Are relative-income effects constant across the well-being distribution?

This paper challenges the common assumption made by economists to date that income comparisons are similarly important in different segments of the subjective well-being (SWB) distribution. The results, based on the 2000-2007 waves of the German SOEP and on a Generalized Ordered Probit for panel data, show that relative income, as measured either by the mean income of the reference group or the individual ordinal ranking within the group, exerts a differential effect across SWB levels. Such divergence is assessed by means of the tradeoff ratio between household income and the relative income variables. The results show that a low rank and falling below the average income in one's group are significant determinants of low SWB but largely irrelevant when accounting for high SWB. The fact that conditionally unhappy individuals are more sensitive to comparisons, particularly if they are unfavorable, is consistent with earlier laboratory studies in the field of psychology. -- subjective well-being ; comparison income ; income rank ; generalized ordered response model