received healthy lifestyle ad\.ice by mail and bi worksite posters. Men in the inter\,ention sites found at baseline to be at high risk for cardiovascular dis- ease \vere provided medical counseling on risk factor change, including smoking cessation. At the end of the intervention in 1977-1978, a small but significant reduction in smoking prevalence had occurred among the high-risk smokers in the intervention site (Rose et al. 1980). Five intervention and five control worksites were resurveyed in 1983, approximately 12 years after the baseline screening and at least 5 years afier the end of the intervention program (Bauer et al. 1985). There was no significant difference in the prevalence of smok- ing between intervention and control factories, but the smokers at the intervention sites reported smoking sig- nificantly fewer cigarettes per day. The initial design and implementation of the North Karelia and Stanford Three-Community trials led to the design of several other cardiovascular disease prevention trials around the world. These included the Swiss National Research Program from 1977 to 1980 (Gutzwiller et al. 1985), the South African Coronary Risk Factor Study from 1979 to 1984 (Steenkamp et al. 1991), and the Australian North Coast Healthy Lifestyle Programme from 1978 to 1980 (Egger et al. 1983). The early trials also influenced the development of two communitywide mass media-based smoking cessation trials implemented in Australia in the 198Os, in Sydney from 1983 to 1986 and in Melbourne from 1983 to 1986 (Pierce et al. 1986, 1990; Macaskill et al. 1992). In the Swiss trial, two towns in the French- speaking and two towns in the German-speaking regions of the country were assigned to either interven- tion or reference status (Gutzwiller et al. 1985). Baseline surveys of risk factors for cardiovascular disease were conducted among random samples of residents aged 16 to 69 years in all four towns in 1977-1978 and repeated at the final assessments in 1980-1981. In the interval, communitywide health education and health promotion interventions were conducted in the two intervention towns, including media campaigns, counseling of high-risk individuals, and community organization efforts to encourage environmental and social changes. The prevalence of smoking in the con- munities declined from 32.8 to 27.4 percent in the intervention towns and from 37.1 to 35.3 percent in the reference towns, a significant net effect of 3.6 percent decline. In the South African Coronary Risk Factor Study, three rural communities, matched in size, socioeco- nomic status, and cultural factors, were assigned to low-intensity prevention, high-intensity prevention, and control status (Steenkamp et al. 1991). Both the low- and the high-intensity sites received a mass media educational campaign using so-called small media, such as posters, billboards, mailings, and coverage in local newspapers. In the high-intensitv community, high-risk individuals, including smokers, received personal interventions from health care pro- viders. Risk factors for cardiovascular disease were measured in a cohort of residents aged 15 to 64 years from each community in 1979 and in 1983. The baseline prevalence of smoking was higher among men (49.2 vs. 44.4 percent) and women (17.0 vs. 14.5 percent) in the high-intensity intervention community than in the control community, but the difference was not sta- tistically significant. After the four-year intervention, the net change in smoking prevalence in the high- intensity community, relative to the control commu- nity, was not significant for men but was significant for women. Women in both the low- and the high- intensity intervention communities had significantly higher rates of quitting than women in the control com- munity, but no differences were observed for men. The Australian North Coast Healthy Lifestyle Programme replicated the design of the Stanford Three-Community Study (Egger et al. 1983). In 1978, three communities in northern New South Wales, Aus- tralia, were assigned to a media intervention, media intervention plus community program, or control sta- tus. A two-year study for preventing cardiovascular disease was conducted, including a smoking cessation component called "Quit for Life." The media inter- ventions used professional commercial media and advertising techniques and a social marketing and health promotion framework involving print, posters, radio, television, and other advertising techniques. The community programs for smoking cessation in- cluded promotions of smoking cessation organizations, kits handed out by doctors, distribution of self-help materials, and telephone help lines. The smoking ces- sation campaigns also incorporated other community activities-such as organized runs, stress management training, and computerized health testing-that con- veved the overall program's broader theme of healthy lifestyles. Risk factors for cardiovascular disease, in- cluding smoking, were measured in random samples of residents aged 18 years and older in each commu- nitv in 1978 (baseline), 1980, and 1981. In the multiple log&tic regression analysis model, which controlled for baseline differences among the three communities in age and sex distributions, there was a statistically greater decline in smoking in the two intervention com- munities than in the comparison community, with the largest differences among young smokers. Declines in the prevalence of smoking in the area assigned to media inter\-ention plus communitv program ranged from 15.7 percent among men aged- 1 H-25 irears to 6.1 percent among women aged 65 Jrears and older. In the 1980s. a community~~ide mass media- based smoking cessation campaign i1.a~ conducted in Sydney and Melbourne, Australia (D\z.ver et al. 1986; Pierce et al. 1986). The Sydney campai@ began in mid- 1983, and the Melbourne campaign began one yeal later (during the preceding year, Melbourne \vas used as a control citv for the Svdnel, campaign). The "Quit for Life" campaigns in\.ol\.ed inno\.ati\,e and pro\`oca- tive smoking cessation messages deli\ cred through paid spots on the radio, on tele\.ision, and in net\-spa- pers. These messages \~ere supported bv a telephone "Quit Line," self-help "Quit Kits," and a Iiospital-based "Quit Centre," all of \j.hicti \\`ere promoted at the end of the paid ad\~ertiscmcnts used in the campaigns. The campaigns ivei-e c\,aluated through monthlv random telephone sur\.evs in the tit.0 iommuniticsl In addi- tion, a cohort of-rcsijents was inter\ it>r\.cd in April- June 1983 and again in Ma\ IYXI. In the c.ohort, 23 percent of smokers in S\,dne\- and Y percent in Melbourne quit during the initial (control) ve`ar before the campaign \Vas begun in Melbourne (Pierce et al. 1986). The monthI\' pre\,alence estimates demonstrated an approximatelv l-percent decline in Svdnev in com- parison with the rest of Australia CD\\ v& et-al. 1986). The media campaigns Lvere continued through 1986, along rvith additional programs in conjunction with physician-, school-, and communitv-based activities. Long-term evaluation of trends in smoking in the t\vo cities from 1981 to 1987 suggests that the sustained campaigns mav have contributed to a decline in smok- ing prevalenceof about 1.5 percentage points per year in both communities among men but had little impact on women (Pierce et al. 1990). An analysis of the campaign's potential differential impact across educa- tional levels suggested that the Australian mass media and community campaigns did not contribute to an increase in the gap in smoking pre\,alence bet\veen educational groups (Pierce 1989; Macaskill et al. lYY2). The lack of a consistently positive effect from these initial community trials was attributed more to an incomplete understanding of comprehensive interlren- tions and to the relatively ITeak, quasi-experimental designs of the studies than to concern about the effi- cacy of the overall approach (Farquhar 1978). The con- tinuing enthusiasm for the potential efficacy of the communitywide approach was reflected in both na- tional and international reviews and guidelines (Blackburn 1983; WHO 1982; USDHHS 1983; National Cholesterol Education Program Expert Panel 1988; Shea and Basch 19YOa,b). Similarly, the positii e results from the Australian communitywide antismoking media campaigns and smoking cessation data from the North Karelia trial encouraged the planning of smoking-specific community efforts in the United States in the late 1980s. Three major community-based trials for prevent- ing cardiovascular disease were funded by the Na- tional Heart, Lung, and Blood Institute (NHLBI) in the early 1980s: the Stanford Five-Citv Project, the Min- nesota Heart Health Program, and the Pawtucket Heart Health Program. Each had comparison and in- ter\-ention communities and stronger designs and e\.aluation methodologies than the studies initiated in the 1970s. Each study \vas developed by an indepen- dent team of investigators, and the NHLBl maintained a collaborati\.e research relationship among the stud- ies (Winklebv et al. lYY7). All three shared common inter\rention approaches that lasted five to eight years and focused on the major risk factors for cardiovascu- lar disease (hypertension, cigarette smoking, high di- ctarv fat, obesity, and sedentarv lifestyle). Each project used mass media, community mobilization, and mul- tiple educational channels, such as health care provid- ers, schools, lvorksites, and \-oluntarv agencies. The programs integrated individual and social change ap- proaches, cmploving some combination of social learn- ing theory, social neti,vork diffusion theory, and social marketing to guide the planning and implementation of the interventions (Bandura 1977; McGuire 1973; Rothman lY7Y; Rogers 1983). The three projects dif- fered initially in their relative emphasis on specific modalities (Stanford emphasized media; Minnesota, population screening; and I'aivtucket, community or- ganizations) (Shea and Basch 1990a1, but frequent col- laborations among projects decreased these differences over time. Manv innovative strategies were devel- oped, and the piocess evaluations on specific smok- ing prevention and cessation interventions were posi- tive (Glasgow et al. 1985; Sallis et al. 1985; Altman et al. lY87; Elder et al. 1987,19Y3; King et al. 1987; Lando et al. lYYO,1991; Perry et al. 1992; Pechacek et al. 1994). n`onetheless, the overall impact of the three interven- tions on smoking prevalence \vas modest. The Stanford Five-City Project began \1-ith baseline survevs in 1979. Five cities in Northern California were selected on the basis of location, size, and media mar- kets (Farquhar et al. 1985). Monterey and Salinas shared a media market and \vere assigned to the intervention group. The three control cities (Modesto, San Luis Obispo, and Santa Maria) Mere isolated from the me- dia market of the intervention communities. The communitvlvide educational campaigns began in 1980 in collaboration 12.ith existing communitv organizations. The two treatment cities recei\,ed con- tinual exposure for five years; each vear, four to five separate risk factor education campaigns took place, one of which focused on smoking. Evaluations in- cluded independent, cross-sectional population samples aged 25 to 74 years surveved at baseline and at 25, 51, and 73 months, as well as a cohort formed from the baseline survey that M'as resurveyed at 17, 39, and 60 months. Initially, the cohort samples in the intervention con;munities experienced a significantly greater decline in smoking prevalence than those in the control communities (-7.66 \`s. -3.76 percent) (Farquhar et al. 1YYO; Fortmann et al. 1993). Bv the end of the intervention in 1986, the cross-sectional surveys showed no such difference in declining prevalence. At the final folloiv-up in 1989-1990, a more rapid though nonsignificant decline was detected in the control communities than in the intervention com- munities (Winkleby et al. 1996). In the Minnesota Heart Health Project, three pairs of communities were selected, with one of each pair assigned to educational intervention and the other to comparison status (Jacobs et al. 1986; Murray et al. 1994). The communities Lvere matched on size, com- munity type, and distance from the Minneapolis- St. Paul metropolitan area. After a lh-month baseline assessment period, a 5- to 6-year intervention program was started in November 1981 in the first education site, Mankato, Minnesota (Luepker et al. 1994). The second and third education sites, Fargo-Moot-head on the North Dakota-Minnesota border and Bloomington, Minnesota, were started 22 and 28 months later in 1983. The staggered entry alloLved for a gradual develop- ment of the intervention program and a stronger e\.alu- ation design (Luepker et al. 1994). Starting in 1980, annual cross-sectional survevs among residents aged 25 to 74 years were conducted in all six sites. A ran- dom sample of residents surveyed before the start of the education program \vas resurveyed. For long-term smoking cessation, the cross-sectional survey data provided evidence of an intervention effect for M'omen but not for men; no such effect was obserlred for ei- ther sex in the cohort sample (Luepker et al. 1994; Lando et al. 1995). Unexpectedly, large declines in smoking prevalence, especiall\; among men, \vere ob- served in comparison communities. In the Pawtucket Heart Health Program, the impact of a communitywide program for reducing risks for cardiovascular disease in Pawtucket, Rhode Island, was compared with trends in a nearby matched community in southern Massachusetts (name Ivithheld to honor a confidentiality agreement with the city gov- ernment) (Carleton et al. 1995). Pawtucket was selected as the intervention site from among a pool of nincX potential northeastern New England cities; thcl comparison site had similar sociodemographic char- acteristics. Surveys of risk factors for cardiovascula, disease were conducted with random samples ot residents aged 18 to 64 years in the two communitieh at two-year intervals, beginning in 1981 and continL1. ing until 1993. Communitywide educational strategies emphasized public awareness campaigns, behavior change through existing community resources and volunteers, and community activation to promote in- volvement and environmental changes (Elder et al. 1987, 1993; Lefebvre et al. 1987). During the seven- year intervention program from 1984 to 1991, more than 500 community organizations were involved, including schools, religious and social organizations, larger worksites, and city government departments. Overall projected risk for cardiovascular disease declined significantly in Pawtucket during the educa- tional program, but the prevalence of cigarette smoking declined only slightly and did so more !n the comparison than in the intervention communitv (Carleton et al. 1995). Concurrent \vith the community-based cardiovascular disease prevention trials in the United States, an antitobacco community education program ivas initiated in India (Anantha et al. 1995). The trial ivas conducted between 1986 and 1992 in the Karnataka State. One intervention area (117 villages) and tLt.0 control areas (136 and 120 villages) were se- lected \\?thin the Kolar District. A baseline survey was conducted in 1986, and follow-up surveys were con- ducted t\vo and fi1.e years later. Villages \yere ran- domly sampled in each of the three areas, and the to- bacco use habits of all residents of each household were assessed. A subsample of the villages selected at baseline \,vas resur\,eyed t\1-o and five years later to provide cohort follow--up. After the baseline survey, a three-year educational campaign used health worker staff from Primary Health Centres to visit each village at least once a lveek and deliver health education mes- sages about the risks of cigarette smoking and other forms of tobacco use, particularly chewing. Handbills, photographs, posters, and films m multiple languages rvere used to reinforce health education counseling de- livered to individuals and small discussion groups. Among tobacco users in the intervention area, preva- lence declined 26.3 percent for men and 36.7 percent for \vomen. The proportional reduction in the preva- lence of any tobacco use was significantly greater in both men and \vomen in the intervention area than in the t\1-o control areas (10.2 vs. 2.1 and 0.5 percent for men and 16.3 vs. 2.9 and 0.6 percent for M-omen). The Federal Republic of Germany began the Ger- man Cardiovascular Prevention (GCP) Studs in the mid-1980s (GCP Studv Group 1988). The se\ren-year prevention campaign in the GCP Studv targeted more than 1 million people in six intervention regions lvhose demographic and socioeconomic structure reflected that of the West German population. The reference population was sampled from the total West German population. The goal of the campaign was to reduce four risk factors for cardiovascular disease (hyperten- sion, hypercholesterolemia, smoking, and obesitv) bv using a multifaceted prevention program. Public health services, voluntary welfare federations, institutions for adult education, sports and consumer associations, and other existing communitv resources and facilities \vere used extensively. The campaigns sought the in\.ol\re- ment of health care providers and emphasized consunl- ers' access to them. Special emphasis M'as placed on improving community knorzrledge and awareness of healthy nutrition, the benefits of physical activity, and the importance of quitting smoking. To identify per- sons at high risk for hypertension and hvpercholester- olemia, screenings were conducted at social t\,ents, in factories, and at other communitv settings in close cooperation with physicians, pharniacists, and health insurance companies. To discourage smoking, non- smoking restrictions were extended in public places, and educational campaigns M'ere conducted in the media and in community settings to promote smoking cessation and to help smokers quit. For the e\.nluntion of risk factor trends, representative samples of residents aged 25-69 years from the intervention regions and of the national population of West Germany were sur- veyed before the intervention (May 1984 to March lY86), at midstudy (February 1988 to April 1989), and at the end of the intervention (April 1991 to April 1992) (Hoffmeister et al. 1996). In the national reference sample, the prevalence of smoking declined from 34.0 percent at baseline to 33.5 percent at the end of the study. In the intervention region, the prevalence of smoking declined from 35.4 percent at baseline to 32.5 percent at the end of the study, for a net change of -6.7 percent (P < 0.001). The decline occurred exclusively among men (net change of -7.9 percent, I' < 0.001). Among women, the prevalence of smoking increased in both the intervention regions and nationwide, and no inter- vention impact was noted (net change of -1 .X percent). Using a somewhat different design, the Com- munity Intervention Trial for Smoking Cessation (COMMIT) was started in the late 1980s (COMMIT Re- search Group 1991). COMMIT focused solely on smok- ing cessation and built on the initial experience in the ongoing trials to prevent cardiovascular disease. COM- MIT M'as planned as a randomized community trial with 11 pairs of communities and had adequate statistical popver to detect relatively small intervention effects (Gail et al. 1992). One community of each pair was randomly allocated to the intervention program, and the other was monitored as a control. The 11 interirention com- munities received a four-year educational program that focused on adult cessation, with special empha- sis on "hea\,):" cigarette smokers (those who smoked 23 or more cigarettes per day). The intervention philosophy of the trial assumed that a comprehensive communitw~ide strategy would make it difficult for residents in the 11 targeted sites to avoid exposure to messages about the importance of nonsmoking and w,ould alert smokers to the many opportunities for cessation. Interventions focused on four primarv edu- cational channels: media-based and communitywide e\~ents, health care providers (e.g., physicians and den- tists), \x,orksites and other organizations, and cessa- tion resources. Within these channels, the centrally de\.eloped protocol specified 58 mandated activities, designed to be carried out largely by community vol- unteers and local staff or agencies with limited external wsources (Lichtenstein et al. 1990-1991). Intervention acti\.ities started after the baseline survev and random- ization, beginning M.ith community mobilization in Januarv 1989 and continuing lvith protocol-defined inter\wtion through December 1992. A telephone sur- vey \\-a~ conducted in each of the 22 sites to estimate baseline prevalence and identify cohorts of heavy and light-to-moderate smokers. Cohort members were contacted annually bv telephone, with a final assess- ment in early 1993. A final prevalence survey was conducted in all 22 communities from August 1993 to January 1994. There was a high degree of community owner- ship rvithin the 11 intervention sites (Bracht et al. 1994; Lichtenstein et al. 1996), and program staff and com- munity organizations diligently delivered the 58 man- dated activities. Hence, the modest effects observed in this trial were sobering for the public health com- munity (Fisher 1995; Susser 1995). No cessation effect was observed for the "heavy" smokers for whom the trial was specifically designed (COMMIT Research Group 1995a). Among the evaluation cohort of light- to-moderate smokers, a significantly greater propor- tion quit in the intervention than in the control communities (30.6 vs. 27.5 percent) over the four-year intervention period, and the effect w'as strongest among the less educated residents of the communi- ties, O\,erall the prevalence of smoking declined slightly (but nonsignificantlv) more in the intervcn- tion cdmmunities (3.5 percentage points) than in the comparison communities (3.2 percentage points) (COMMIT Research Group lYY5b). The quality and statistical power of the overall trial design (Gail et al. 1992) make it unlikely that any true intervention effects were missed. The COMMIT intervention pro- tocol sought to apply the most effective smoking ces- sation strategies as defined bv the published literature (Lichtenstein et al. 1990-1991; COMMIT Research Group 1991). The investigators were limited, however, in their ability to be involved in many of the recom- mended ecological and policy-oriented health promo- tion strategies (WHO 1979; Green and Richard 1993) because of restrictions imposed by federal funding of the study (Fisher 1995; Susser 1995). In addition, process data showed that implemented protocol did not have a significant impact on many important in- termediate variables (e.g., physician and dentist coun- seling rates, worksite smoking bans, public attitudes toward smoking) (Glasgow et al. 1997; Ockene et al. Statewide Interventions 1997; Taylor et al. 1998). Therefore, the failure of the COMMIT interventions to use certain strategies or to change intermediate social and policy variables suggests that the study was not an adequate test of the efficac-, of the social-environmental approach to reducing to- bacco use. Several reviewers have provided some perspec- tives on the modest smoking cessation effects observe(j in these community trials (Green and Richard 1YY3; Luepker 1994; Winkleby 1994; Fisher 1995; Susse, 1995). Common themes are (1) the difficulty in oh- serving intervention effects because of the largtx secular declines in risk factors for cardiovascular dis- ease, including smoking, that occurred during tllc, period when the trials were implemented and (2) tllc need for a more comprehensive health promotion a~>- preach. A more complete understanding is needed ot why such modest and mixed smoking cessation effects have been observed in numerous well-designed and well-implemented communitywide trials. Concurrent with the implementation of the corn- munity intervention trials, a broader national move- ment to reduce tobacco use began to emerge in the 1980s. Unlike the communitv intervention trials, this movement, and the large-scale inter\.cntions that developed from it, was not structured around research hypotheses and preplanned evaluation designs. Rather, the movement was characttri7ed by commu- nity mobilization at the national, state, and local levels and encompassed the principles of health pro- motion as a social movement that evolves (Kickbusch 1989; Allison and Rootman 1996; Dolvnie et al. 1996; Nutbeam 1998). Funding for these efforts came from both federal and private sources; hoLyever, an impor- tant manifestation of this national mo\.ement \vas the establishment of statewide interventions funded by increases in cigarette excise taxes or settlements \Vith the tobacco industry. Such increases Lvere the result of voter initiatives, beginning M'ith those in California in 1990 and Massachusetts in 1993. The next section of this chapter reviews the main elements of the na- tional movement. Community Mobilization A significant step in organizing the movement to reduce tobacco use w'as the founding in 1981 of the Co.as to be set aside for preventing tobacco use. Further, this legisla- tion authorized the Commissioner of Health to launch a major state\Vide initiative-the Minnesota Tobacco- Use Prevention Initiative-to promote nonsmoking and established state aid for school-based programs to prevent tobacco use. The legislation allocated funding to support the school-based programs at the rate of SO.52 per student during the 1985-1986 school year and SO.54 per student during future years. School districts \vere authorized to use these nets funds for staff in-service training, cur- ricula and materials, community and parent alvareness programs, and e\,aluation. Three principles guided the state's tobacco con- trol programs. First, a broad base of public support was developed by the collaboration of the Minnesota Coalition for a Smoke-Free Society 2000, the Associa- tion for Nonsmokers-Minnesota, \wluntary health agencies, health professionals, and insurers. Second, the program maintained a positive approach that stressed the consequences of tobacco use rather than attacked the tobacco industry or blamed smokers. Third, the program focused on preventing tobacco use among adolescents and young \vomen ~t.ho had not yet become addicted to cigarettes or smokeless tobacco. The mass media campaigns were the most vis- ible component. The campaigns included paid televi- sion, radio, and outdoor/transit advertising directed at two target populations: 12- to 13-year-old boys and girls and 18- to 24year-old women. The goal of the media campaign was to change a social climate that encouraged the use of tobacco. Advertisements fo- cused on increasing the awareness of the negative as- pects of tobacco use that are most important to young people-unpleasant social and personal consequences, such as bad breath, smelly clothes, and addiction. To foster community tobacco control programs, T/w Mirlrwsofo PIorr for Nor7srrmkirzg nrzd Hcalfh recom- mended that schools, health services, and other community organizations be involved in providing prevention and education programs about tobacco use. A granting program was established in 1986 to fund 21 proposals from local organizations that could dem- onstrate a coordinated approach for involving multiple local organizations in the prevention effort. A second cycle of local projects was funded in 1988. Schools throughout the state were involved in an intensive effort to plan, implement, and evaluate effec- tive programs for students from kindergarten (K) to grade 12 and in technical institutes. Since the start of these programs in the 1986-1987 school year, the per- centage of school districts addressing smoking in grades K-4 steadily increased but remained fairly constant in grades 5-10. The number of school districts in the state rz.ith a tobacco-free policy, howelrer, steadilv increased. Each of the main program elements- funded by the Minnesota Tobacco-Use Prevention Initiative has been e\.aluated (Minnesota Department of Health 1989, 1991). L'outh and adults targeted by the program were a\\`are of the media campaign, and the evaluation data suggested that the campaign improved young people's attitudes tolvard tobacco use (Minnesota Department of Health 1991). There \vas a steady increase in the number of school districts wfhose curricula included components for pre\,enting tobacco use (Minnesota Department of Health 1991). Nonetheless, a prospec- tive study indicated that schools using the prevention curricula pvere not more effective in reducing adoles- cent tobacco use than bvere a randomized control group of schools (Murray et al. 1992). In that study, a com- parison of trends in adolescent tobacco use in Minne- sota and Wisconsin betbveen 1986 and 1990 found a slightly larger (but nonsignificant) net decline in Min- nesota: The investigators suggested that greater reach and penetration of preventive efforts may be required to produce statelride reductions in adolescent tobacco use (Murray et al. 1992). California In November 1988, the Tobacco Tax and Health Promotion Act (Proposition 99) \vas passed by Cali- fornia voters, thus mandating the start of California's Tobacco Control Program. The program is the largest and most comprehensive undertaken in the United States to reduce tobacco use. Initially, the program defined three long-term objectives: (1) to reduce the initiation of cigarette smoking by children and youth under age 19 from the 1987 rate of 26.4 percent to no more than 6.5 percent bv 1999, (2) to reduce cigarette smoking among adults aged 20 years and older from the 1987 rate of 26.0 percent to 6.5 percent by 1999, and (3) to reduce smokeless tobacco use among males aged 12-21 years from the 1987 rate of 8.9 percent to no more than 2.2 percent by 1999 (Tobacco Eclucation Oi,ersight Committee 1991). The excise tax rate on cigarettes in California rose from $0.10 to $0.35 on Janu- ary 1, 1989, Lvhen Proposition 99 \\`as implemented. On Januarv 1, 1993, the tax increased to $0.37, \vhere it remained in 1999. Funding for tobacco control efforts began during fiscal year lY8Y (July 1989-June 1990). The fiscal year 1999 budget in California was $126.8 million ($3.90 per capita) for tobacco control activities funded bv the Department of Health Services and the Departn&t of Education. The NCI's planning frame\vork (NC1 1991) was used to establish the program's target groups, in- tervention channels, and interventions to reach them (Bal et al. 1990). Community mobilization is a key part of California's extensiire program for reducing tobacco use. Community-based programs are the responsibil- itv of the California Department of Health Services and 6i local health departments (58 county and 3 city). These local agencies, advised by local coalitions, es- tablished multiple subcontracts with community- based organizations to conduct events, programs, and presentations for diverse racial and ethnic groups (To- bacco Education Oversight Committee 1991). Local lead agencies have been a cornerstone of the program by mobilizing communities to eliminate exposure to ETS, by closing channels for minors' access to tobacco, and by advising local policymakers. The local lead agencies receive approximately 20 percent of funds allocated for education programs to achieve these ends. The statewide media campaign, which receives about 12 percent of funds, has been the program's most visible element. Launched in 1990, the media cam- paign has focused primarily on changing public opin- ion to denormalize tobacco use. In particular, it has sought to raise public awareness of the tobacco industry's manipulative and deceptive marketing tactics and of the dangers of ETS. Although young people are a direct target audience for some campaign messages, the campaign has focused more on chang- ing social norms and reducing adult tobacco use to influence youth, many of whom begin using tobacco to be more adultlike. Funding for the statewide me- dia campaign was about 924 million ($0.75 per capita) in 1998 but has varied considerably over the years, as is discussed later in this section. About 16 percent of education funds are spent on competitive grants to community-based organizations. More than tlvo-thirds of these grants have targeted racial and ethnic minoritv communities. The competi- tive grants program has-had multiple funding cycles, and 46 separate projects were funded in 1993. In ad- dition, the competitive grants program funds several state\vide projects, such as the Tobacco Education Clearinghouse of California, which distributes library and video materials, and the California Tobacco Con- trol Resource Partnership, Lvhich provides technical assistance and training to local lead agencies. The com- petitive grants program has also been used to estab- lish regional linkages among local governments and local nongovernmental organizations. Twenty-four percent of the education funds go to school-based programs to prevent tobacco use and are distributed through the California Department of Education. The project estimated that it would reach approximately 350,000 students through programs implemented be- tkveen 1993 and 1996. The single largest share, by far, of the education funds-59 percent through 1996-goes to the medical care programs. This percentage is notably higher than the 35 percent specified by the legislation (Novotny and Siegel 1996). As a result of this redistribution, the portions of the program that deal with reducing to- bacco use-designated for 20 percent of the fund- have never been fully financed. In the first year, 16.5 percent of funds were allocated for such program efforts; in the second cycle, 12 percent were allocated; in the third, 10 percent. This diversion of funds was the result of executive decisions and was strongly sup- ported by the tobacco industry and the California Medical Association. After the third diversion, civil action was initiated by Americans for Nonsmokers' Rights, supported by the American Lung Association and the ACS, to prevent the reallocation. The Sacra- mento Superior Court found in favor of the plaintiffs in early 1995. The state appealed, and the judgment for the plaintiffs \vas upheld in December 1996 (A~~rcri- cfl115 for Norrsmoke~s' Ri~llfS i'. stfite of Cdifmifl). The complicated course of these events, as de- tailed by Novotny and Siegel (1996), has highlighted the role of the tobacco industry in countering efforts to reduce the use of its products and the opposing strat- egy of health advocates. Begay and colleagues (1993) have pointed out that since Proposition 99 passed, the tobacco industry's political expenditures in California have risen tenfold, from $790,050 in the 1985-l 986 elec- tion to S7,615,091 in the 1991-1992 election, during which the tobacco industry contributed more heavily to candidates for the California legislature than to can- didates for the U.S. Congress. In a further analysis, this same research group (Traynor et al. 1993) detailed the specific industry strategies to prevent local con- trol of tobacco use. Using case studies, they docu- mented the industry's use of front groups to conceal its involvement, its organization of local referenda to defeat or suspend local ordinances, and its financing of local election campaigns to repeal ordinances by popular vote. Glantz and Begay (1994) have also analyzed the relationship between campaign contri- butions and votes on individual tobacco-related bills in the California legislature. Using a "tobacco policy score" (p. 1178) that ranked legislators according to their stance for or against reducing tobacco use, they found a significant relationship between the amount of money received from tobacco sources and a protobacco position. This ongoing documentation of tobacco industry influence, though not a formal part of the California Tobacco Control Program, has been one of its notable features, and it provides a model of health advocacy for other states and localities. The program, xvhich has e\,olved considerably since 1989, remains a multifocal, multichannel ap- proach to the broad range of issues that confront large- scale efforts to reduce tobacco use (Tobacco Education and Research Oversight Committee 1995; Pierce et al. 1998a). In 1993, the California Tobacco Control Program was revised, and program priorities ivere refocused (Pierce et al. 1998a). Four broad priority areas, or policy themes, Ivere established for use in the program planning and funding decisions: . Protecting people from exposure to ETS. . Revealing and countering tobacco industry influence. o Reducing young people's access to tobacco products. o Providing cessation services. The California Tobacco Control Program contin- ues to place its primary emphasis on a broad statewide infrastructure that reaches into communities across the state. The program's basic structure is composed of a state-level office and several statewide and regional programs that foster a collaborative grassroots approach to serve a decentralized structure of community pro- grams across the state (Pierce et al. 1998a). Surveillance and evaluation activities to assess program performance and impact were established as part of the initial program structure (Bal et al. 1990; Tobacco Education Oversight Committee 1991). The evaluation is composed of large triennial surveys (Pierce et al. 1994, 1998a) and smaller ongoing sir- veys (Pierce et al. 1998b), a more targeted evaluation of program components (Independent Evaluation Consortium 19981, and a wide array of local program evaluation efforts. Evaluation is complicated, how- ever, by the multiplicity of prevalence surveys avail- able and by potential error from using data from surveys with differing methods (Novotny and Siegel 1996; Siegel et al. 2000). Establishing specific rela- tionships between large-scale social interventions and a change in tobacco use is difficult, but the temporal relationship between the decline in California's to- bacco consumption and the efforts generated by Proposition 99 can be clearly observed. Per Capita Cigarette Consurnptiolt Before the implementation of the program in 1989, the rate of decline in monthly per capita ciga- rette consumption was 0.42 packs, which was signifi- cantlv greater than the rate of 0.36 in the rest of the couniry (Pierce et al. 1998a,b). From Januarv 1989 through December 1993, the decline in California in- creased significantly, to 0.65 packs, while the decline in the rest of the United States increased nonsignifi- cantly, to 0.45 packs. Until early 1992, the media pro- gram M'as the only part of the tobacco control program that M'as fully implemented. An econometric analysis (Hu et al. 1995) has estimated that of the 1,051-million pack decrease in sales between 1990 and 1992, approxi- matelv 232 million (22 percent) \vere attributed to the media campaign and the remaining 819 million (78 percent) to the excise tax increase. Between 1993 and 1996, the rate of decline in per capita consumption in California slokved significantly, to 0.17, but virtually halted altogether in the rest of the country (at 0.04 packs) (Pierce et al. 1998b3). Consumption decreased more rapidly in California than in the rest of the coun- try, even though the California cigarette excise tax changed only slightly during this period (from $0.35 in 1993 to $0.37 in 1994). Between 1993 and 1996, how- ever, expenditures for tobacco control were reduced by more than 50 percent from their initial funding lev- els in fiscal year 1990 and 1991. During 1989-1993, spending for advertising and promotions by the tobacco industry exceeded tobacco control expendi- tures in California by a ratio of about 5 to 1; from 1993 to 1996, that ratio increased to nearly 10 to 1 (Pierce et al. 1998b). Adult Smoking Prevalence Data on adult patterns of smoking pre\,alence are not as consistent or as easy to evaluate as consump- tion trends (Novotny and Siegel 1996). Nevertheless, the trends in these data are consistent with the pat- terns noted in the per capita consumption analyses. From 1989 to 1993, smoking prevalence declined in California almost twice as rapidly as in the rest of the country (Pierce et al. 1998b). However, from 1993 to 1997, the rate of decline in California appeared to ~10~`. Overall, smoking prevalence has declined from 26.7 percent in 1988 to 16.7 percent in 1995 in California and from 30.2 percent in 1988 to 24.7 percent in 1995 in the rest of the country (CDC 1996; Pierce et al. 1998b). A recent analysis of trends in adult prevalence of smok- ing in California compared with the rest of the United States observed a significant decline in smoking preva- lence in California from 1985 to 1990 and a slower but still significant decline from 1990 to 1994, a period in which there was no significant decline in the remain- der of the nation (Siegel et al. 2000). Youth Tobacco Use Prevaleuce The lack of consistent youth smoking surveil- lance data between California and other states has impeded the evaluation of program impact on tobacco use among young people in California. Holz-ever, one multivariate analysis of data from the school-based Monitoring the Future survey of Sth-, 1 Oth-, and 12th- grade students showed that irom 1992 to 1994, the in- crease in youth smoking rates that was experienced nationwide was slowed significantly in California (I' < 0.001, controlling for price, smoking policies, and other nonprogram effects) as a result of the combined effect of the tax increase in 1994 and the implementa- tion of the state's tobacco control programs (Chaloupka and Grossman 1996). Pierce and colleagues (1994) have concluded that the media campaign was successful in stopping the rise in teen smoking that had been oc- curring in California before the campaign launch. Results from other analyses of youth tobacco use data are consistent with the result found by Chaloupka and Grossman (1996). In data reported by the Califor- nia Independent Evaluation Consortium, between 1991 and 1996, rates of smoking during the past 30 days among California youth in the 8th and 10th grades in the Monitoring the Future survey increased, but the increase in California was less pronounced than in other states (Independent Evaluation Consortium 1998). Among Sth-grade youth, since 1993 the preva- lence of smoking during the past month has varied from 12 to 14 percent in California while steadily in- creasing from 17 to 22 percent in the rest of the coun- try. Similarly, among IOth-grade youth, past-month smoking prevalence in California has been about 18 to 19 percent since 1992 while increasing from 22 to 32 percent in the rest of the country. Data from the telephone-based California Youth Tobacco Survey in- dicate that the prevalence of smoking during the past 30 days among 12- to 17-year-olds increased from ap- proximately 9 percent in the early 1990s to 11.9 per- cent in 1995. Prevalence declined gradually after 1995, to 10.9 percent in 1997, while increasing in the rest of the country (Pierce et al. 1998a). Other Fidiugs Since the start of the program in 1990, numerous changes in intermediate outcomes have been noted related to changes in social norms; clean indoor air policies in public places, worksites, and bars; and vol- untary policies to ban smoking in homes. Massachusetts In November 1992, Massachusetts voters ap- proved an initiati\,e petition known as Question 1, establishing the Health Protection Fund with revenue generated from a 25-cent increase in the state's ciga- rette excise tax and a 25-cent increase in the wholesale price of smokeless tobacco products. Revenues have been used to fund the Massachusetts Tobacco Control Program, a comprehensive set of activities and services that emphasize prevention programs at the local level and that focus on young people. The Massachusetts program was modeled, in part, on California's pro- gram. The overall goal of the program was to reduce tobacco use in Massachusetts by 50 percent by the end of 1999 (Abt Associates Inc. 1995). With the passage of Question 1, the excise tax on cigarettes in Massa- chusetts rose from $0.26 to $0.51 on January 1, 1993. This tax was fully absorbed by the industry through wholesale price reductions (CDC 1996). However, in October 1996 the cigarette tax increased to $0.76 per pack (with comparable increases on smokeless tobacco products), where it currently remains. Funding for tobacco control efforts began with a large media campaign in October 1993. In late 1993 and early 1994, funding for local agencies was begun, and several statewide initiatives were undertaken to provide direct services, as well as technical assistance, training, and materials for localities. Starting in late 1994, with the first year of complete implementation, the program received $43.1 million (33.7 percent) of the $127.8 million placed in the Health Promotion Fund created by the revenues from the excise tax increase. Other key programs receiving appropriations from the Health Promotion Fund were those for comprehensive school health education ($28.8 million, or 22.5 percent of the Health Promotion Fund in fiscal year 1995), drug education ($5.0 million, or 3.9 percent), and other health-related programs ($50.7 million, or 39.7 percent) (Abt Associates Inc. 1995). After the first funding year, the program's budget declined to $41.8 million in 1995- 1996 and to $36.8 million in 1996-1997. Funding was increased for other programs receiving appropriations from the Health Promotion Fund (Abt Associates Inc. 1997). Community-based education activities and pre- vention activities are two main elements of the Mas- sachusetts program. The state's 10 regionally based, primary care Prevention Centers have added a com- ponent for reducing tobacco use and provide ongoing technical assistance and training to local community programs. Local community initiatives have included programs to increase community awareness about the hazards of tobacco use, to promote tobacco-free workplaces and public facilities, and to enforce local regulations and ordinances for reducing tobacco use; needs assessments in the community; mobilization of youth service agencies to prevent and reduce tobacco use among children and adolescents; funding of community-based agencies to Lvork Lvith at-risk adult populatiot&, including cultural and linguistic minority groups, women of childbearing age, and blue-collar workers; and funding of school-based health centers (Abt Associates Inc. 1995). Per Capita Cigarette Corrszmption As in California, Massachusetts has experienced a persistent pattern of I c ecline in per capita cigarette consumption. Before the 1993 implementation of these tobacco control programs, per capita cigarette con- sumption was declining in Massachusetts at a rate approximately equivalent to that of the rest of the coun- try (6.4 percent in Massachusetts and 5.8 percent in the states other than California [CDC 19961). Between 1992 and 1997, per capita consumption in Massachu- setts declined by 31 percent (from 117 to 81 packs per adult), while the decline in the remaining 48 states was only 8 percent (Abt Associates Inc. 1997). Between 1993 and 1996, the decline in per capita consumption has been more consistent in Massachusetts than in California (CDC 1996). Although program funding declined about 15 percent in Massachusetts from 1995-1996 to 1996- 1997 (Abt Associates Inc. 1997), it declined less than in California. Adult Smoking Prevalence Adult smoking prevalence has been monitored in Massachusetts both by the annual survey conducted through the Behavioral Risk Factor Surveillance Sys- tem (BRFSS) and by special Massachusetts Adult To- bacco Surveys conducted in 1993,1996, and 1997. Data from the BRFSS indicate that adult smoking prevalence in Massachusetts declined from an average of 23.5 percent for 1990-1992 to 20.6 percent in 1997. In the rest of the country (excluding California), prevalence declined from 24.1 percent in 1990-1992 to 23.4 per- cent in 1993-1995 (CDC 1996; Abt Associates Inc. 1997). The Massachusetts survey produced different preva- lence estimates but corroborated a similar decline in the prevalence of smoking among adults in Massachu- setts (from 22.6 percent in 1993 to 21.1 percent in 1996 and 20.6 percent in 1997) (Abt Associates Inc. 1997). Youth Tobacco Use Prevalence As in California, the observed nationwide in- crease in the prevalence of smoking among young people from 1992 to 1994 was significantly less ev- dent in Massachusetts (Chaloupka and Grossman 1996). Follow-up data from the Youth Risk Behavior Survey (YRBS) indicated that the prevalence of cur- rent smoking among Massachusetts high school stu- dents (grades 9 to 12) declined from 35.7 percent in 1995 to 34.4 percent in 1997 while increasing from 34.4 to 36.1 percent nationwide (CDC 1996, 1998). Data from the YRBS and other survey sources suggest a dif- ferential pattern by age: the prevalence of current smoking increased in Massachusetts among older stu- dents in a manner similar to that of the rest of the coun- try but declined among younger students. Between 1993 and 1996, the prevalence of smoking during the past 30 days among 8th-grade students in Massachu- setts declined from 26.5 to 26.0 percent but increased from 16.7 to 21.0 percent nationwide (Briton et al. 1997). For Massachusetts, the prevalence of current smoke- less tobacco use among 9th-12th graders decreased from 8.4 percent in 1995 to 6.0 percent in 1997; for males, the decline was from 15.1 to 10.3 percent (Kann et al. 1998). In the nation as a lvhole between 1993 and 1996, lifetime use of smokeless tobacco among 9th- 12th graders decreased from 25 to 20 percent, and cur- rent use decreased from 9 to 6 percent (Briton et al. 1997). The most recent data from the 1999 YRBS in Massachusetts indicated a continuing decline in the prevalence of current smoking, down to 30.3 percent among 9th-12th graders (GoodenoLv 2000); however, national comparison data for 1999 are not yet available. A 1996 survey of 12- to 14-year-olds in Massachu- setts and a national comparison sample (Houston Herstek Fa\,at, Youth exploratory 1996, Massachusetts Department of Public Health, presentation of findings, unpublished data) found that Massachusetts youth had significantly higher le\Tels of agreement \vith issues addressed in the state media campaign. For example, 59 percent of Massachusetts youth but only 35 percent of youth in the national sample agreed M'ith the state- ment, "Smoking cigarettes decreases vour stamina and smokers have a hard time keeping up in sports." Re- sults from a longitudinal survey of Massachusetts youth provided additional support for the efficacv of the Massachusetts antismoking media campa-ign (Siegel and Biener 2000). In a four-year follow,-up of youth aged 12 to 15 vears in 1993, this studv found that among the younger-adolescents (aged 12 tb 13 years at baseline), those exposed to antismoking advertisements were significantly less likely to progress to established smoking. However, among older adolescents (aged 14 to 15 years at baseline), exposure did not prevent pro- gression to established smoking. Other Findings There have been multiple changes in intermedi- ate measures of program impacts related to youth access, protection of nonsmokers from ETS, and avail- ability of cessation services (Abt Associates Inc. 1999). For example, by 1999, nearly two-thirds of Massachu- setts residents lived in cities and towns with some kind of smoking restriction in restaurants, and 26 percent were protected bv complete bans. Prior to the start of the program, less than 1 percent of Massachusetts resi- dents lived in towns with complete bans. Additionally, the local restaurant smoking restrictions were found to be more restrictive in communities receiving funding from the Massachusetts Tobacco Control Program. Arizona In November 1994, Arizona voters passed Propo- sition 200, which increased the state cigarette excise tax from $0.18 to $0.58. Revenues from the tax increase were earmarked for the state's Medicaid program (70 percent of revenues), for programs for preventing and reducing tobacco use (23 percent), for research on pre- vention and treatment of tobacco-related disease and addiction (5 percent), and for an "adjustment account" (Arizona Tobacco Tax and Health Care Act 1994, sec. 2C4) to offset lost revenue to other state programs currently funded by revenue from the existing $0.18 excise tax (2 percent). The petition drive to place the initiative on the November 1994 state ballot and the campaign to win voter approval was led by the Ari- zona for a Healthy Future coalition. Although public support for the initiative was strong when it was first proposed in 1993 (71 percent in favor, with 56 percent indicating strong support), the initiative was vigor- ously opposed in a well-funded advertising effort on television, in posters, and by direct mail. Proposition 200 was narrowly approved, garnering approximately 51 percent of the vote (Nicholl 1998). With the passage of Proposition 200, analysts estimated that the revenues earmarked for tobacco prevention and education programs would be ap- proximately $25 million per year (Meister 1998). However, measures passed during the 1995 session gave the legislature control over the funds and lim- ited expenditures to $10 million per year (Madonna 1998). Additionally, multiple restrictions were placed on hots the funds could be used, and an advisory com- mittee was appointed that included legislative and business representatives hostile to the program (Meister 1998). Although the Coalition for Tobacco- Free Arizona led an effort to keep the goals of the newly created Arizona Tobacco Education and Prevention Program (AzTEPP) "comprehensive," the program efforts were narrowed to a focus on youth prevention; adult cessation activities were restricted to pregnant women and their partners. Not until the fiscal year that began on July 1, 1997, with a new governor and health department director, were the programmatic restrictions lifted from the health department and the program allowed to proceed with the implementation of the "draft" comprehensive tobacco control plan originally proposed by the Coalition for Tobacco-Free Arizona. The expenditures of AzTEPP reflect the political history of the program: $9.7 million in fiscal year 1996, $18.2 million in 1997, and $28.2 million in 1998. Al- though the countermarketing campaign has expanded (with spending increasing from $7.4 million in 1996 to $13.2 million in 1998) (Riester and Linton 1988), the greatest expansion in the program has been in the scope and focus of the local programs (Meister 1998) (with funding increasing from $1.7 million in 1996 to $9.4 million in 1998). Recent program efforts have focused on all of the elements in the coalition's draft comprehensive tobacco control plan (Meister 19981, thereby expanding its adult cessation activities (discussed at the fourth annual AzTEPP meeting in February 19991, but one of the factors that had been minimized in earlv health department efforts was evaluation. Only recently have baseline data coilec- tion surveys been initiated (Meister 1998); as a result, no outcome data have been reported on the program, and subsequent evaluation efforts will be compro- mised by the lack of baseline data collected before the start of the multiple large-scale program efforts. Respondents to an initial statewide telephone survey conducted in 1998 (Arizona Cancer Center 1998); about two and a half years after the media campaign's launch, reported that the advertising cam- paign, which stressed how damaging tobacco use is and how unappealing it is to the user, to peers, and to the opposite sex, had influenced their attitudes in the intended direction. For example, HO percent of young people reported that the advertisements made them think about the negative aspects of tobacco use, and 58 percent of pregnant or postpartum women said the advertisements made them uncomfortable around smokers. Young people who had been exposed to the television advertisements in the previous 30 days were less likely to be susceptible to using tobacco than were youth who had not seen the advertisements. The campaign's impact on reported behaviors is less clear, especially among young people. Among respondents who were using tobacco at the start of the campaign, 23 percent of adults, 37 percent of pregnant or post- partum women, and 27 percent of young people said the advertising campaign had convinced them to try quitting. However, 23 percent of young people also reported that the campaign had convinced them to irzcwnsc their tobacco use. Cummings and Clarke (1998) noted that such an unintended effect, if it is real, might represent young smokers' negative reaction to a narrowlv focused youth campaign with no messages directed at changing broader social norms. In response to a request from the Arizona Joint Legislative Audit Committee, the State Auditor Gen- eral conducted a performance audit of the AzTEPP (State of Arizona, Office of the Auditor General 1YYY). This audit noted that evaluations of the state and local levels of programs have not yet produced an adequate assessment of the program's tobacco control efforts. Thus, the audit recommended that the program needed to improve its evaluations to measure its ef- fectiveness in preventing youth from starting to use tobacco, encouraging and assisting tobacco users to quit, and reducing exposure to secondhand smoke. Specifically, the audit found that the program had been unable to establish a baseline on tobacco use among youth and had only preliminary assessments in place to assess cessation services. The program has estab- lished adequate methodologies to measure the preva- lence of adult smoking; however, follow-up results are not yet available. Thus, the audit concluded that "The program's evaluation approach to date leaves it far short of knowing whether its programs are working" (p-ii). In response to this audit, the Arizona Department of Health Services (AzDOHS) has implemented changes in its surveillance and evaluation systems. Expanded surveillance systems for youth have been planned and will be implemented in 2000; however, no baseline data are available on youth smoking rates, For adults, a baseline survey of adults was conducted in 1996 and repeated in 1999. Using methodology simi- lar to that used by the state BRFSS, the 1996 and 199~ Arizona Adult Tobacco Surveys were conducted by telephone interviews on representative samples of more than 4,500 adults in Arizona aged 18 years and older. Results from these surveys indicate that the prevalence of smoking among adults declined from 23.8 percent to 18.8 percent overall (AzDOHS 2000). Among adults aged 18 to 24 years, a significant de- cline was observed also, from 27.5 percent in 1996 to 21 .O percent in 19YY. Both of these rates compare very favorably to national trends, where rates overall among adults have not declined in recent years and rates among younger adults have been increasing. Finally, smoking rates among Hispanics declined from 23.5 percent to 13.6 percent, which was the largest decline seen in any race/ethnic group in the state. Multiple other indicator variables suggest that these changes may be related to increases in smoke-free policies, ad- \?ce from doctors and dentists, and exposure to tele- \,ision antismoking information. Finally, these declines in smoking prevalence are consistent with declines in per capita sales (Orzechowski and Walker 2000) that indicate that declines in Arizona since 1996 are larger than those observed in the rest of the country. Oregon On November 5, 1996, Oregon voters approved Measure 44, raising the state cigarette excise tax from SO.38 to $0.68 (with a proportional increase in the tax rate on other tobacco products) and designating 90 percent of the increased revenue for the Oregon Health Plan (to expand insurance for medically underserved state residents) and the remaining 10 percent for a statewide tobacco prevention and edu- cation program managed by the Oregon Health Divi- sion. Survey data indicated that support for the initiative was increased by having the new revenue earmarked in this way (CDC 1997; Nicholl1998). The Oregon campaign to place the initiative on the Novem- ber 1996 ballot was initially led by the Committee to Support the Oregon Health Plan, tvhich represented primarily the private health care sector. Nonprofit and public health organizations added their support and worked in a loosely organized network led by the ACS. Later in the campaign, both groups combined efforts and resources. The measure had strong support from state media (receiving endorsements from all major newspapers and a majority of the smaller ones), from leading business groups, and from the governor, who conducted a three-day supportive media tour before the election. The Oregon Health Division used its existing Oregon Tobacco Control Plan as the model for the new statewide program. Revenue from Measure 44 dur- ing the 1997-1999 biennium was projected to be $170 million; of this, 10 percent (approximately $17 million) per biennium was appropriated to fund the Tobacco Use Reduction Account administered by the Oregon Health Division. The resulting Oregon Tobacco Pre- vention and Education Program has eight elements: (1) local community-based coalitions, (2) comprehen- sive school-based programs, (3) statewide public awareness and education campaigns, (4) a cessation help line, (5) tribal tobacco prevention programs, (6) multicultural outreach and education, (7) demon- stration and innovation projects, and (8) statewide leadership, coordination, and evaluation. The 1997-1999 biennium budget for these eight elements is combined into five categories: (1) local coalitions-$6.5 million (38 percent), (2) public aware- ness and education-$4.6 million (27 percent), (3) state- wide and regional projects-$2.75 million (16 percent), (4) schools-$2 million (12 percent), and (5) statewide coordination and evaluation-$1.2 million (7 percent). Evaluation data from Oregon indicate that the program has successfully implemented each of the program elements and is achieving its performance objectives (Oregon Health Division 1999). Local community-based coalitions were created in all 36 Oregon counties. Twenty-four school projects were funded, reaching 58 of the 198 (30 percent) school dis- tricts in the state. Surveys indicated that approxi- mately 75 percent of adults and 84 percent of the young people recalled seeing the state's public awareness campaign. In January 1999, more than 1,500 Orego- nians called the cessation help line. All nine federally recognized Indian tribes in Oregon are now receiving funding to implement prevention and education pro- grams to reduce tobacco use. Multicultural outreach and education programs have been established for Hispanic, Asian/Pacific Islander, and African Ameri- can populations in Oregon. Five demonstration projects have been funded focusing on pregnant women, health care delivery systems, and creative ways to reach youth audiences. The program has also established a comprehensive and multifaceted surveil- lance and evaluation system and has strengthened program management. Trends in per capita consumption in Oregon were compared with the remainder of the country (exclud- ing California, Massachusetts, and Arizona) for the period before program implementation (1993-1996) and after (1997-1998). From 1993 to 1996, consump- tion increased 2.2 percent in Oregon and decreased 0.6 percent in the rest of the country (CDC 1999b). In 1997 and 1998, per capita consumption declined 11.3 percent in Oregon (from 92 to 82 packs per adult). Be- tween 1996 and 1997, per capita consumption in the rest of the country declined only 1 .O percent (from 93 packs per adult to 92 packs per adult). Smoking prevalence among adults in Oregon has been consistent with the observed declines in per capita consumption. Data from the BRFSS indicate that the prevalence of smoking among adults aged 18 years and older in Oregon declined from 23.4 percent in 1996 to 21.9 percent in 1998 (Oregon Tobacco Prevention and Education Program 1999). The proportion of women who smoked during pregnancy, as reported on state birth certificates, dropped from 17.7 percent in 1996 to 15.2 percent in 1998. Data suggest that smoking rates among young people are continuing to increase as in the rest of the country. Maine In June 1997, the Maine legislature approved HP. 1357, An Act to Discourage Smoking, Provide Tax Re- lief and Improve the Health of Maine Citizens, which increased the state cigarette excise tax from $0.37 to $0.74 and earmarked the increased revenue for the Tobacco Tax Relief Fund. The act established the To- bacco Prevention and Control Program within the Maine Bureau of Health and provided $3.5 million in funding for fiscal years 1998 and 1999. The legislative effort to gain passage of the act was a combined effort of the state public health community, legislative lead- ership, and executive branch support. The Bureau of Health has developed the Maine Tobacco Prevention and Control Program to expand the existing ASSIST program structure and to meet the legislative requirement of the 1997 state statute. The legislation specified that the program include an on- going, major media campaign; grants for funding community-based programs; program surveillance and evaluation; and law enforcement efforts regard- ing transportation, distribution, and sale of tobacco products. The program's initial $4.35 million annual budget included $1.6 million for a multimedia cam- paign, $1.25 million for community and school grants, $625,000 for statewide cross-cutting activities, $400,000 for state staffing, $400,000 for evaluation, and S75,OOO for enforcing youth access provisions. In April 2000, legislation was passed in Maine that appropriated additional funds to expand the Maine Tobacco Prevention and Control Program; a total of $18.3 million from the settlement is going to tobacco control. Of this total amount, $8.35 million will be used for community and school-based grants, funding communities and schools to achieve the goal of reducing tobacco addiction and use and resulting disease, with a focus on those at highest risk such as youth and disadvantaged populations. About $6.75 million will be used for cessation and statewide mul- timedia campaigns; 51.2 million is for evaluation for independent program evaluation, research, and out- comes monitoring; $200,000 funds five positions in the Bureau of Health for administering the programs; and $1.8 million for improved prevention and treatment of tobacco-related diseases for those with Medicaid Insurance. Programs Funded by State Settlements With the Tobacco Industry As was discussed earlier in this report (see "Legislative Developments" and "Master Settlement Agreement" in Chapter 51, all 50 states, the District of Columbia, and five commonwealths and territories have settled lawsuits with the tobacco industry to rc- claim statewide costs spent treating Medicaid patients for diseases related to tobacco use. Four of those states settled their individual lawsuits vzith the industry- Mississippi in July 1997, Florida in September 1997, Texas in January 1998, and Minnesota in Mav IYYH- and the remaining parties jointlv settled in November 1998 in the multistate Master Settlement Agreement. Because of a "most favored nation" clause (ex- plained in "Recovery Claims by Third-Party Health Care Payers" in Chapter 5), the four separate settle- ments have been closely linked, particularly in 110~ the terms of their awards affect the kind of compre- hensive programs discussed in this chapter. Most notably, when the State of Florida received in its settlement $200 million that was earmarked for a two-year pilot program to reduce tobacco use among young people, the State of Mississippi, though it had settled its lawsuit earlier, received $62 million for the same type of pilot program specified in its lawsuit. Texas and Minnesota received no such additional aw.ard, because their lawsuits did not specifically set aside funds for a parallel pilot program, although Min- nesota received funds earmarked for smoking cessa- tion and tobacco-related research. Language in the Texas and Minnesota settlements, however, released Florida and Mississippi from existing requirements to use their pilot program funding within two years and to direct their programs exclusively to young people. Because program planning in Florida and Mis- sissippi was already in place when the youth-onlv restriction was removed, an emphasis on preventing tobacco use among young people has been evident in their pilot programs' first years of activities. These activities are described in the next two sections of this chapter. Brief descriptions of settlement-funded plans in Texas and Minnesota follow. This report does not attempt to describe the various plans and legislative proposals that are developing (at the time of this writ- ing) in the 46 states, the District of Columbia, and the five commonv~ealths and territories included in the joint settlement of January 1998. Mississippi The Partnership for a Healthy Mississippi, a nonprofit corporation representing a broad range of public and private interests, plans and manages the state's pilot program. The program's mission is to cre- ate a youth-centered, statewide collaboration dedicated to fostering a healthier Mississippi and eliminating to- bacco use among Mississippi youth. The partnership ivill award grants in five designated areas: (1) commu- nitv/school/youth activities and partnerships, (2) law enforcement, (3) public awareness, (4) health care ser- vices and research, and (5) evaluation. In the first year, with a budget of $23.7 million, approximately 25 community and youth partnership coalitions were funded, and more are planned for the second vear. Local coalitions-one-quarter of whose membership must be young people-are among the stateM-ide and regional organizations supported by community assistance statevvide partner grants to provide training, tobacco prevention activities for ra- cial and ethnic minority groups, and other technical as- sistance. Specific programs that have been funded by the partnership are 4-H Youth Programs, Frontline (an advrocacy organization for 14- to 18-year-olds), com- prehensive school health programs, and a comprehen- sive school health nurses pilot project. In the first two years, $4 million has been allocated to these activities. The law enforcement program has awarded grants to municipalities to enforce the Mississippi Juvenile Tobacco Access Prevention Act of 1997. These awards lvill range (accordin g to population size) from a minimum of 55,000 per municipality to a maximum of $250,000. A total of 512.65 million has been bud- geted over the first trz,o years of the program for these aM-ards. The grants \\-ill require municipalities to con- duct periodic enforcement checks on the illegal sale of tobacco to minors, provide retailer education pro- grams, provide education programs in schools, orga- nize vouth partnerships, and 1% ork \vith communit\ coalitions on enforcement issues. Other enforcement activities are being performed statetridc bv the Mis- sissippi Attorney General's Office. The partnership has budgeted $12.3 million fol a countermarketing media campaign and other pub- lic aw-areness acti\,ities to be conducted during the first t\vo years. The health care ser\.ices and research con- ponent focuses on nicotine addiction and cessation among young people. An expenditure of 55 million is anticipated for the first and second years for training health providers in cessation counseling, for research- ing childhood and adolescent tobacco abuse, and for coordinating cessation services in the state, including a telephone help line. The Mississippi State Depart- ment of Health will manage the e\~aluation of the pi- lot program and \Vill focus on program effecti\-eness in preventing initial tobacco use among young people, helping young people quit smoking, and reducing young people's exposure to ETS. An expenditure of S2 million is anticipated for the first and second years' evaluation activities. Since 1998, the Partnership for a Healthy Missis- sippi has managed the pilot program to reduce youth tobacco use through a seven-member Board of Direc- tors (www.healthy-miss.org) (McMillen et al. 1999). The major youth programs that have been implemented have included (1) the Reject All Tobacco (RAT) pro- gram among students in grades K-3, (2) the Students Working Against Tobacco (SWAT) Program for students in grades 4-7, and (3) the Frontline youth advocacy movement. Community programs have involved 26 community/youth partnership grants, targeted pro- grams in collaboration with statewide organizations, and the school nurse program in 52 Mississippi school districts. Grants have funded 245 municipalities and 74 counties to empower the local law enforcement agen- cies to reduce sales to minors. Cessation services have included the Adolescent and Child Tobacco Treatment Center and a Mississippi Tobacco Quitline. Finally, a "Question It" public awareness campaign has focused on the 12- to 17.year-old audience. The Mississippi State Department of Health has established a consortium of evaluation contracts in- volving multiple state universities to implement program evaluation efforts. The o\rerall coordination is being managed by the Social Science Research Cen- ter at Mississippi State University, with the evaluation of the media component conducted by the University of Mississippi, community programs conducted by Jackson State Universitv, law enforcement component by Mississippi State University, and the school nurses component by Mississippi State University (McMillen et al. 199Y). A baseline Social Climate Survey of To- bacco Control and Tobacco Use was conducted in 1999 among 3,040 adults aged 18 years and older that provided benchmark data on several social norm intermediate indicator variables (McMillen et al. 1999). Sur\.eillance of youth tobacco use patterns is being con- ducted by the Mississippi State Department of Health. The Youth Risk Behavior Survey was conducted among students in grades 9 to 12 in 1993,1995, 1997, and 1999 and among students in grades 6 to 8 and 9 to 12 in 1998 and 1999. Results indicate that in Mississippi, smok- ing rates among students in grades 9 to 12 had been increasing, as in the rest of country, between 1993 and 1997 (Mississippi State Department of Health 2000). Betlveen 1997 and 1999, smoking rates among students in grades 9 to 12 appear to have stopped increasing and leveled off. Among students in grades 6 to 8, smoking rates did not decline betlveen 1998 and 1999. Florida Program planning and implementation initially were managed by the Governor's Office, with direct leadership provided by Governor Lawton Chiles, who 12-as a partv to the state's lawsuit and a member of the small tean; who negotiated the settlement agreement. The Florida Tobacco Pilot Program is now managed by the Office of Tobacco Control within the Florida De- partment of Health. The program has sought the input of Florida youth in planning the program focus and materials and in working toward the main goals of changing voung people's attitudes about tobacco use, increasing youth empowerment through community involvement, reducing young people's access to tobacco products, and reducing youth exposure to ETS. These four goals will be addressed through program compo- nents similar to those of the Mississippi program: . Marketing and communications initiatives are planned to directly counter the tobacco industry's marketing efforts. A commercial advertising firm, working closely \vith teen advisors, has developed the "Truth" campaign, a direct attack on the image of smoking as cool and rebellious. The campaign's multichannel approach-based on techniques used by the tobacco industrv-incluhcs tcle\,isinn, print, and billboard acl\.ertising, as ~fell as consumer items, such as "Truth'`-imprinted T-shirts and stickers. Youth programming and community partnership activities recruited young people to a Teen Tobacco Summit in early 1998 to ad\rise on the overall de- velopment of the program. Chapters of Students Working Against Tobacco are currently active in all 67 counties. Education and training programs focus on school- aged children. Conducted in partnership with communities, schools, voluntary agencies, profes- sional organizations, and universities, these pro- grams ensure that effective tobacco prevention curricula are presented in middle and high schools across the state and that tobacco prevention strat- egies are being implemented in grades K-12 in conjunction with the Sunshine State Standards. Enforcement initiatives are aimed at improving Florida's efforts to reduce the accessibility of to- bacco products to minors. The Florida Department of Business and Professional Regulation, Division of Alcoholic Beverages and Tobacco, provides en- forcement, educational, and marketing initiatives to ensure compliance \vith all tobacco laws. The evaluation and research component monitors the performance of each of the program initiati\,es and the progress of the overall program in meeting goals and objectives. Under the leadership of the Florida Department of Health, and \vith the con- sultation of the Unilrersitv of Miami, baseline data were collected by Florida universities in all major areas before the pilot program began in earlv 1998. In the first full year of operation, the program budget was approximately 570 million, \\ith program component allocations of approximately $26 million for marketing and communications, SlO million for youth programming and community partnerships, S13 mil- lion for education and training, S8.5 million for enforce- ment, and $4 million for evaluation and research. An additional $5 million was budgeted for programs tar- geting minority populations and 53.5 million for ad- ministration and management. In the second year, approximately $45 million more r\.as appropriated for program operations; however, there were significant unexpended funds from the first year of operations that enabled major program components, such as the mar- keting and communications activities, to continue a level of expenditure similar to the first year. Youth Tobacco Use Preualerlce Between 1998 and 1999, the prevalence of i.,I'. rent cigarette use among middle school student\ (grade? 6 to 8) declined from 18.5 to 15.0 percent (CD( 1999~). Among high school students (grades Y to I?), current cigarette use declined from 27.4 to 25.2 pL,1`- cent. However, these declines were significant on], for non-Hispanic white students; the change in cLIr- rent smoking among non-Hispanic black and Hispanic middle and high school students was small and llcjll- significant. Current cigar use declined significantl\ only for middle school students (from 14.1 to 11.9 per- cent), and this decline was almost entirely among males. Similarly, current smokeless tobacco use C~C- clined only among middle school students (from 6.q to 4.9 percent) and remained unchanged among high school students. In early 2000, additional declines in youth to- bacco use were observed (Florida Department of Health 2000). Current cigarette use among middle school students declined to 8.6 percent, or an overall 54-percent decline since the 1998 baseline. Among high school students, current cigarette use declined to 20.9 percent, or an overall 24-percent decline since the 1998 baseline. Although declines between 1998 and 1999 \vere significant only for non-Hispanic white students, the declines observed in 2000 were significant among all racial/ethnic groups, except among the non- Hispanic black and "other" categories of high school students. Declines in current tobacco use, which in- clude the use of cigars and smokeless tobacco, also l\.ere significant. Since the 1998 baseline survey, cur- rent cigar use declined by 46 percent among middle school students and 21 percent among high school stu- dents. Smokeless tobacco use declined by 54 percent among middle school students and by 19 percent among high school students. Declines in current to- bacco use w'ere consistent across grade, gender, and ethnicity as \vell. Using additional data collected as part of the overall program evaluation, the Florida Tobacco Con- trol Program has connected the declines in youth smoking prevalence with program activities (Univer- sitv of Miami 1999). Results suggest that students who reported recei\?ng elements of a comprehensive to- bacco use prevention education in school had greater declines in smoking between 1998 and 1999 than those students who reported not receiving such education in school. Similarly, the Community Partnerships in the 67 Florida counties were classified as "excellent," "a\rerage," or "needing improvement" based upon program record data, and these ratings were linked to data from the Florida Youth Tobacco Survey fur I YY8 and lYY9 in those counties. Declines in smoking prevalence lvere related to the classification, lvith the greatest declines among middle and high school stu- dents in counties rated as "alwage" or "excellent." Similar ratings of counties on the level of local enforce- ment of youth access laws ivere related to vouth smok- ing prei.alence, Lvith the highest le\,els of enforcement in counties with the lotz.est prevalence. Finally, data from the Florida Anti-Tobacco Media E\.aluation (FAME) have indicated that the "Truth" campaign is producing impressi\~e alvareness among vouth and changes in attitudes and knwzledge consistent lvith the campaign themes. BetIveen lYY8 and lYYY, the prevalence of Florida youth aged 16 years and uncle1 1%.ith antitobacco attitudes increased from 59 to 64 per- cent but decreased slightlv nationlvide. National data against trhich to compare the Florida data from 1998 and 19YY are not yet a\.ailnble, but some data suggest that the prevalence of tobacco use among young people mav ha\,e peaked nation- lvide and could be starting to decline (Universitv of Michigan 1998). III addition, the impact of state exiise tax increases that have occurred since the 1 YYX baseline data collection might be assessed. Adult Smoki?lg Preualerzce In 1998, the Florida Behavioral Risk Factor Surveillance System (BRFSS) expanded its assessment of tobacco issues. The tobacco module will enable changes to be assessed in tobacco use prevalence, cessation behalriors, family rules about tobacco use, environmental tobacco smoke exposure at home, and workplace policies regarding smoking. Texas The legislative plan developed by the Texas Interagency Tobacco Task Force (1998) incorporated the CDC recommendations for community and school- based programs to reduce tobacco use. The plan in- cludes a public awareness campaign, cessation and nicotine addiction treatment, programs for diverse or special populations, enforcement of laws to reduce minors' access, surveillance and evaluation, and state- wide program administration. The plan requests $20.75 million for fiscal year 2000 and $61.25 million for fiscal year 2001 to implement, evaluate, and ad- minister the programs proposed. In the fall of 1999, the Texas legislature created an endowment fund of $200 million and requested the Texas Department of Health to conduct a pilot study based upon recommended interventions included in the 1998 tobacco task force plan. This pilot would be funded by int-estment revenue from the endowment fund, approximately S9 million per year. In response to this requirement, the Texas Department of Health has begun an Intervention Effectiveness Pilot Study in conjunction \\ith uni\,ersities in the state. To assess the impact of tobacco use prevention activities in the state, the Texas Department of Health has conducted the Texas Youth Tobacco Survey in 1998 and 1999 among middle and high school students from a sample of students statewide and in eight regions of the state. Results from the 1998 survey indicated 31 percent of middle school students and 43 percent of high school students \yere currently using some form of tobacco products (Texas Department of Health). For cigarettes alone, 21 percent of middle school students and 33 percent of high school students were current smokers. Minnesota Settlement Program In Minnesota, the Minnesota Partnership for Ac- tion Against Tobacco, the Tobacco Work Group of the Minnesota Health Improvement Partnership, and the Minnesota Blue Cross and Blue Shield (which received a separate S469-million settlement award [see "Recov- ery Claims by Third-Party Health Care Payers" in Chap- ter 51) all have developed plans for the statewide effort to reduce tobacco use. In the 1999 Omnibus Health and Human Services appropriation bill, the Minnesota legislature set aside $968 million from the state's tobacco settlement to establish two health-related endowments: one for preventing tobacco use and supporting local public health efforts (S590 million) and the other for tobacco-related medical education and research ($378 million). The interest earned from these endowments will support long-term programs. The 1999 Minnesota Omnibus Health and Human Services bill established an ambitious goal to reduce tobacco use among young people by 30 per- cent bv the vear 2005. In response to this, the Minne- sota Department of Health developed the Mirz~~esotn y~ifll Tol~rrccc, Pww~~ti(~l Illifintiw: Strrrtl;tric Plnrr (Min- nesota Department of Health 1999). This plan defined major activities that will be funded from January 1, 2000, through June 30,2001, in four component areas: Statewide Public Information and Education Cam- paign, Statewide Programs, Community-Based Pre\,ention Programs, and Youth Leadership Projects. The strategic plan established "initial indicators of suc- cess" for each program component to enable program performance to be assessed. The Statelvide Public Information and Educa- tion Campaign Ivill have a proposed budget of $7.5 million for the l&month period. The campaign will include both a media component and grassroots organizing efforts focused on the target audience of 12- to 1 T-year-old youth. The Statewide Programs will be budgeted at $3.55 million for the initial l&month period. Evaluation activities, training, and technical assistance services will be funded along with statewide organizations to support the community-based efforts. The Community-Based Prevention Programs will be budgeted at S4.4 million for the initial 1%month period. Community-based prevention efforts will include tobacco-use prevention activities at the local level and projects that focus on populations at risk. Finally, the Youth Leadership Projects will be budgeted at $1 million for the initial 18-month period and will work in conjunction with the community-based prevention efforts. These activities will seek to em- power Minnesota's youth to take leadership in the planning and implementation of tobacco prevention and control programs at the local level. The Minne- sota Department of Health has established an evalua- tion plan to track progress of the initiative, with the first comprehensive report on program effectiveness to be delivered to the legislature in January 2003. Programs Meeting the Needs of Special Populations The recent Surgeon General's report K~l~~rn) USC? Anzor1g U.S. Rarinl/Efl~lic Mirlorit!/ Gvo~r;j$ provided a summary of the various approaches that have been used to prevent and control tobacco use among racial/eth- nic minority groups in the United States (USDHHS 1998). This report highlighted the need for more re- search on the effect of culturally appropriate programs to address this problem. Few new findings ha\,e emerged since the publication of that report; hence, the elimination of disparities in health among population groups remains hampered by the lack of culturally ap- propriate programs of proven efficacy. Belolv are some examples of community-based interventions that have proven to be effective and that may serve as examples for the development of future program initiatives. Uniting and mobilizing the movement to reduce tobacco use among racial/ethnic groups have not been easy. Tension frequently occurs between various orga- nizations within the community regarding appropriate strategies to achieve particular goals, "turf" disagree- ments, competition for fund-raising dollars, and other issues. Many of these problems were identified during the 1989-1992 COMMIT trial. Though COMMIT researchers did not attribute to internal dissension the program's inability to reach its goals (Thompson et al, 1993), internecine rivalry can splinter community mo- bilization efforts and greatly impair the effectiveness of any program trying to reduce tobacco use. Diverse views and dissent are an expected part of organizing activity. A more serious issue for com- munity mobilization has been a lag in engendering support from all segments of society. Historically, the movement to reduce tobacco use has been dominated by organizations composed of middle- and upper-class white Americans and often led by white males (see Chapter 2). For many years, participation in the move- ment was further limited to organizations concerned with health and medical issues and nonsmokers' rights. In the early 198Os, increasing dissatisfaction was voiced by women and underrepresented communities who felt that their issues and contributions were not adequately integrated into mainstream efforts to reduce tobacco use (Jacobson 1983). In recent years, a number of persons and organizations representing more diverse perspectives have assumed a greater role (see the text boxes "Uptown, " "X," and "Dakota"). Particularly in view of the tobacco industry's targeted marketing to women, African Americans, Hispanics, and young people (USDHHS 1994,1998), such heightened activity is of critical importance to ensure a nonsmoking norm Mithin diverse communities. In some instances- exemplified by the low and declining smoking preva- lence among African American youth (USDHHS lY91)-such a norm may have already taken hold. Programs for the African American Community Several leadership groups, such as the National Black Leadership Initiative on Cancer, which is funded by the NCI, and the National Association of African Americans for Positive Imagery, funded in part by the CDC, have begun to have a voice in activities to re- duce tobacco use in the African American community. For example, in 1989, a strong coalition guided com- munity mobilization efforts to mount a successful cam- paign against the test-marketing of Uptown, a neM brand of cigarettes targeting African Americans (see the text box "Uptown"). A similar community- organized campaign in 1995 resulted in the withdrawal of X, another new brand seemingly intended for the African American community (see the text box "X"). In lYY2 and 1993, the ACS provided funds for community demonstration projects to use Patlrwm~s to FI.CC~OV~: Wi~lj~ill:; tile fi,$t Apimt Tobncco, a self-help guide for African American smokers (Robinson et al. Uptown I n mid-December 1989, R.J. Reynolds Tobacco Company announced that on February 5,1990, it would begin test-marketing a new cigarette in Phila- delphia, Pennsylvania. The cigarette, to be named Uptown, was the first to be marketed directly to African American smokers. Within 10 days of this announcement, the Coalition Against Uptown Ciga- rettes (CAUC) was formed. Using existing church and community organizations and word of mouth, the coalition gre\v to include 26 diverse organiza- tions representing health, religious, and community groups. The group's leaders bvere African Ameri- cans w.ith long-standing ties to the Philadelphia African American community. The Philadelphia chapter of the National Black Leadership 1nitiatiL.e on Cancer, an organization funded in part LX the National Cancer Institute and dedicated to r&luc- ing cancer in the African American community, and the Committee to Prevent Cancer Among Blacks facilitated the coalition's formation. Also acti\.e in the CAUC were se\,eral other organizations that addressed local issues on cancer control. These groups included chapters of the American Cancer Society and the American Lung Association, as \\.ell as the Fox Chase Cancer Center. The CAUC decided that its initial goal \~ould be to limit R.J. Reynolds' ability to use Philadelphia as a test market by convincing African American smokers to boycott the ne\,v cigarette. The coalition mobilized both smokers and nonsmckers in support of this goal by focusing on R.J. Revnolds' strategv to promote tobacco use among Africin Americans. %he coalition initially used local media to reinforce the messages being sent through grassroots channels and did not seek out national coverage, which the coali- tion members believed would hinder their goal of 1992). Awardees used Pnthnys to Frccdoril to bring tobacco control efforts to the African American com- munity. Through these demonstration projects, many ACS divisions began or enhanced their work in the African American community. A recent study in three predominately low- income, African American neighborhoods has demon- strated that culturally appropriate interventions can produce significant declines in smoking behaviors (Fisher et al. 1998). The Neighbors for a Smoke Free building a local, grassroots constituency. On behalf of the CAUC campaign, Dr. Louis Sullivan, then Secretary of Health and Human Services, addressed the Cniversity of Pennsylvania School of Medicine on January 18, 1990. In his remarks, Secretary Sullivan said that "at a time when [African Ameri- cans] desperately need the message of health pro- motion, Uptown's message is more disease, more suffering and more death for a group of people al- ready bearing more than its share of smoking- related illness and mortality" (quoted in Heller lYY0, pp. 32-3). The national media embraced the story. Sec- retary Sulli\,an's remarks were prominently fea- tured in the evening news and were front-page headlines across the country. R.J. Reynolds initially responded by defending their targeted marketing strategy, but the companv later claimed that Up- to\vn \vas not aimed specifically at African Ameri- cans. On January 19, lYY0, R.J. Reynolds canceled the Philadelphia test-marketing of Cptolvn. On Januarv 31, 1990, the company canceled production of the cigarette. The course of events suggests that the Uptown coalition played a decisive role in altering R.J. Reynolds' targeting strategy. A united response from Philadelphia's African American community, an or- ganized local grassroots effort, the strategic alliance \vith a national figure, and media management were associated M.ith product cancellation less than two months after introduction. The episode highlights the importance of timing in measures to reduce to- bacco use. In this instance, a marketing campaign appears to have been derailed in its beginning stages by short-term, high-intensity media advocacy (see "Media Advocacy," later in this chapter). North Side organized residents in wellness councils to encourage nonsmoking in their areas. A citywide advisory council, composed mostly of African Ameri- cans, carried out central planning for the program and provided linkages to community resources and tech- nical assistance to neighborhood councils. The pro- gram implemented a wide range of activities over a 24month period, including smoking cessation classes, billboard public education campaigns, door-to-door campaigns, and a "gospelfest." A quasi-experimental X I n early 1945, the memory of the grassroots vic- torv against Uptown cigarettes (see the previous text box, "UptoMn") served as a rallying cry in the African American community in Boston against the potential threat of a new brand-X cigarettes. As with Uptown in Philadelphia, the first information about this cigarette brand came in local media- in X's case, in articles in the Bo.sto/~ C/oh and the Bostorr Htwld. This distinctive menthol cigarette brand was packaged in the Afrocentric colors red, black, and green and featured a prominent "X," a symbol fre- quently associated with the well-knoMn, deceased African American leader Malcolm X. Community leaders in Boston and throughout the United States thought that the product had the potential to attract young African Americans-a group whose smok- ing rates had dropped dramatically in recent years. The use of "X" on a cigarette brand also \vas seen as a defamation of Malcolm X, a noted nonsmoker. Al- though manufactured and distributed by two com- panies tvithout large marketing budgets, there M'as a fear that even a small success with X cigarettes would stimulate the creation of similar products by the major tobacco companies, which would have significant resources for advertising and promotion in African American communities. The National Association of African Americans for Positive Imagery (NAAAPI) and the Boston- based organization Churches Organized to Stop Tobacco took the lead in opposing X cigarettes. T\VO NAAAPI leaders, Reverend Jesse W. Bro\vn, Jr., and design l\.as used to evaluate the impact of this pro- gram. The three intervention neighborhoods in St. Louis were matched by ethnicity, income, and educa- tion with three comparison zip code areas in Kansas City, Missouri. Baseline and follow-up random-digit dialing telephone surveys were conducted among adults (aged 18 years or older) in the three interven- tion and three comparison areas in 1990 and in 1992. Smoking pre\.alence declined significantly in the St. Louis neighborhoods, from 34 to 27 percent, but declined only slightly in the Kansas City comparison areas, from 34 to 33 percent. Thus, the results of this trial suggest that a culturally appropriate community- organizing approach to smoking cessation that Chary:-b D. Sutton, both of whom had been involved in the Coalition Against Uptown Cigarettes, spoke in Boston in February 1995 about the need for com- munities to mobilize against tobacco marketing, Their visits were covered extensively by print and broadcast media. As a result of NAAAPI's orga- nizing efforts, the manufacturer and distributor of X cigarettes received calls from around the coun- try, most notably from the organizations involved in the African American Tobacco Education Net- work of California. Because the brand's marketing seemed to be confined to the Boston area, NAAAPI decided to demand in writing that X cigarettes be withdrawn immediately to prevent any wider distribution. The manufacturer (Star Tobacco Corporation, Petersburg, Virginia) and distributor (Stowecroft Brook Distribu- tors, Charlestown, Massachusetts) both responded within 10 days to that request, although they contin- ued to insist that the cigarette brand had not been specifically targeted to the African American com- munitv. On March 16,1995, news conferences were held in Boston and Los Angeles by tobacco advo- cates to announce the withdrawal of X cigarettes from the market. The course of events suggests that the actions of activist groups had direct influence on the out- come. As Leas the case writh the Uptown protest, the X experience suggests the critical role of a rapid but organized community response in efforts to pre\,ent the targeted marketing of tobacco products to racial and ethnic minority groups. emphasizes local authority and involvement in pro- gram planning can have a significant impact on the smoking behavior among residents of low-income, African American neighborhoods. Programs for Women The Women vs. Smoking Network, a project of the Advocacy Institute, was the first national network of lvomen's organizations and women's leaders to focus on reducing tobacco use among rvomen. With financial support from the NCl, the network provided technical assistance and information to `Lz'omen's orga- nizations in an effort to interest them in the movement to reduce tobacco use. The network also focused on obtaining media coverage for issues concerning rvomen and smoking. The netxvork's most notable ef- fort was the release of a plan by R.J. Reynolds to mar- ket cigarettes to young, uneducated women (see the text box "Dakota"). Subsequent media attention made this one of the most widely covered tobacco stories of 1990 (Pertschuk 1992). The netwwk XV~S short-lived (1989-19911, however, because of lack of funding. The International Nettvork of Women Against Tobacco (INWAT) \vas established in 1990 as an international organization to counter the marketing and promotion of tobacco products to \vomen and to foster the dc\.el- opment of programs for the pre\rention and cessation of tobacco use among \vomen. Through support from the American Public Health Association, INWAT has 11-orked to dralv attention to issues concerning rz~~men and tobacco and has sought to unite and inform Ivomen's advocates around the \vorld. As a record of its Herstories project, INWAT assisted in preparing an issue of L%r,lll SuwX-itl;< :711d Htv7ltil (INWAT 1994) that \~as a collection of brief essavs about the role of to- bacco in \vomen's li\res in \`arious countries. INWAT has also published and distributed an international directorv that lists tvomen \~ho are advocates for reducing tobacco use and includes their areas of spe- cialization (American Public Health Association 1991). The National Coalition for Women Against Tobacco, lvhose sponsoring organization is the American Medi- cal Women's Association, provides educational mate- rials and advocacy messages to counteract tobacco industry marketing and combat tobacco use among Lvomen and girls (http:// M~w.womeliagainst.org). Federal and State Programs At the federal level, the CDC's IMPACT program awarded three-year cooperative agreements in 1994 to selected national organizations to enhance their work in reducing tobacco use at the national, state, and local levels. Organizations were chosen on the basis of their ability to provide services and outreach to young people, women, blue-collar and agricultural workers, African Americans, Hispanics, Asian Ameri- cans and Pacific Islanders, and American Indians. Among the states, California has made a concerted effort to involve racial and ethnic minority groups and \vomen in its efforts funded-by Proposition 99-to reduce tobacco use (see the section on California, ear- lier in this chapter). In 1990, four organizations were funded to form networks among Hispanics, African Americans, Asian Americans and Pacific Islanders, and American Indians. Members of the networks convene meetings, share experiences, participate in the devel- opment of culturally appropriate materials, and help community organizations reach their respective com- munities. These net\\Torks currently conduct programs and campaigns to build a strong statewide coalition among their respective populations (Tobacco Educa- tion Oversight Committee 2000). California also has funded a statelvide organization, Women and Girls Against Tobacco, to focus on tobacco product market- ing that targets females. Created in 1992, the organi- zation focuses on empowpering women's and girls' organizations to divest themselves of tobacco indus- tr). sponsorship and funding and on eliminating tobacco ad\-crtising in leading magazines with read- ership among voung ivomen (Women and Girls Against Tobacco-, n.d.). Religious Organizations Although not specifically representative of minority or underserved groups, some religious orga- nizations that have an important impact in minority communities have had long-standing invoivement in issues related to reducing tobacco use. The Interfaith Center on Corporate Responsibility, a coalition of 250 Roman Catholic and Protestant institutional investors, pioneered the corporate responsibility movement in the early 1970s. The value of their combined portfo- lios is estimated at $40 billion. In 1981, the Province of St. Joseph of the Capuchin Order was the first mem- ber of the coalition to file a shareholder resolution with a tobacco company on the issue of smoking and health. Since then, the coalition has filed numerous share- holder resolutions with the major tobacco companies. These resolutions are a unique opportunity to engage in a public dialogue with executives of major tobacco companies; the shareholder meetings frequently re- cei\-e media attention. A more recent effort to involve religious organiza- tions and thereby diversify efforts to reduce tobacco use is the formation of the lnterreligious Coalition on Smoking OR Health. The stated purpose of the group is to mobilize the faith communities in the United States to improve the effectiveness of public policy concerning tobacco. The Coalition is con- cerned with policies affecting United States cor- porations in\,olved in the manufacture and sale of tobacco products. The primary focus of the Coalition is educating policy makers within both the legislative and executive branches of the United States fecleral government (Interreligious Coalition on Smoking OR Health 1993, p. 1). Dakota T he Women vs. Smoking Network, under the aegis of the Advocacy Institute, ~`as a project aimed at informing and uniting women's organ- zations to oppose the tobacco industrv's efforts to market its products specifically to women. In No- vember 1989, the network sent a letter to the editor of more than 100 newspapers nationwide. Several newspapers printed the letter, which responded to a Philip Morris advertisement that had previously run in these newspapers as a mock apology to women for alleged "shortages" of their new ciga- rette, Virginia Slims Super. As a result, several ma- jor national papers and ABC Nc~rjs subsequently ran stories on tobacco ad\rertising that targeted lvomen. Soon thereafter, the controversy and media cover- age surrounding the planned test-marketing of Uptown cigarettes to African Americans began (see the text box "Uptown"). In response, many jour- nalists wrote stories on the related issue of targeted marketing to women. These stories prepared the public for the events that followed. In February 1990, an anonymous source sent the Women vs. Smoking NetLvork copies of confi- dential marketing documents for a new cigarette brand, Dakota. The cigarette, produced by R.J. Reynolds Tobacco Company, was scheduled for test- marketing in April 1990. The marketing documents, entitled "Dakota Field Marketing Concepts," con- sisted of more than 200 pages of test-marketing pro- posals from t\yo different ad\,ertising firms. The marketing documents described Dakota, Mhich \vas The coalition was formed in cooperation Lvitli leading organizations within the mainstream tobacco control community. As of Januarv 1991, the coalition had enlisted 16 main religious orianizations, includ- ing Catholic, Muslim, and Protestant denominations, in the effort to support a large increase in the federal excise tax on a pack of cigarettes (Interreligious Coali- tion on Smoking OR Health 1994). Special Efforts to Reduce Chewing Tobacco Use In 1995, Oral Health America established the National Spit Tobacco Education Program (NSTEP), code-named Project Virile Female, as a cigarette ex- plicitly for young women (18-20 years old). The demographic and psychological profile prepared by Trone Advertising Inc. of the typical Dakota smoker described her as a "Caucasian female, 18-20 years old, with no education beyond high school, work- ing at whatever job she can get" (Butler 1990, p. I, citing Trone Advertising Inc.). She aspired to have an ongoing relationship with a man and "to get married in her early twenties and have a family." She spent her free time "with her boyfriend doing \1-hate\,er he is doing." The marketing documents also included specific promotional strategies to attract young women to the new cigarette. Recognizing the value of the documents, staff of the Advocacy Institute negotiated with the Wnsll- ir2$orr Post for front-page coverage of the story in exchange for initial exclusive release of what the institute staff called "Dakota Papers." The Wad- irlglo,~ Posf ran the story on Saturday, February 17, 1990, tvith the headlin;, "Marketers Target `Virile Female': R.J. Reynolds Plans to Introduce Ciga- rette" (Specter 1990). The Advocacy Institute held back further details on the documents until Tues- day, February 20, so that the director of the Women 1's. Smoking Net\vork could appear on CBS T1lis Morriill:; M'ith Dr. Louis Sullivan, then Secretary of Health and Human Services, to "release" the story of the documents. Secretary Sullivan strongly con- demned R.J. Reynolds' plans to target women in its marketing strategies. fan effort aimed at reducing the use of smokeless to- bacco among youth in sports. Oral Health America teamed up former major league baseball players, such as Joe Garagiola, Hank Aaron, and Bill Tuttle, to help get the message out that smokeless tobacco products arc not a safe alternativ-e to smoking. The components of NSTEP include in-stadium events, public service announcements that have been televised during ma- jor league baseball games, printed materials, and edu- cational videos. An external evaluation of NSTEP is being developed to address all levels of the program and its public health impact. Significant successes of the program include the inclusion of spit tobacco on the national tobacco policy agenda, M'ith specific credit to NSTEP and national