A. Gaggero, E. Zucchell, O. Ajnakina, R. A. Hackett
The extent to which heavy smoking and retirement behaviour are causally related remains to be determined. To overcome the endogeneity of heavy smoking behaviour, we employ a novel approach by exploiting Mendelian Randomisation and use genetic predisposition to heavy smoking, as measured with a polygenic risk score (PGS), as an instrumental variable. A total of 3578 participants from the English Longitudinal Study of Ageing (mean age 64.41 years) had data on smoking behaviour, employment and a heavy smoking PGS. Heavy smoking was indexed as smoking at least 20 cigarettes a day. Our results show that being a heavy smoker increases significantly the probability of early retirement (B= 0.635, standard error = 0.209, p < 0.001). Results were robust to a battery of robustness checks and a falsification test. Overall, our findings support a causal pathway from heavy smoking to retirement behaviour.
Keywords: Smoking; Retirement behaviour; Polygenic Risk Scores; Instrumental variable; Mendelian Randomisation
Scheduled
Data Analysis in Social Sciences
June 8, 2022 4:00 PM
Audiovisual room
