Religion, happiness & choice

What’s the link between religion and happiness? This paper (pdf) suggests that it is not a linear one, but rather U-shaped:

All else constant, people are happier in countries in which the level of religiosity is either high or low, but are less happy in countries with intermediate levels of religiosity.

One reason for this, the authors say, lies in network effects. A country with middling religiosity will be a heterogenous one with little social cohesion. Religious folk will worry about a loss of values whilst atheists will fret about the social restrictions imposed by the pious. No-one will be satisfied. Alternatively, nations with middling levels of religiosity are often moving from one thing to the other, which breeds uncertainty about the future.
However, I wonder if there might be something else going on too. Although this finding holds for societies, might the same be true for individuals? Atheists and fanatics can be happy, but the mildly religious struggle.
Take two examples. In Jolie Holland’s song Corrido por Buddy, she sings of the “bad religion” that “brings you down and can never lift you up”, leading to the suicide of its victim.
And in Corrie, Sophie Webster recently tried to kill herself, motivated in part by her struggle between her faith in God and her lesbianism. Her problem is one of being middlingly religious - sufficiently so to fear God, but not so much so as to have convinced herself He loves her.  Her more religious pastor seems convinced that God loves him, which only shows that God has lousy taste. 
If I’m right, it suggests there is a problem with the notion that people can maximize utility. It would have been pointless to suggest to Buddy or Sophie that they change their beliefs - even though doing so seems to us an obvious way out of despair. This is because we cannot always freely choose our beliefs. They are the result of socialization or upbringing or a perception that God does or does not exist.
This suggests that there is not only the obvious budget constraint upon utility maximization, but a belief constraint too. Which raises the question of how many other such constraints there are, and so what a well-defined and psychologically plausible “utility maximization” problem would really look like.

Comments

[BenSix]

A counterexample: Mitchell Heisman. The vaguely religious may be comforted by their beliefs without ever having to interrogate their premises.

[markey]

In my experince, relgious people only tend to be happier because they reckon the non-religous are going to Hell, or that they will have a bigger, better cloud to sit on.