Non-pecuniary returns to higher education: the effect on smoking intensity in the UK.

This paper investigates whether higher education (HE) produces non-pecuniary returns via a reduction in the intensity of consumption of health-damaging substances. In particular, it focuses on current smoking intensity of the British individuals sampled in the 29-year follow-up survey of the 1970 British Cohort Study. We estimate endogenous dummy ordinal response models for cigarette consumption and show that HE is endogenous with respect to smoking intensity and that even when endogeneity is accounted for, HE is found to have a strong negative effect on smoking intensity. Moreover, pecuniary channels, such as occupation and income, mediate only a minor part of the effect of HE. Our results are robust to modelling individual self-selection into current smoking participation (at age 29) and to estimating a dynamic model in which past smoking levels affect current smoking levels.