This study uses the ECHP 2000 dataset to investigate the determinants of individuals' happiness. We model happiness as a latent trait that stochastically determines responses to four satisfaction questions; the relations between social characteristics and the trait, and between the trait and the responses, are estimated in a unified semiparametric model. The results provide a succinct characterization of the role of covariates in determining the distribution of happiness within and between European states. Relative income plays an important role in the analysis and its effects are further examined by estimating sub-models for groups defined by incomes relative to cohort averages. This more detailed analysis suggests that that the genesis of happiness is affected by relative social status, and that income is more important to high status individuals.