I was the author of that paper from which you quoted Robin Murray. We actually used a measure of voter turnout (in each electoral ward) at local, not general elections. The distinction is important, and at best either measure is only a proxy for social cohesion, but we argued that voter turnout at a local election would correlate with the extent to which people invested and cared in local issues. Voter turnout was measured in a local election year only (no concurrent general election) around the same time we estimated the number of new cases of psychosis in each electoral ward. The two were correlated, after taking into account other possible explanations, like differences in the composition of the population ofdifferent electoral wards, or differences in deprivation. I hope this is helpful. BW James

But perhaps the correlation between the two might in part be due to the dazzle of disorder. The mind cannot rest in one place if control levels are low, but must dance erratically from one thought, vision etc to another. It could be seen as a survival necessity.
In a higher and more disordered population the possible threat to a person is less controlled and more imediate and could come from a higher number of sources. In a modern urban environment we are constantly bombarded by information which we barely have time to register or reason through before we must move on.
It makes sense therefore that there will be higher incidences of mental illness and less sense of permanance or reality.