An Examination of the Dynamics of Well-Being and Life Events using Vector Autoregressions

We use a panel vector autoregressions model to examine the coevolution of changes in mental well-being and changes in income, health, marital status and employment status for the British Household Panel Survey (BHPS) data set. This technique allows us to simultaneously analyze the impact of the aforementioned factors on each other. We find that increases in well-being are associated with subsequent increases in income, marriage, employment, and health variables, while increases in the these life-domain variables (except health) tend to be followed by decreases in well-being in subsequent periods, suggesting adaptation dynamics in all domains. In addition to the robust observation that well-being is associated with subsequent longer-term improvements in health status, we observe that health improvements themselves have a long-lasting positive contribution to employment status. Employment status, in turn, has lasting positive effects on income growth. Although well-being per se may be a short-lived phenomenon, our results suggest that it may have longer-term effects through these particular channels. These findings are quite robust to different model specifications.