It's probably a cognitive psychology effect rather than a social psychology one. Don't assumed that the social context is always the most important. There's a lot of our thinking that is very robust across all known social circumstances, probaly hard-wired in, so that atention to the social context is actually misleading. The endowment effect is probably a very good example of this. Also along these lines: cognitive dissonance and the related phenomenon of sour grapes (adaptive preference formation, or "I do not want what I have not got"), and various other things usefully explored by Jon Elster in his books on the social implications of cognitive spychology, Sour Grapes, Ulysses and the Sirens, and other books jks
WS: Excellent point. Not everything is socially constructed or determined. However, cognitive factors work on a very general level i.e. we may have a tendency to cognitively lump or separate things, but how we do that is affected by socio-historic conditions. For example the 'endowment effect' Barkley mention may apply to thing that we think of as 'being something of value that we possess' (hence we engage in the packrat behavior) - but it is socially deteremined what those things are. Fifty years ago, we would not think of a patch of swamp as being such a thing, but today we define it as "wetlands," see it its irreplaceable ecological value and protect from destruction.
Stated differently, cognitive processes involve a priori schemata in organizing perceptual data - and the use of such schemata is probably 'hard wired' as you say. However, a particular schema used by a particular person or a group of people is constructed and defined by patterns of social interaction (as Emile Durkheim aptly observed in _The Elementary forms of the Religious Life_ which arguably is the greatest constribution linking Kantian philosophy to modern social science).
Wojtek