Evidence suggests that smoking restrictions are accompanied by quitting intentions and behaviors and help denormalize smoking (Albers et al., 2004
; Wakefield et al., 2000
;). We sought to develop a theoretical model and test the influence of smoke-free legislation on adult smokers’ quit intentions through the mediation of normative beliefs of smoking. Our findings are consistent with the Focus Theory of Normative Conduct, which predicts that individuals will conform to a relevant norm, provided it is prominent in their consciousness (Cialdini & Goldstein, 2004
; Cialdini, Kallgren, & Reno, 1990
). In our study, prior to the smoking ban in Scotland, support for a ban significantly heightened smokers’ perceived social unacceptability of smoking and further strengthened these perceptions postban. To the extent that the relevant norm (in this case a nonsmoking directive) is focal or salient, increased unacceptability of smoking would be expected, possibly through media portrayals and peer communication about the implementation of smoke-free laws (Cialdini & Trost, 1998
; Real & Rimal, 2007
). Our results provide evidence that smokers’ perceptions of nonsmoking directives at baseline can transform their smoking norms, which legislation serves to reinforce. These findings support previous research demonstrating a link between approval of bans and perceptions of smoking as less normative (Albers et al., 2007
; Borland, Mullins, Trotter, & White, 1999
; Trotter, Wakefield, & Borland, 2002
), which are propagated in peer networks, through communication about the relevant norm (Perkins, 1997
; Real & Rimal, 2007
Similar findings were obtained for smokers in both Scotland and the rest of the United Kingdom in terms of changes in social unacceptability of smoking. However, social unacceptability of smoking among smokers in Scotland postban was slightly greater than for smokers in the rest of the United Kingdom. Thus, as the normative directive (i.e., smoke-free legislation) became evident in Scotland postban, smokers in Scotland possibly perceived smoking as more socially unacceptable than smokers from the rest of the United Kingdom, which had no ban at the time.
Comparable with the rest of the United Kingdom, in Scotland, perceived social acceptability of smoking at follow-up was associated with higher quit intentions. This finding is consistent with several studies that have found perceived social acceptability of smoking among referent groups to be independently associated with both strength of intention to quit and actual quit behavior at follow-up (de Vries, Mudde, Dijkstra, & Willemsen, 1998
; Dotinga et al., 2005
; Hammond, Fong, Zanna, Thrasher, & Borland, 2006
). Nonetheless, as there was no significant difference in quit intentions between two countries, this association with quit intentions is perhaps indicative of the filtering effects of the Scottish ban coupled with media depictions about the enactment of smoke-free laws in the rest of the United Kingdom, which were implemented in summer 2007. To the extent that a nonsmoking directive is enacted, quit intentions and behaviors will be guided largely by normative considerations, and this will likely impact upon neighboring environments, especially on account of media campaigns and accessibility to and from both settings.
Our study also found that in the rest of the United Kingdom, support for a ban at baseline significantly increased support for a ban at follow-up, although support had no effect on quit intentions at follow-up. That support for a ban at follow-up did not affect quit intentions is perhaps suggestive of the strength of perceived unacceptability to influence quit intentions in a country preparing to introduce a ban rather than support for a ban. Likewise, among smokers in Scotland, support for a ban increased postban, but quit intentions at follow-up were associated with support for a ban at baseline rather than support for a ban at follow-up. This increase in support for a ban, and corresponding association with quit intentions, at follow-up may be partly due to the marked decreases in secondhand smoke evident in Scottish pubs (Semple, Creely, Naji, Miller, & Ayres, 2007
). The variance in quit intentions at follow-up explained in Scotland was greater (19%) than that for the rest of the United Kingdom (12%), but this difference was not significant, similar in this respect to the findings of a recent Scotland/U.K. ITC study that compared smoking cessation indicators and exposure with secondhand smoke in a range of venues between Scotland and the rest of the United Kingdom (Hyland et al., 2009
). These findings reflect the effectiveness of smoking bans, whether implemented or due to be implemented, in influencing quit intentions, and support the view that smoke-free laws should be a public health priority for legislators (Edwards et al., 2008
Like all research, our study has limitations. Our model did not include moderator variables (i.e., general demographic information), which may have impacted upon the findings, although we found no significant gender differences in additional analyses (not reported). The failure to include other potential mediating variables (e.g., reduced opportunity to smoke in the workplace, dislike of smoking outside, associated mass-media campaigns and unpaid media coverage) may have similarly impacted upon the findings, and the absence of other normative constructs weakened the explanatory power of the model. Additionally, a relatively small sample size was employed as a result of the low response rate, and almost half the sample was lost to attrition at follow-up, which is slightly higher than with other research (Albers et al., 2007
). Furthermore, we found significant differences between respondents and nonrespondents in terms of age, with smokers aged 25–54 years more likely to drop out. This may have impacted upon the results, although past research reporting similar response bias in terms of age suggests that this does not affect the conclusions drawn from these studies (Benfante, Reed, MacLean, & Kagan, 1989
; Forthofer, 1983
; Heilbrun, Nomura, & Stemmermann, 1991
Despite these limitations, the use of a longitudinal design allowed us to assess the influence of a population-level policy measure (smoking ban) on quit intentions via a suitable general mediator (unacceptability). Longitudinal designs can overcome many of the problems associated with cross-sectional research and allow causality to be demonstrated, permitting valuable insights into the pathways involved in behavior change. Our findings shed light on how a smoking ban can increase the social unacceptability of smoking, which, in turn, is associated with quit intentions. Future research using tobacco industry perceptions as an additional normative mediator, aside from unacceptability, would be of value to examine whether smoke-free legislation influences quitting partly via the creation of less favorable industry perceptions.