When the latest 'happiness maps' show that the happiest people live in the least populated places (Outer Hebrides, Orkney and Shetland) I am once again convinced that we can learn little, if anything, from spatial differences in self-reported happiness. Quite simply, if people care about their wellbeing and if this drives decisions about where to live then the most heavily populated places must offer something that low populated places don't. So either the people in the Outer Hebrides aren't very representative of the overall population or self-reported happiness misses something important (or both).
I provide a more detailed discussion in my March post on 'Miserable Londoners' but this is the basic idea and the reason why I take these things with a pinch of salt.