We consider a dynamic macroeconomic model where households derive utility from wealth comparisons. Measure of relative affluence can be the ratio to, or the difference from, the social average. These two specifications lead to quite different equilibrium consequences: under the ratio specification full employment is always realised, whereas under the difference specification persistent shortage ... [Show full abstract] of aggregate demand can result. Using data of an experiment on status concern, we empirically find that the difference specification of relative affluence fits the data better than the ratio specification. Our work may help in understanding mechanisms behind recent stagnation in developed economies.