Philipp Biermann, Jürgen Bitzer, Erkan Goeren
University of Oldenburg, Department of Economics,
(Oldenburg Working Papers V-421-19 )
Based on data from the German Socio-Economic Panel (SOEP), we used a correlated random effects econometric framework to simultaneously estimate the within and between effects of age on subjective well-being. The proposed approach overcomes the ambiguity in the relationship between age and subjective well-being reported in a series of studies based on cross-sectional and/or longitudinal panel data. Our results suggest that a cubic-type functional relationship between well-being and age fits the data best, leading to highly significant coefficient estimates associated with the age variables, and consistent within and between effects of age on subjective well-being. A linear or quadratic functional relationship between well-being and age is not empirically supported, as the between and within estimates of age on well-being differ significantly from each other. The main findings are robust to the inclusion of a broad range of individual-level sociological, demographic, and economic controls, and to the inclusion of various interviewer controls such as survey experience, survey type, and interviewer fixed effects.
Keywords: Subjective Well-Being, Life Cycle Happiness, Cohort Eﬀects, Mundlak Approach, Correlated Random Eﬀects, Fixed Eﬀects, Between- and Within-Person Eﬀects