One of the significant developments in the last four decades of economics is the growing empirical evidence that individual consumption preferences, as mea- sured by self-reported life satisfaction, are neither fixed nor self-centred but are instead overwhelmingly dominated by externalities, partly in the form of reference levels set by others and by one’s own experience. Welfare analysis recognising this fact is likely to indicate enormous revisions for macroeconomic policy and social objectives as well as for what is taught in economics at all levels. Yet the task of constructing general equilibrium models based on this microeconomic re- ality is still in its infancy. In this work I take the conventional stance that decision makers understand their own utility function. Therefore, they can choose the mi- lieu in which they immerse themselves with the sophisticated understanding that it will affect their own consumption reference levels and therefore the degree of satisfaction they derive from their private consumption. At the same time, their private consumption will help to set the reference level for others in their chosen group. I treat theoretically the problem of such endogenous formation of consump- tion reference groups in the context of a simultaneous choice of neighbourhoods and home consumption amongst a heterogenous population. For both discrete and continuous distributions of types, I find general equilibrium outcomes in which differentiation of neighbourhoods occurs endogenously and I compare the welfare implications of growth in such economies.