Abstract
The objective of this article is to describe the relationship between past changes in daily cigarette consumption and happiness using a national longitudinal panel survey. Data from 724 smokers who participated in the first 11 waves of the British Household Panel Survey were analysed using a fixed effects regression model. An increase in daily smoking frequency corresponded with a decrease in happiness (β = −0.004, 95% confidence interval −0.006 to −0.001). Independent of this relationship, happiness decreased as health deteriorated with a transition to the poorest health group associated with greatest unhappiness (β = −0.190, 95% confidence interval −0.258 to −0.123) followed by those in the next poorest health group (β = 0.114, 95% confidence interval −0.176 to −0.051) and as health improved further, the effect on happiness diminished (β = −0.077, 95% confidence interval −0.127 to −0.027) indicating a dose–repose relationship between health and happiness. Smokers who reduce cigarette consumption can expect greater happiness in a relationship that is not mediated by changes in health. This relationship is interpreted as cigarettes showing reference dependence