Evidence that such a correlation exists for “white evangelicals” in the US tells us nothing about whether it exists for religious people in general - including religious Americans from other racial/religious backgrounds, and non-Americans. Your observation might not be true for religion at all, only for a very specific type of it