This analysis has 4 principal findings: (1) American immigrants support smoking bans more strongly than their US-born counterparts, with a clear gradient of support across immigrant generation and assimilation measures (e.g. length of stay, citizenship); (2) demographic composition and smoking-policy related context each explain part of the higher immigrant support, although residual higher support among immigrants remained; (3) there was some initial support for the family socialization theory of smoking ban support (since 2nd generation immigrants had higher support than 3rd generation), but it was largely accounted for by differential composition and smoking policy context; (4) Lastly, the US has experienced a substantial secular increase in support for smoking bans over the seven years spanning this analysis among all groups. We address these findings below.
Smokefree policies are workplace protection measures that prevent disease and premature death1,2
. However, certain groups are less likely to be protected by smokefree policies at work including Hispanic immigrants and workers in low-wage jobs33
. Since illegal immigrants, citizen children of illegal immigrants, and refugees are vulnerable populations compared to their counterparts who are US citizens, they may be less willing to advocate for worker-protection policies like workplace smoking bans, for fear of retribution and/or fear of deportation22
. Indeed in the last decade, states have passed hundreds of laws that restrict immigrant rights22
. Public health advocates must therefore protect immigrant health by advocating for smokefree policy on their behalf.
The finding that immigrants to the US consistently are more supportive of banning smoking than their US-born 3rd
generation counterparts suggests that they are potential allies in the tobacco control movement, as partners for coalition building and campaign mobilization. Publicizing their higher support may also protect them from industry manipulation within legislative battles12
The tobacco industry has targeted and exploited vulnerable populations like immigrants and racial minority groups, not only to increase product sales, but also to obstruct passage of smokefree policies; indeed, the industry has a complex understanding of the immigrant market, and has strategized and targeted immigrants for sales and marketing since the 1970’s19
. Profit is the industry’s motivation factor, as this internal document from Brown & Williamson (B&W) articulates:
“Clearly, the sole reason for B&W’s interest in the black and Hispanic communities is the actual and potential sales of B&W products within these communities and the profitability of these sales … this relatively small and often tightly knit [minority] community can work to B&W’s marketing advantage, if exploited properly”34
(Bates No. 531000141–2).
The tobacco industry has moreover cultivated relationships with leaders in minority communities not only to increase their tobacco use, but also for legislative ends: “to advance and defend industry policy positions and to … obstruct tobacco control efforts”20
(p.342). For example, an internal memo enumerates the Hispanic and Asian constituent organizations to which Philip Morris donated $145,655 in 1988. Forty-three percent of these organizations “provided support on legislative issues” including on “restrictive smoking”, “excise tax”, and “Prop 99” (a 1988 California tobacco control ballot initiative) issues35
. In other instances, the industry has compelled organizations to which it has donated to write letters to support its legislative agenda20,21
. Industry has thus forced many groups to oppose smokefree legislative efforts that would likely have improved the health of their constituents had the efforts passed. The tobacco industry has also specifically targeted voters in black and Hispanic communities with mass media campaigns to oppose ETS legislation36,37
. Therefore documenting higher smokefree policy support among immigrants (including among immigrants who are US citizens and therefore voters) provides some protection against tobacco industry manipulation, including making it more difficult for minority constituency groups to lobby on behalf of industry’s positions, against the interest of constituents.
We found that population composition, including race/ethnicity and smoking status, did account for a portion of the higher support that immigrants expressed for smoking bans. Prior discussions of the implications of demographic differences in support for smokefree policies have centered on compositional explanations, including insufficient control for smoking status13
. In immigrant health research, nativity and race/ethnicity are often conflated38
, despite that these constructs are distinct dimensions of inequality that may influence tobacco use and support for tobacco control policy differently. Dispelling population compositional explanations may strengthen deductions from observational studies about the contextual effect of clean-indoor air policies. For example, Gilpin et al. concluded that California exhibited higher support for smoking bans compared to the rest of the US from 1993–1999 after controlling for demographic variables, and attributed the association to the strong tobacco control policy environment cultivated there14
The pattern of stronger policy support among immigrants holds for most countries of origin, and erodes with longer residence in the US. We found a gradient of support by immigrant generation, including that children of immigrants (2nd
generation) were initially more supportive of smoking bans than the 3rd
generation in unadjusted models, as well as by the number of parents who were foreign born. These 2nd generation associations may signal effects of family anti-tobacco socialization e.g. about the unacceptability of smoking39
. Norms and expectations play an important role in conditioning behaviors and attitudes, including those related to smoking, and immigrant kin and community networks have been documented as reinforcing certain behavior-related norms and sanctioning those who defy them40
. Although we found only small residual associations among the 2nd
generation after covariate adjustment, these covariates may have overadjusted for mediators since socioeconomic advancement is an important aspect of immigrant assimilation41
. We lastly do not have the most appropriate data to test family socialization pathways that may be proxied by generation.
We found that immigrants who have remained in the US for a longer period of time held weaker policy support attitudes, after adjustment. This erosion of attitudes with time aligns with other evidence that immigrants adopt the health behaviors or health profile of the US-born with greater assimilation to the US27,42
. Erosion of attitudes may reflect residual effects of tobacco advertising and marketing to immigrants, since tobacco industry has researched, tracked, and targeted immigrants by levels of assimilation19
Americans increased their support for smoking bans from 1995 to 2002, regardless of nativity, subgroup, or venue. These findings align with prior research12,14,43
, and seem to be due to the success among the tobacco control movement to pass local and state-wide laws prohibiting smoking in public places9,10
, the voluntary passage among employers, and the promotion of social norms that smoking is unacceptable44
. Both of these pathways (policies and norms) are important for reducing population rates of smoking. As with seat belt use and drunk driving prevention45
, the passage of laws inhibiting smoking in public areas may increase the level of support for such policies43,46
. However, even though more restrictive smoking policies might cause an increase in attitudes supporting such policies, a certain level of baseline support for policies needs to be present before such passage is politically feasible47
The majority of the US population supports smokefree policy, yet the US does not have national-level laws mandating smokefree environments. Sixteen countries worldwide have passed comprehensive national smokefree policies, including Ireland, New Zealand, and Uruguay31
. Despite current strong support for workplace smoking bans, and despite the documented, well-publicized harmful health effects of secondhand smoke1,2
, inequalities of workplace smokefree policy coverage remain11–13,33
. These inequalities are generated by the U.S. tobacco-control policy patchwork of smoking ban policies at local and state levels, and reliance on voluntary smoking bans that disproportionately benefit white collar workers48
. Mandating smokefree workplaces must be a matter of federal policy to effectively protect the health of all workers and reduce health inequalities.
Strengths and limitations
We used the Current Population Survey, one of the few surveys representative of the U.S. civilian noninstitutionalized population with the power to examine smoking patterns across different American subgroups, with valid and reliable measures on tobacco use, tobacco control, and demographics. We also adjusted for the complex survey design using replicate weights, to report corrected standard errors.
Since this was a repeat-cross sectional study, we cannot deduce that attitudes changed within individuals across time. However, since the CPS is rigorously designed and executed by the Census Bureau, the TUS is representative of the US and state populations, thus our results indicate changes in population attitudes across time. The measurement of the construct of support for tobacco control policies may be underestimated by our method, which measures support for smokefree policies in only 6 venues. Other instruments, such as the Smoking Policy Inventory (SPI) have been developed to capture a range of different domains of tobacco control policy49
. The TUS should consider adopting this measure in future surveys. Lastly, our measure of workplace smokefree policy may be measured with error since it was self-reported. Although we did not account for enforcement, US smokefree policies are largely self-enforcing50
In this study we found that attitudes in support of smokefree policies were stronger for immigrants than for the US-born 3rd generation, and that composition as well as context explained part, but not all, of these patterns. Increasing the support for smokefree laws among all subgroups, and passing comprehensive smokefree policies, are important goals for preventing tobacco-related disease and improving population health.