Estimating linear birth cohort effects: revisiting the age-happiness profile.
Open University, Economics Department, Milton Keynes, UK.
Full text available as:
This paper provides a simple way of accounting for linear birth cohort effects, together with linear age and calendar time effects. It relies on the discreteness of the data and on the fact that not all individuals are
born/interviewed in the same day. This creates an exogenous source of age variation within the same cohort that breaks the linear dependence between the three variables. This method is applied to a happiness equation and shows that, once a linear birth cohort term is included in the regression equation, together with linear age and calendar time terms, the robustly found U-shape profile of happiness in age disappears.
Actions (login may be required)