Tobacco Documents Online

436 Journal of Health Politics, Policy and Law PerceM Year Figure 4 The Stigmatization of Cigarettes, 1964-1987. Source: U.S. DHHS 1989. Note: Survey items (percent agree): (1) Cigarette smoking causes lung cancer, (2) cigarette smoking causes heart disease; (3) it is annoying to be near a person who is smoking cigarettes; (4) smoking is hazardous to nonsmokers. Item wording varied. and 1987, support for more draconian policies was considerably weaker. Neither a ban on cigarette sales nor even a ban on smoking in restaurants was ever supported by more than 23 percent of those surveyed. Among the goals articulated most clearly by early antismoking activists was to make smoking so unpopular that smokets would be forced to quit. The data presented in Figure 4 are evidence of how successful the effort has been to stigmatize smokers as harmful to themselves and to others. Translating this stigmatization into public policy consensus has proved more difficult. ~ Public Policies. Stronger local, state, and federal regulation has been among the major goals of the smoking/tobacco control movement (and of the gun control movement as well). Major regulatory achievements at the federal level were noted in Table 2. However, the antismoking move- ment has had its greatest success at the local and state levels. By 1993, over 500 local communities had enacted some form of smoking regu- lation, almost all since 1980 (U.S. DHHS 1993). By 1995, all but ten states regulated smoking in state government work sites; all but twenty restricted smoking in restaurants (CDC 1996a). ao PerceM GO w M ea Nathanson • Social Movements as Catalysts 437 74 ]9 Year 9a Figure 5 Public Support for Legal Restrictions on Cigarettes, 1964- 1987. Source: U.S. DHHS 1989. Note: Survey items (percent agree): (1) Smoking should be allowed in fewer public places; (2) smoking should be limited in restaurants; (3) smoking should be banned in restaurants; (4) cigarette advertising should not be permitted; (5) selling cigarettes should be stopped completely. Items varied. Guns Mortnliry. Gun-related deaths fall into three categories: homicide, sui- cide, and accidents. Homicide and suicide account for approximately equal numbers of deaths (in 1994, 17,527 and 18,765 respectively); accidents are a distant third (1,356) (Anderson, Kochanek, and Murphy 1997). Of these categories, only accidental deaths have shovm a consistent decline over several decades (Spitzer 1995: 74). Since the early 1960s, age-adjusted death rates for firearm homicides have doubled and such rates have increased by half for firearm suicides (Karlson and Hargarten 1997: 1).'Ihe age-adjusted homicide firearm death rate declined slightly in 1994 (and the crude rate has continued to decline, along with the total homicide rate). Half of all homicide victims are minorities (double their representation in the total population); the highest rate is among African Americans: "For African Americans, the risk of being killed in a homicide is nearly seven times greater than for whites and more than twice as great as for Hispanics" (Karlson and Hargarten 1997: 6). Firearms are the leading cause of death for adolescent black males. 9£SOZBZLOZ

434 Journal of Health Politta, Policy and taw sp Percent 10 1966 1970 1974 1979 1984 1989 1994 Year Figure 2 Percent Current Smokers by Education, 1966-1995. Sources: Giovino et al. 1994; CDC 1994, 1996b, 1997. in 1994 was about the same as the percentage of the highly educated who smoked in 1966. Furthermore, absence of recent change-even increase -in the percentage of smokers is most evident among the least well edu- cated. In contrast to its democratic pattern in the 1960s, smoking is now concentrated in the lower socioeconomic strataJs A third basis for cau- tion, one that has received a great deal more public attention, are recent increases in the percentage of smokers among high school students. These data are presented in Figure 3. Smoking prevalence among high' school seniors declined steadily until the mid-1980s, rose and fell errati- cally until around 1992, and since then has sharply increased. Attitudes and Beliefs. Attitudes and beliefs about cigarettes and guns may be compared using two dimensions: beliefs about the degree and nature of the danger that these products represent, and responses to that danger 15. Unfortunately, current smoking prevalence date arc grouped by educational level rather than by occupation or income (with theezception of poveny level statua, which is a fairly crude income index). Data on smoking prevatence by occupetion in 1955 published by the National l:ent¢r for Health Statistics show remarkably tiuie variation in prevalence by occnpation: among men, between 67 and 70 penxnt of white collar workers and between 69 and 74 percent tfblue co))arworkers reponpdcurrentd regular, smoking. A much smalkr penxntage of women smoked, but there waa an equal absence of variation by occupation (U.S. DH9W 1956). Nathanson • Soclal Movements as Catalysts 435 Figure 3 Percent Current Smokers among High School Seniors, 1976-1995. Source: Giovino et al. 1994. in the form of support for or opposition to legal restrictions on their use.l6 Data on changes over the past several decades in attitudes and beliefs about cigarettes are presented in Figures 4 and 5. Two dimensions of belief about the danger of cigarettes are included in Figure 4: first, the belief that cigarette smoking causes harm to the smoker, and second, the belief that cigarette smoking causes harm to (or at the very least) incommodes individuals in the smoker's environment. The former belief stigmatizes the cigarette; the latter belief stigmatizes the smoker as well. All of these beliefs increased markedly during the period of observation. Particularly notable is the near doubling between 1974 and 1987 (from 46 percent to 81 percent) of the proportion of the survey sample who believed that smoking is hazardous to the health of nonsmokers. Public support for legal restrictions on smoking, advertising, and sales of cigarettes is described in Figure 5. While support for limitations on where people could smoke clearly increased over the period between 1964 16.'tlsese are net, of courx, the only relevant attitudes and beliefs. They are, howeverr the ones for which data are av®itabk. In evatuating the data to be prescntcd in this section, it is important to be aware that the exact wording of the questions about cigarettes varied in the dif- ferent surveys carried out by differentorganieations. Some ofthe apparent variation in aa0udea over titro may be explsined by these diffrrenccs.'Ihedata on guns arc frmn avrveys mrdu4ed by the Gallup organization and NORC, using the satne question wording. 5080ZSZLOZ

422 Journal of Hea@h Politics, Policy and Law ments have received relatively little attention from social movement scholars. This article employs an analytic framework drawn from social movement and related sociological theories to account for the relative success of the United States' campaigns against smoking and guns in bringing about change in the policies and behaviors targeted by these movements. While the article's primary focus is on the smoking/tobacco control movement, the analytic strategy is comparative. Aspects of the move- ment that may account for its relative success (as well as for the pitfalls it may still face) are best identified through comparison with the experi- ence of another contemporary health-related social movement. Gun con- trol was selected because it has certain initial similarities with smoking/ tobacco control. Cigarettes and guns were both widely used "democratic" consumer products accessible to and enjoyed by millions in all walks of life, romanticized in film and (in the case of cigarettes) advertising. Movements against these products were initiated in the late 1960s and early 1970s. Both movements confronted well-connected, well-financed opponents. Based on a detailed comparison of the two campaigns I argue that the success of the smoking/tobacco control movement may be accounted for by an ideologically persuasive construction of the relevant health risks, by grassroots mobilization for nonsmokes' rights, and-in the end-by important weaknesses in the movement's opposition. Sources for the analysis presented include interviews with movement activists and observers, participant-observation in movement-related activities, archival materials, and published books and articles by advo- cates, journalists, and scholars.' Research was largely completed by the end of 1996, but some reference is made to later events. The choice and definition of variables to be examined as well as the analysis of research materials were guided by two bodies of theory: social movement theory as elaborated by sociologists and political scientists (e,g., Gamson 1990; Tarrow, 1994; Kriesi 1995; McAdam, McCarthy, and Zaid 1996) and work on the social construction of public problems and per- ceptions of risk by Joseph R. Gusfield (1981), Mary Douglas and Aaron Wildavsky (1982), and others (Kunreuther and Linnerooth 1983; Wynne 1987). Where quantitative measures of particular variables were avail- able and appropriate, they have been used. However, for the most part the method used was qualitative and comparative (see, e.g., Glaser and Strauss 1967; Lofiand 1996). t. A detailed descdpuon of these saurre.c is available from the author. Nathanson • Social Movements as Catalysts 423 The analytic framework employed includes three sets of variables. These are, first, movements' supporting ideologies; second, each move- ment's capacities for mobilizing potential constituencies and organiza- tional resources; and, third, political opportunity structures, defined to include a broad range of opportunities and threats external to the move- ments themselves. Each variable is more fully described in the context of analysis of the movements themselves. The article is organized in four sections. The first section consists of brief historical overviews of each movement. Measures of movement success are described and then the movements' reiative success by these criteria are compared in the second section. The third section is devoted to a comparative analysis of the two movements, and in the fourth and concluding section I reflect on what can be learned from this analysis. Historical Overview Movements have histories. The content and relative importance of each of the elements I have identified-ideologies, organization, and political opportunities-shift over time. Aspects that appear highly important at an early stage of movement evolution may become much less important and, indeed, change their character altogether as the movement either declines or becomes institutionalized. Political opportunities, in particu- lar, are subject to marked change as elites more or less friendly to the movement's cause gain or lose power, What Charles Tilly (1986) has labeled "repertoires of contention"-the sit-in, the protest march, the courtroom battle-shift over time in popularity and in value as vehicles of expression and influence. Particularly in the case of movements that attract substantial media attention, immediate drama (or what the media defines as drama) tends to obscure the complex reality of how change comes about. This complexity is illuminated by close attention to each movement's history. Onl ll January 1964, less than two months after President John F. Kennedy's assassination, the Commission on Smoking and Health that Kennedy had appointed two years earlier issued its report. The assassi- nation and the report each generated substantial pressure for change toward greater control, respectively, of guns and smoking, and produced corresponding opposition to the changes proposed. Here I present a sum- mary and highly selective account of these events. 6Z80Z9ZLOZ

488 Journal of Health Politics, Policy and Law Research Could Endanger Public Safety. Ameican Journal of Public Health 1 87(6):918-921. Wills, G. 1995. To Keep and Bear Arms. New York Review of Books, 21 September, 62-73. -.1995. To Keep and Bear Arms: An Exchange. NewYork Review ofBooks, 16 November, 61-63. Walinskg H., and T. Bmne. 1994. The Serpenr on the Staff. New York: Putnam. Wynder, E. L., and E. A. Graham. 1950. Tobacco Smoking as a Possible Etiologic Factor in Bronchiogenic Carcinoma. A Study of 684 Proved Cases. Journal of the American Medical Association 143(4):329-336. Wynne, B. 1987. Risk Managemenr and Hazardous Waste: Implementation and the Dialectics of Credibiliry. Berlin: Springer Verlag. Zimring, E B., and G. Hawkins. 1987. The Citizen's Guide to Gun Control. New York: Macmillan. The Missing Millions: Organized Labor, Business, and the Defeat of Clinton's Health Security Act Marie Gottschalk University of Pennsylvania Abstract During the battle over comprehensive health care reform in the early 1990s, organized labor was not only unable to put together a winning coalition but also found itself divided and on the defensive as it struggled to prevent any further erosion of the private-sector safety net of the U.S. welfare state. Labor's mlative incf- fectiveness has deep institutional and political roots and was not merely a con.ce- quence of its dwindling membership base. Several key institutions of the private wel- fare state, notably the Taft-Hartley health and welfare funds and the Employtnent Retirement Income Sccuriry Act (ERISA) preemption, brought the interests of orga- nized labor more closely in line with thuse of large employers and commercial insur- ers and aggravated divisions within organized labor and between unions and public interest groups. In addition, several political factors conspired to reinforce labor's tendency to stick to a policy path on health care issues that was predicated on an employer-mandate solution and that had been charted primarily by business and lead- ing Democrats. As a result, organized labor did not emerge from 1he 1993-1994 struggle with its political base fortified nor with a viable long-term political strategy to achieve universal health care and to shift the political debate over health policy in a more desirable direcdon. In the 1993-1994 struggle over health care reform, organized labor proved to be politically ineffective. Unions, which have played a critical part in the passage of a number of major pieces of social welfare legis- lation, notably Medicare (Matmor 1973; Derthick 1979), were not only unable to put together a winning coalition in the early 1990s, but also I am gateful to 6avd Cameron, David M.yhew, Cethte lo Martln, Rogers Smith, Mark Peter- snn, Srephen Skowmnek, and an anonymous reviewerforlheir helpful commcnts on earlier ver- sioas of thls anicle and to Eric Lonazolf for his indispensable research assistanee Journal uf 8ealrh Fbliricr, Pollcy and Jnw.Vol.1d. No. 3, /une 1999. Copyright ©1999 by Duke tlnivershy Press. 2980Z5ZGOZ

442 lournal of Health Politics, Policy and Law Comparative Analysis of the Smoking/ Tobacco and Gun Control Movements Constructions of Risk Guns and cigarettes are hardly newcomers to the American scene. Spit- zer describes "the long-term sentimental attachment of many Ameri- cans to the gun"grounded in cultural myths about the role of the gun in the struggle for independence and the taming of the frontier (1995: 8). During the early part of the twentieth century cigarettes became power- ful symbols of sexuality, power, autonomy, and modernity (Brandt 1992). The two symbolic traditions merge in the Marlboro man, the fiercely independent, (possibly) gun-mting cowboy smoker. Mobilization against these powerful cultural symbols demanded that their meanings be transformed so as to legitimate and guide protest activity (McAdam 1994: 37). Central to this transformation has been the construction of credible risks. Dangers abound. Whether or not these dangers rise to a commu- nity's threshold of awareness and become defined as risks to the public's health, much less elicit responsive action, depends on the active inter- vention of human agency, on what I have called "constructions of risk" (Nathanson 1996:614-615). The first element essential to this construc- tion is the existence of groups or individuals with the authority to define and describe the danger that threatens. The second element is the asser- tion of a causal framework to account for the danger. The Making of a Credible Threat Snroking. The consequences of smoking for health and morality have been argued at least since James the First wrote his famous polemic in 1604 (I{Iuger 1996: 15). Progressive Era antivice campaigns encom- passed cigarette smoking, along with drinking and sexual adventurism, as immoral, unhealthful, and a corrupter of youth. States passed restric- tive legislation, setting the age at which cigarettes might be legally pur- chased, or in some states, prohibiting the sale of cigarettes altogether. The Progressive Era antismoking movement petered out in the 1930s, along with the other campaigns against vice, only to reappear in new guises more suitable to the tenor of the times. The late-twent9eth-century construction of cigarette smoking as a cred- ible risk was the work of the Surgeon General's Advisory Committee on 6£80ZSZL0Z i Nathanson - Sociaf Movements as Catalysts 443 Smoking and Health.m The committee brought together the large body of existing evidence that cigarette smoking was a danger to human health, summarized that evidence in a scholarly fashion, and put the imprimatur of the federal government on its central conclusion, that "cigarette smok- ing is causally related to lung cancer in men" (U.S. DHEW 1964: 31, emphasis mine). The committee did not collect any new data or make any new discoveries. Its report did. nevertheless, construct new knowledge. It transformed scattered evidence into official authoritative knowledge, pro- viding "power and legitimacy to the epidemiologic findings" (Brandt 1992: 67).'s Gusfield argues that public consensus on the dangers of smoking "represents the hegemony of medical science over the culture of health: ,.. by the time the Surgeon General issued the report of 1964, social conditions had become favorable to the transmission and credibility of medical science and the position of the federal government as a source of authoritative advice and activity in the promotion of health" (1994: 54-55). Perhaps. Medical authorities have, indeed, made substantial inroads as arbiters of personal conduct; the latest scientific reports are carried on television and the public is highly attentive (but substan- tially more skeptical, I would areue, than it was in 1964).24 Neverthe- less, the smoking story is more complex than Gusfield allows. The federal government did not embrace ownership of the smoking and health issue in the 1950s and 1960s. A reluctant executive branch had ownership thrust upon it by the efforts of the health voluntaries (principally by the American Cancer Society [ACS]) and by a small minor ity of dissident members of Congress who believed that the authoritative voice of the federal government was essential to confer legitimacy on their cause. Further, while Congress in 1965-with its mandate for annual reports on the health consequences of smoking-awarded continuing ownership to the Surgeon General, it was a limited-purpose ownership. The Surgeon General was given authority to define and describe the problem-to assert the risks of smoking-but very little power to limit 22. The seleceion pmcess and the background leading up to the commiua's fomulion are descdbed in many other sources and will not he repeated here (U.S. DUEW 1964: U.S. DHHS 1999; Brandt 1992; Kluger 1996). 23. An important aspect of the Surgeon General's report, as Brandt points out. was that it markedthe beginning of s new, rnte for epideniologists in the construction of risk (1992: 67). 24.InJuly 1995, thejoumal Srientr published an article with the titk,'Eptdemiology Faces tts Limits;'that began, •71te news about health risks comes thick and fast these days, and it scems dnwst constitutionally contradinory" (7hubes 1995). Science's primary cusxzm was the public, not the sciennfic response to a dtluge of "mntrzdictory advice:' ~. .

460 Journal of Health Politics, Policy and Law bunch of chapters going we organized all the names we had by states and localities and mailed those names out to local Lung Associations and local GASP groups around the country, saying these are people who have written to us from your area. Contact them, get them active in your group. ... We actually sent them envelopes withjust stacks of little mail- ing labels" (Gouin 1995). Although Gouin mailed the first issue of the newsletter to members of Congress and federal officials were aware of GASP's existence, the group's initial tactics were almost entirely local and on a small scale, focused on getting smoke-free meeting rooms (particu- larly meeting rooms of obvious groups, like environmental groups and the Lung Association itself), doctors' offices, hospitals, natural food stores, and the like. Funding requirements for these activities were minimal (Gouin's budget never went above $10,000 per year) and came from con- tributions and from the pockets of the organizers themselves. Although GASP stmck a responsive chord in some quarters, non- smokers' rights were by no means immediately popular with the gen- eral public: "When I started getting up there an my soapbox, it was a very unpopular issue... they thought who was this crackpot telling us we shouldn't be smoking in public" (Gouin 1995). The media, never- theless, were very interested and "we got a lot of free publicity;" at first locally and then nationally and internationally. Exposure brought new recmits. By the mid-1970s, the focus of nonsmokers' rights activists began to shift from "passing out leaflets and buttons" to the passage of state and local antismoking regulations (Hanauer, Barr, and Glantz 1986: 2). Sev- enteen of the fifty-four GASP groups listed in 1974 were in California. These groups incorporated as California GASP in 1976 and focused increasingly on regulatory action, first at the state and later (after narrow defeats in 1978 and 1990) at the local level.m These activities brought additional groups and individuals into the movement: the Califomiachap- ter of the ACS in particular played "a very crucial role" in the initiative campaigns (interview with smoking/tobacco control activist, 1995). Data presented in Figure 8 on the timing of grassmots organization in relation to the passage of local restaurant nonsmoking ordinances (the first type of 56. In 1981, Califomia GASP became Califomians for Nonsmok<ce Rights and in 1986. Ameriexns for Nonsmokers' Rights (ANR). In an anviversary issue% the ANR newslelter com- mented on the shin from srate to local priaitier.'The experiercx uf hsving bas buvel twlce at Ihe stat<level whik succeeding locally iin &rkeley in 1977 and in Los AIIOS in 19"19] taught the nonsmokere righls sctivists a great deal. They learned that the tnbacco fntluslry, while wealthier and more powerfut at the stxte level. just wasn't ahte to thwan our cRoes lacatly" (ANR 1996). 8ti80Z9Z10Z Nathanson - Social Movements as Catalysts 461 No. Groups -oroaw -ad.nea. Figure 8 Grassroots Organizations and Nonsmoking Ordinances, 1964- 1992: Number of Nonsmokers' Rights Groups and Passage of Local Restau- rant Nonsmoking Ordinances. Sources: U.S. DHHS, NIH 1993; U.S. DHHS 1990. ordinance to be widely adopted) suggest that organization at the tocat level played a critical role in subsequent regulatory action. There are no published surveys of nonsmokers' rights activists, How- ever, some idea of the population groups to whom their activities had most appeal can be gained from an analysis of the relationship between the average family income of Califomia counties and the timing of smok- ing control otdinauce adoption. These data are shown in Figure 9. Counties that adopted smoking control ordinances in 1985 or before are defined as early adopters, counties that adopted ordinances after 1985 are late adopters, and the remaining counties are nonadopters. Wealthiercounties are substantially more likely to be early adopters: the difference in mean income levels across the three adopter groups is highly significant. It is a reasonable inference from these data that individuals active in the non- snickers' rights movement were also likely to be above average in edu- cation and wealth. As the nature and scope of their activities changed, funding require- ments increased and GASP groups became somewhat more profession- alized (although they were still largely voluntary); drafting legislation and engineering its passage demanded legal experience and organizing a No. (hdinances

432 Journal of Health Politia, Potky and Law (1995: 281), while Edwin Amenta and colleagues argue that achievement of "new advantages" not anticipated by the movement should also count as success (and that new advantages are more "meaningful" than accep- tance as a measure of success) (1994: 681). An additional problem arises in the case of ongoing social movements: how successful the movement appears depends on at what point in the movement's trajectory (unknow- able, except in retrospect) success is measured The long-range goals of the smoking/tobacco and gun control move- menrs are to bring about a decline in the relevant parameters of mortality and morbidity. In the short run, both movements advocate behavior change in, respectively, patterns of cigarette smoking and in gun ownership and use. Beyond these generalities, smoking/tobacco control movement repre- sentatives have from the beginning been quite explicit in their aims of stig- matizing cigarettes and the smoker and, more recently (in the words of one activist), of "gening the bully [the tobacco industry] off the block." There is less consensus among gun control advocates, in part because of dis- agreement on the meaning of the word control. In the following paragraphs I evaluate the extent to which each movement has achieved its goals.14 Smoking Mortalitk On 14 November 1996, the New York Times reported an over- all decline in cancer death rates "for the first time since 191p" (Brody 1996). Philip Cole and Brad Rodu (1996), whose work was the basis for the Times' report, attributed about half of the reduction in cancer mor- tality to declines in smoking since 1965 and anticipate continued reduc- tion "as the now rising lung carcinoma mortality rates among women stabilize and then decline," a consequence, presumably, of declines in women's smoking. Their work was foreshadowed in several earlier reports. Morbidity and Mortality Weekly Reporr (MMWR) noted in November 1993 that "the declines in smoking prevalences have resulted in a stabi- lization or decline in the lung cancer death mte for men aged <55 years and for women aged <45 years, respectively" (CDC 1993). Cardiovas- 14. The pmbtem ofoutcwnx measvrcmcnt is particulariy tmuhlesome in ihe caseof gun con- trot for several reasors. Fint, thae is disagrrernent among advocates about the movemeni s goals; second, the involvement of the public health communiry is relatively recent; and third, the movcment itsetf is a moving target, evolving in several directions simuhaneously. In addi- tion, measures of soccess have been setccted, in part, for their comparability across the two muvenmts. As a rewlt, some rrteasures thatmight work to the detriment ofthe smokinyfaUaxo enntrut muvement-for exarnple, suecess in achieving tobacco raa inereaxs or In curbing cig- areite advertising-heve not been used. Nathanson • Social Movements as Catalysts 433 f9e4 1989 1974 1979 1984 Y.w 1999 1994 amm. +an,w«. ~w~al Figure 1 Smoking Behavior Change, 1964-1995. Sources. Giovino et al. 1994; USDA 1995-96, 1998; CDC 1996b, 1997. cular disease (CVD) is responsible for nearly twice as many deaths as lung cancer and it is clear that cigarette smoking plays a substantial role in CVD mortality. Cardiovascular disease mortality, however, has been declining for the past three decades among both women and men and the relative part in this decline played by changes in smoking behavior is dif- ficult to determine with precision (see, e.g., Hunink et al. 1997). Behavioc Change over the last three decades in the behaviors targeted by tobacco advocates are presented in Figure 1. Two measures of change are employed: cigarette consumption per capita per year (in numbers of cigarettes) and smoking prevalence (percent of the population age eigh- teen and over who are current smokers). Both cigarette consumption and the reported prevalence of smoking have declined substantially since 1964. Despite these declines, there are a number of reasons for caution in evaluating changes in cigarette smok- ing. First, both of the aggregate measures employed in Figure 1 have recently plateaued. Second, as shown in Figure 2, by far the sharpest declines in cigarette smoking have occurred in that proportion of the population with a college or graduate school education. Among individ- uals with less than a high school education, the percentage who smoked V£80Z9ZL0Z

430 Journal of Health Politics, Policy and Law the NRA and its congressional supporters, who began work to overturn the act as soon as it had passed (Spitzer 1995: 147). Their efforts culmi- nated in the Firearms Owners Protection Act of 1986, also known as the McClure-Volkmer Act after its principal House and Senate sponsors. As enacted, the bill eased a number of provisions of the 1968 act but retained the ban on interstate sale (but not transport) of handguns. Despite the lat- ter concession to gun control advocates, the McClure-Volkmer Act is regarded by most observers as a victory for the NRA (Davidson 1993; Spitzer 1995). A significant consequence of the bill, however, was that it "solidified the split between the NRA and police organizations" (Spitzer 1995: 151). McClure-Volkmer had been actively opposed by a newly orga- nized coalition of police groups. After a fallow period of over twenty-five years, two pieces of federal legislation strengthening gun restrictions were enacted in quick succes- sion in 1993 and 1994. Neither of these measures was particularly dra- conian; both include exemptions and other provisions favored by gun supporters. Nevertheless, both took over five years to pass from the time of their first introduction in Congress and both were the subject of vim- lent debate between supporters and opponents. Emergence of the Gun Control Movement. Josh Sugarmann reports that gun control became an issue for some women's groups in the early 1930s (1992: 29). However, there is little evidence of an organized gun control move- ment in the United States before the mid-I970s. The National Coalition to Ban Handguns (NCBH) (renamed the Coalition to Stop Gun Violence in 1990) evolved out of church groups' response to the assassination of Presi- dent John F. Kennedy. Following a period of dormancy between 1968 and 1974, the group reorganized in 1975 as the NCBH. Handgun Control, Inc. (HCI) was founded at about the same time by a Republican businessman whose son had been shot and killed with a handgun (Spitzer 1995: 115). Beginning in the early 1980s, the NRA found itself in a series of high- profile political conflicts with the police around the regulation of "cop killer bullets" and plastic guns and around the McClure-Volkmer bill. 13 Important outcomes of these conflicts were independent mobilization by the police to advance their own interests in gun control and a political realignment of sorts, with the Democrats on the side of the police against the Republicans and the NRA. A final important development in this brief history is the redefinition 13. These conHicts are colurfuny described by Osha Gray Davidson (1993:85-127). Nathanson • Social Movements as Catalysts 431 of guns and gun control as issues in the domain of public health as well as, or in addition to, the domains of crime and law enforcement. Among the proponents of this definition have been the federal Centers for Dis- ease Control and Prevention (CDC), which established a violence pre- vention unit in the mid-1980s, and former Surgeon General C. Everett Koop (Koop and Lundberg 1992). Medicalization of the gun question has important consequences for how the problem is characterized and into whose province it falls. This redefmition has triggered a new wave of gun control organization, and in the last few years several gun control groups led by health professionals and by lawyers with training in public health have emerged. Not surprisingly, ntedicalization of guns has been heavily contested by the NRA and by some members of Congress. Measures of Movement Success While scholars' interest in social movements "stems from their belief that movements represent an important force for social change" (McAdam, McCarthy, and Zald, cited in Burstein, Einwohner, and Hollander 1995: 275), the measurement of change and its causal attribution present con- siderable methodological problems (see, e.g., Diani 1997). For example (following Diani), the adoption of arestaurant smoking ban may be due to nonsmokers' rights activism; the activism may have been generated by politicians in support of the ban; or both activism and the ban may result from other social forces (e.g., increased consumer sensitivity to health threats and accompanying media hyperbole around these threats). The problems of causal attribution are somewhat alleviated when (as in the present case) attention is narrowly focused on specific campaigns and the time span over which movements are observed is fairly long. Neverthe- less, while I will argue that nonsmokers' rights activism played a causal role in bringing about change in smoking-related policies and behaviors in the United States, this argument does not preclude the existence of addi- tional causes and causal paths that I may not have fully considered. Conceptualizations of movement success generally begin with William Gamson s two-dimensional criteria of "new advantages" for its constituency and/or "acceptance" of the movement by its targets or by the public as the legitimate spokesperson forthe interests itrepresents (1990:31-34; e.g., Amenta, Dunleavy, and Bernstein 1994; Burstein, Einwohner, and Hollander 1995). Beyond this reference point, there is little consensus. For example, Paul Burstein and colleagues state that "assessing a move- ment's success involves determining whether it has achieved its goals" ££80ZSZLOZ

446 Journal of Health Politics, Policy and Law gered and these cultural presuppositions are remarkably consistent over time and place (Douglas and Wildavsky 1982; Douglas 1992; Dobbin 1994; Nathanson 1996). In the U.S., we concem ourselves with dangers to the individual, not (e.g., as in France) to the state. Furthermore, the dimensions of risk are highly predictable. In any given case, risks may be portrayed as acquired deliberately or involuntarily (and the victims as correspondingly culpable or innocent), as universal (putting us al] at risk) or as particular (only putting them at risk), as arising from within the individual or from the environment, as visible or invisible.n The most acceptable risks are universal, are attributable to the external environ- ment, and are incurred involuntarily by innocent victims. Smoking. The remarkable transformation of the cigarette and smoking from symbols of "modernity, autonomy, power, and sexuality" to sym- bols of weakness, irrationality, and addiction (Brandt 1992: 70) was accompanied and driven by shifts in underlying constructions of risk. The initial constmction of the cigarette as a danger to the health of the male smoker did not disappear, but has been added to and elaborated ever time. Throughout the period in question, advocates for the dangers of smoking were aided and abetted by a highly interested and attentive news media. During the first phase of the tobacco wars, the messages conveyed to the U.S. public were that the hazards of smoking are attributable to the risky choice of the person who smokes and that the individual is respon- sible for risk reduction by making the necessary changes in his or her behavior. The most consistent advocate of this perspective, the American Cancer Society, was also the dominant player in the early stages of the smoking and health drama. The following is from the society's 1957 annual report: "The society believes that at our present state of knowl- edge, the question of whether to start smoking or to give it up must be left to the judgment of individuals. For intelligent decisions everyone should know the facts: There is a definite association between cigarette smoking and cancer" (19). The ACS approach is striking in two respects: first, in its validation of individual choice and, second, in the limitation of its own responsibility to 'just the facts:' Knowledge of the association between cigarette smoking and cancer as well as other diseases accumulated, but the posture of the ACS as 27. Social movement mtrepreneurs (and other interesred actors) pomuy risks in what they ktrow to be culturally resonant terms, thereby reinforcing, in chcular fashion, preexiating eu4 tural preconceptions about the relevant dimensions of dsk. Nathanson . Social Movements as Catalysts 447 articulated in the cited quotations did not, in fact, begin to change until the early 1980s. Lest it be thought that the ACS was unique in its indi- vidualization of the smoking and health issue or in the timidity of its rec- ommendations, federal health officials took much the same tack. Their approach to the preparation of the Surgeon General's report was that of an individual physician advising an individual patient: "What do we [that is, the Surgeon General of the United States Public Health Setvice] advise our patient, the American public, about smoking?" (cited in Brandt 1992: 66). Implicit in this question is not only, as Allan Brandt points out, "a particular model of public health and the role of the state," but also a particular model of where the hazards of smoking are socially located. Themes of personal responsibility for health (and the limits of govern- ment intervention) were echoed in influential publications throughout the 1970s: cigarette smoking became the quintessential exemplar of lifestyle change within the individual's control (Lalonde 1974; Knowles 1977; U.S. DHEW 1979).22 There is a striking disjunction between the Surgeon General's 1964 proclamation of smoking as "a health hazard of sufficient importance in the United States to warrant appropriate remedial action"(U.S. DHEW 1964: 33) and the actions that were, in fact, taken by the major players at the time. While there is a strong libertarian bias in Americans' approach to health protection, early constructions of smoking and health were dri- ven as much by political as by philosophical concerns. In an interview, an ACS official who had worked in the national office from 1960 through 1990 described the society's internal struggles: I was there when the great debates were held on how far the American Cancer Society should go as an organization in taking up an antito- bacco position. And you could well realize that there was tremendous resistance within the American Cancer Society in the late 1950s and early 1960s, because here you have tobacco-growing states and here you have divisions, North Carolina, South Carolina, Kentucky, Ten- nessee, that said, "You are going to destroy us. We are not going to be 28. Bssa7 on quamhative analysis of stories in a range of inedia (major newspapers, televi- sioq news megazines) in 1960 and 1984, Elcanor Singcr md Phyllis M Endreny report that "stories about akuhol and tobaccu diaproportionately blamed vittims for risks associated with these hazuds. In the case of tobacco, victims alsn appear to be disproportionately held respon- sible forprevention., „7hus,judging from theevidence ofthese stories, smohingwas seen pri- ntarily ac an aclivity within the individual's eontmL whereas prevention of thc risks of drink- ing was acen, in the majority of stories, as requiring govertmcn intervention thmugh the imposition of laws and the like' (1993: 117). I 6ti80Z9ZLOZ I

4" Journal of Health Politics, Policy and Law the gun control movement received a major boost when police organiza- tions came to its support in the mid-1980s. Interest in gun control among the larger population is driven, in Spitzer's terms, by a "cycle of outrage, action, and reaction usually [begin- ning] with the sensational and the horrific" (1995: 13). This point was eloquently made by a gun control activist we interviewed: From 1975 [when his group was founded] to 1980 we were a very small, basically inconsequential organization. We werejust a voice out there in the wilderness.... In 1980-81, the Pope, John Lennon, and the president were all shot.... Suddenly, the issue became hot. We were one of the few sources that people had to go to for information, especially the news media. So in 1980, we grew rapidly. And then, it died. The president took that "aw shucks, it didn't hurt" approach, the Pope didti t say anything about the issue, and all we were left with were John Lennon fans. The other folks disappeared.60 The problematic character of its recruitment base may be responsible for what another movement activist described as "probably the greatest short- coming of the gun control movement: even though we have strong pub- lic suppott, we have no organized grassroots." Grassroots support for the gun control movement has been sporadic and difficult to sustain, and the movement has only recently begun to attract support from established organizations comparable to that of the health voluntaries for smoking/tobacco control. Further, while both move- ments have experienced and continue to experience conflict between more conservative and more radical factions, the impact of this conflict on the movement's ability to present a united front may be more serious in the case of gun co¢trol.61 In the mid-1990s, the national stage was shared by two gun control organizations, Handgun Control, Inc. (HCt) and the Coalition to Stop Gun Violence (CSGV). Spitzer gives a concise summary of the groups' background, beginning with HCL' "[HCI] began in partnership with the National Coalition to Ban Handguns (NCBH), formed at about the same 60. The ^cyek" this activist descritxs eppears ckady in daU from ouranalysfs of NewYnrk Times cuverage of the gun control issue.7Te number of anklesjumped frqn fony-two in 1980 to sixty-twu in 1981 fallowing the Rcagan and aher ahootings; i[ dnpped to betwern sixtcen and thirty-nine in the petiod fmm 1982 to 19Ne,JUmped again to seventy-six in 1989 following the Smekton, California schonlyard shooting, and dropped again to forty-six in 1990. 61. Recaitly, however, the prospenive tobacco settlement led to iniense conflict within the smaking/lobeccoeonttul nm¢ment between propptents who belieVed it was the best deel they could get and opponents who believed it was a sellout. 0980Z5ZL0Z Nathanson • Social Movements as Catalysts 465 time. The groups soon parted ways. The NCBH was renamed the Coali- tion to Stop Gun Violence in 1990. It has generally pursued a tougher stand on gun regulation than HCI, and has been overshadowed by HCI's greater size and visibility, especially since Sarah Brady has become a prominent HCI figure" (1995: 115). Only two state-level gun control advocacy groups were identified by our respondents, in Maryland and Illinois. One activist described most other organizations as consisting of "dedicated people with a hunk of lenerhead:' A range of low-profile groups do exist, however. Missing Peace, for example, is a primarily educational rather than political advocacy group (although it has employed direct action occasionally, e.g., a candlelight vigil in front of the White House). The group was founded by a suburban Maryland woman after a twelve-year-old brought a gun to her son's school; its director would like to expand nationally and pointed to affili- ates in eighteen states. At the same time, she suggested that she might quit if additional funds were not forthcoming soon. Missing Peace has much in common with the original GASP: it is small, with meager funds, and is staffed almost entirely by volunteers. A major (and, I believe, crit- ical) difference is the absence of anything comparable to the American Lung Association's support for GASP. A standard question in the interviews we conducted asked respondents to comment on the relationship of their own group with other groups working on the gun control issue. The level of perceived competition, even conflict, elicited by this question was striking. Reactions ranged from strongly negative-extending to an unwillingness to share mailing lists and to uneasiness about appearances on the same platform-to expressions of competition over financial resources, to more thoughtful analyses: There are a couple of splits in the gun control movement. One is between the controllers and the banners. The controllers who want to control handguns through licensing, registration, things like that, tend to view the banners as radicals who give them a bad name. The con- trollers also tend to see gun violence as a crime issue versus gun vio- lence as a public health issue. The banners, we're talking handguns here, tend to be public health people, those who take a public health perspective and one of the problems with the public health argument is that it points out many of the limitations and internal flaws of the gun control's control argument! Which doesn't lead to good relations. I think that's one of the problems in this movement in that whereas in