920.
COMMIT Research Group. Community intervention trial for smoking cessation (COMMIT): I. Cohort results from a four-year intervention. American Journal of Public Health 1995;85:183–192.
Cromwell J, Bartosch WJ, Fiore MC, et al. Cost-effectiveness of the clinical practice recommendations in the AHCPR guideline for smoking cessation. Agency for Health Care Policy and Research. Journal of the American Medical Association 1997;278:1759–1766.
DeCicca P, Kenkel D, Mathios A. Putting out the fires: will higher cigarette taxes reduce youth smoking? Presented at the Annual Meetings of the American Economic Association, 1998.
Dee TS, Evans WN. A comment on DeCicca, Kenkel, and Mathios. Working Paper. School of Economics, Georgia Institute of Technology, 1998.
DiFranza JR. Are the federal and state governments complying with the Synar amendment? Archives of Pediatrics and Adolescent Medicine 1999;153:1089–1097.
Evans WN, Farrelly MC, Montgomery E. Do workplace smoking bans reduce smoking? Cambridge, MA: National Bureau of Economic Research Working Paper No. 5567, 1996.
Farrelly MC, Bray JW, Office on Smoking and Health. Response to increases in cigarette prices by race/ethnicity, income, and age groups—United States, 1976–1993. Morbidity and Mortality Weekly Report 1998;47:605–609.
Ferrence R, Slade J, Room R, Pope M, eds. Nicotine and Public Health. Washington, DC: American Public Health Association, in press.
Florida Department of Health, Office of Tobacco Control. Report Regarding the Progress of the Tobacco Pilot Program, March 17, 1999.
Food and Drug Administration. Regulations restricting the sale and distribution of cigarettes and smokeless tobacco to protect children and adolescents; final rule . Federal Register 1996 (August 28);61(168):44396–45318.
Forster JL, Wolfson DM. Youth access to tobacco: policies and politics. Annual Review of Public Health 1998;19:203–235.
Frank E, Brogan DJ, Mokdad AH, et al. Health-related behaviors of women physicians vs. other women in the United States. Archives of Internal Medicine 1998; 158:342–348.
Glantz SA. Editorial: preventing tobacco use—the youth access trap. American Journal of Public Health 1996;86:156–158.
Glantz SA, Smith LR. The effect of ordinances requiring smoke-free restaurants and bars on revenues: a follow-up. American Journal of Public Health 1997;87:1687–1693.
Goldman LK. Glantz SA. Evaluation of antismoking advertising campaigns. Journal of the American Medical Association 1998;279:772–777.
Goodman J. Tobacco in History: The Cultures of Dependence. New York: Routledge, 1994.
Grossman M, Chaloupka FJ. Cigarette taxes: the straw to break the camel's back. Public Health Reports 1997; 112:290–297.
Grossman M, Chaloupka FJ, Saffer H, et al. Alcohol price policy and youths: a summary of economic research. Journal of Research on Adolescence 1994;4:347–364.
Harris JE. Cigarette smoking among successive birth cohorts of men and women in the United States during 1900–80. Journal of the National Cancer Institute 1983;71: 473–479.
Harris JE, Connolly GN, Brooks D, et al. Cigarette smoking before and after an excise tax increase and an antismoking media campaign: Massachusetts, 1990–1996. Morbidity and Mortality Weekly Report 1996;45:966–970.
Henningfield JE, Cohen C, Pickworth WB. Psychopharmacology of nicotine. In: Orleans CT, Slade J, eds. Nicotine Addiction: Principles and Management. New York: Oxford University Press, 1993.
Jacobson PD, Warner KE. Litigation and public health policy making: The case of tobacco control . Journal on Health Politics, Policy and Law 1999;24:769–804.
Jason LA, Ji PY, Anes MD, et al. Active enforcement of cigarette control laws in the prevention of cigarette sales to minors. Journal of the American Medical Association 1991;266:3159–3161.
Kennedy M. Smoking hazards to be depicted on cigarette packs. Government set to force warnings that show graphic cancer photos. Ottawa Citizen January 6, 2000.
Lancaster T, Stead LF. Interventions for preventing tobacco sales to minors. Cochrane Database of Systematic Reviews 1999;4.
Lantz PM, Jacobson PD, Warner KE, et al. Investing in youth tobacco control: A review of smoking prevention strategies. Tobacco Control 2000; 9:47–63.
Levy DT, Friend K, Coggeshall M. A simulation model of tobacco youth access policies. Working paper. Rockville, MD: National Center for the Advancement of Prevention, December 1998.
Lynch BS, Bonnie RJ, eds. Growing Up Tobacco Free: Preventing Nicotine Addiction in Children and Youths. Washington, DC: National Academy Press, 1994.
Marsh A, Matheson J. Smoking attitudes and behavior. An enquiry carried out on behalf of the Department of Health and Social Security, Office of Population Census and Surveys. London: Her Majesty's Stationary Office, 1983.
McGinnis JM. Foege WH. Actual causes of death in the United States. Journal of the American Medical Association 1993, 270:2207–2212.
Mendez D, Warner KE. Has smoking cessation ceased? Expected trends in the prevalence of smoking in the United States. American Journal of Epidemiology 1998; 148: 249–258.
Monardi F, Glantz SA. Are tobacco industry campaign contributions influencing state legislative behavior? American Journal of Public Health 1998;88:918–923.
Monitoring the Future, http://monitoringthefuture.org/, 2000.
Murray CJL, Lopez AD. Evidence-based public health policy—lessons from the Global Burden of Disease study. Science 1996;274:740–743.
Napier K. Cigarettes: What the Warning Label Doesn't Tell You. The First Comprehensive Guide to the Health Consequences of Smoking. New York: American Council on Science and Health, 1996.
National Association of Attorneys General, http://www.naag.org/tob2.htm, 2000.
National Cancer Institute. The impact of cigarette excise taxes on smoking among children and adults: summary report of a National Cancer institute expert panel. Bethesda, MD: Division of Cancer Prevention and Control, NCI, August 1993.
National Cancer Policy Board. State Tobacco Programs Can Reduce Tobacco Use. Washington, DC: National Academy Press, 2000.
Nelson DE, Giovino GA, Emont SL, et al. Trends in cigarette smoking among US physicians and nurses. Journal of the American Medical Association 1994;271:1273–1275.
Ohsfeldt RL, Boyle RG, Capilouto EI. Tobacco taxes, smoking restrictions, and tobacco use. Cambridge, MA: National Bureau of Economic Research Working Paper No. 6486, 1998.
Philip Morris, Inc. http://www.philipmorris.com/tobacco_bus/tobacco_issues/youth_ smoking_prevention.html . 2000.
Pierce JP, Gilpin EA, Emery SL, et al. Has the California tobacco control program reduced smoking? Journal of the American Medical Association 1998;280:893–899.
Pomerleau OF. Individual differences in sensitivity to nicotine: Implications for genetic research on nicotine dependence. Behavior Genetics 1995;25:161–177.
Popham WJ, Potter LD, Hetrick MA, et al. Effectiveness of the California 1990–1991 tobacco education media campaign. American Journal of Preventive Medicine 1994; 10:319–326.
Prochaska JO, DiClemente CC, Velicer WF, et al. Standardized, individualized, interactive and personalized self-help programs for smoking cessation. Health Psychology 1993;12:399–405.
Ramirez AG, Gallion KJ. Nicotine dependence among blacks and Hispanics. In: Orleans CT, Slade J, eds. Nicotine Addiction: Principles and Management. New York: Oxford University Press, 1993.
Rigotti NA, DiFranza JR, Chang Y, et al. The effect of enforcing tobacco-sales laws on adolescents' access to tobacco and smoking behavior. New England Journal of Medicine 1997;337:1044–1051.
R.J.Reynolds Tobacco Co., Inc. http://www.rjrt.com/TI/Pages/TIcover.asp. 2000.
Roper. A Study of Public Attitudes toward Cigarette Smoking and the Tobacco Industry in 1978. Vol. 1. Roper Organization, May 1978.
Saffer H, Chaloupka FJ. Tobacco Advertising: Economic Theory and International Evidence. Cambridge, MA: National Bureau of Economic Research, Working Paper No. 6958, February 1999.
Samuels B. Glantz SA. The politics of local tobacco control. Journal of the American Medical Association 1991;266:2110–2117.
Siegel M, Carol J, Jordan J, et al. Preemption in tobacco control. Review of an emerging public health problem. Journal of the American Medical Association 1997;278:858–863.
Tax Treatment of Organizations Providing Health Care Services, and Excise Taxes on Tobacco, Guns and Ammunition. Hearing before the Committee on Finance, United States Senate, 103rd Congress, 2nd Session, April 28, 1994. Washington, DC: U.S. Government Printing Office, 1994.
Taylor P. The Smoke Ring: Tobacco, Money, and Multi-National Politics. New York: Pantheon, 1984.
Tilley NM. The RJ Reynolds Tobacco Company. Chapel Hill, NC: University of North Carolina Press, 1985.
Tobacco Dependence: Innovative regulatory approaches to reduce death and disease. Food and Drug Law Journal 1998;53(suppl).
Tobacco Institute. The Tax Burden on Tobacco. Historical Compilation 1995. Vol. 30. Washington, DC: Tobacco Institute, 1998.
Townsend JL, Roderick P, Cooper J. Cigarette smoking by socioeconomic group, sex, and age: effects of price, income, and health publicity. British Medical Journal 1994;309: 923–926.
U.S. Department of Health, Education, and Welfare. Smoking and Health. Report of the Advisory Committee to the Surgeon General of the Public Health Service. Washington, DC: U.S. Department of Health, Education, and Welfare, Public Health Service, Center for Disease Control. Public Health Service Publication No. 1103, 1964.
U.S. Department of Health and Human Services. The Health Consequences of Smoking:Nicotine Addiction. A Report of the Surgeon General. Rockville, MD: U.S. Department of Health and Human Services, Public Health Service , Centers for Disease Control, Center for Health Promotion and Education, Office on Smoking and Health. Department of Health and Human Services Publication No. (CDC) 88–8406, 1988.
U.S. Department of Health and Human Services. Reducing the Health Consequences of Smoking: 25 Years of Progress. A Report of the Surgeon General. Atlanta, GA: U.S. Department of Health and Human Services, Public Health Service , Centers for Disease Control, National Center for Chronic Disease Prevention and Health Promotion, Office of Smoking and Health. Department of Health and Human Services Publication No. (CDC) 89–8411, 1989.
U.S. Department of Health and Human Services, 1991. Strategies to Control Tobacco Use in the United States: A Blueprint for Public Health Action in the 1990's. Smoking and Tobacco Control Monographs No. 1. Department of Health and Human Services, Public Health Service, National Institutes of Health. NIH Publication No. 92–3316. Bethesda, MD: National Cancer Institute, 1991.
U.S. Department of Health and Human Services. Smoking and Health in the Americas. A 1992 Report of the Surgeon General in Collaboration with the Pan American Health Organization. Atlanta, GA: U.S. Department of Health and Human Services, Public Health Service , Centers for Disease Control, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health. Department of Health and Human Services Publication No. (CDC) 92–8419, 1992.
U.S. Department of Health and Human Services. Preventing Tobacco Use Among Young People: A Report of the Surgeon General. U.S. Department of Health and Human Services, Public Health Service , Centers for Disease Control, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health. Washington, DC: U.S. Government Printing Office, 1994.
U.S. Department of Health and Human Services. Smoking Cessation. Clinical PracticeGuideline Number 18. Public Health Service, Agency for Health Care Policy and Research, Centers for Disease Control and Prevention. AHCPR Publication No. 96–0692. Rockville, MD: AHCPR, 1996.
U.S. Department of Health and Human Services. Tobacco Use Among U.S. Racial/Ethnic Minority Groups—African Americans, American Indians and Alaska Natives, Asian Americans and Pacific Islanders, and Hispanics: A Report of the Surgeon General . Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, 1998.
Wagner S. Cigarette Country: Tobacco in American History and Politics. New York: Praeger, 1971.
Walsh DC. Gordon NP. Legal approaches to smoking deterrence. Annual Review of Public Health 1986;7:127–149.
Warner KE. Clearing the airwaves: the cigarette ad ban revisited. Policy Analysis 1979;5: 435–450.
Warner KE. Selling Smoke: Cigarette Advertising and Public Health. Washington, DC: American Public Health Association, 1986.
Warner KE. Effects of the antismoking campaign: an update. American Journal of Public Health 1989;19:144–151.
Warner KE. Reducing harms to smokers: methods, their effectiveness, and the role of policy. In: Rabin RL, Sugarman SD, eds. Regulating Tobacco: Premises and Policy Options. New York: Oxford University Press, in press.
Warner KE. Cost-effectiveness of smoking cessation therapies: interpretation of the evidence and implications for coverage. PharmacoEconomics 1997; 11:538–549.
Warner KE, Murt HA. Impact of the antismoking campaign on smoking prevalence: a cohort analysis. Journal of Public Health Policy 1982;3:374–389.
Warner KE, Murt HA. Economic incentives for health. Annual Review of Public Health 1984;5:107–133.
Warner KE, Chaloupka FJ, Cook PJ, et al. Criteria for determining an optimal cigarette tax: the economist's perspective. Tobacco Control 1995;4:380–386.
Warner KE, Citrin T, Pickett G, et al. Licit and illicit drug policies: a typology. British Journal of Addiction 1990;85:255–262.
Warner KE, Goldenhar LM, McLaughlin CG. Cigarette advertising and magazine coverage of the hazards of smoking: a statistical analysis. New England Journal of Medicine 1992;326:305–309.
Warner KE, Slade J, Sweanor DT. The emerging market for long-term nicotine maintenance. Journal of the American Medical Association 1997;278:1087–1092.
Wasserman J, Manning WG, Newhouse JP, et al. The effects of excise taxes and regulations on cigarette smoking. Journal of Health Economics 1991; 10:43–64.
Whelan EM, Sheridan MJ, Meister KA, et al. Analysis of coverage of tobacco hazards in women's magazines. Journal of Public Health Policy 1981;2:28–35.
World Bank. Curbing the Epidemic: Governments and the Economics of Tobacco Con trol. Washington, DC: World Bank, 1999.