Does preference diversity impact on Optimal currency area demarcation? My feeling is that back when I was at School in the early 80's the answer most people would have given was no.
People thought labor and capital mobility sufficient to constrain preference diversity to that which could drive gains from trade but not vitiate it or render the working of fiscal automatic stabilizers (like Social Security) perverse and incentive incompatible in their working.
Now, post Greece, Portugal, Spain etc. the question must be readdressed. What if a small moral hazard gets amplified such that you get a new population mix with sharper preference diversity? Then you could get a dependency culture- islands of work supporting hinterlands of leisure preference. This may still be politically acceptable if people value having a menu of life-style choices. However language and other barriers may militate against this. Still, provided assets in work-shy or dependent areas depreciate relative to assets in the islands of productivity, presumably a condition for the optimal currency area is met.
What if those assets instead of depreciating have their price either artificially inflated or stabilized by a Financial system run amok or, in its after-math, a Central Bank intervention?
It appears to me that a way to approach questions of this kind is by reversing the chain of causation and asking what sort of entities are conducive to optimal preference diversity. How and why, over their life-cycle might those entities generate perverse preference diversity matrices. What checks on this generate- for example India's political instrumentalization of a standardized poverty for reasons of internal cohesion- and how those checks can be disabled. These are the sort of questions that might profitably be posed.