Abstract

Background
There is substantial empirical evidence on the benefits of smoking bans; however, the unintended consequences of this anti-smoking measure have received little attention. This paper examines whether workplace smoking bans (WSB's) are associated with higher self-perceived, work-related stress among smoking workers.

Methods
A longitudinal representative sample of 3,237 individuals from the Canadian National Population Health Survey from 2000 to 2008 is used. Work-related stress is derived from a 12-item job questionnaire. Two categories of WSB's, full and partial, are included in the analysis, with no ban being the reference category. Analysis also controls for individual socio-demographic characteristics, health status, provincial and occupational fixed-effects. We use fixed-effects linear regression to control for individual time-invariant confounders, both measured and unmeasured, which can affect the relationship between WSB's and work-related stress. To examine the heterogeneous effects of WSB's, the analysis is stratified by gender and age. We check the robustness of our results by re-estimating the baseline specification with the addition of different control variables and a separate analysis for non-smokers.

Results
Multivariate analysis reveals a positive and statistically significant association between full (β = 0.75, CI = 0.19-1.32) or partial (β = 0.69, CI = 0.12-1.26) WSB's, and the level of self-perceived, work-related stress among smoking workers compared to those with no WSB. We also find that this association varies by gender and age. In particular, WSB's are significantly associated with higher work stress only for males and young adults (aged 18-40). No statistically significant association is found between WSB's and the level of self-perceived work-related stress among non-smoking workers.

Conclusion
The results of this study do not imply that WSB's are the main determinant of self-perceived, work-related stress among smokers but provides suggestive evidence that these may be positively related.