Actually from Catz data it seems like

1. The US has issues, we all know this and blame different things.

2. There is a general "peak" at a midrange of religiousness particularily as measured by attendence.

I.e. if I did a least squares fit of the data points I'd get a curve that goes up and then back down. Go on and look at the figures yourself.