Communication à un séminaire :
This paper uses life satisfaction (LS) and Body Mass Index (BMI) measures from three waves of the GSOEP data to test for the existence of well-being spillovers in body shape between spouses. Using semi-parametric regressions, we find that partner’s BMI is, beyond some threshold representing partner’s ideal BMI, negatively correlated with individual’s own LS. Own BMI tend to be positively correlated with LS in thin men, and negatively correlated with LS, over some threshold defined as own ideal BMI. Own ideal BMI tend to move upward when partner’s BMI increases in the range of overweight. Hence, the marginal negative impact on well-being of partner’s BMI in overweight and obese individuals is lower when the individual is also overweight and obese. This finding may be interpreted as evidence in favour of contagion effects, whereby individuals will tend to follow their partner if the latter gains weight. Instrumental variable estimates show however that these correlations do not reflect causal effects, but only welfare comparisons between individuals. Yet, the latter have an interest in themselves for policy-makers, as they identify sub-population obesity policies may target.