To put in another way, we view religious beliefs as the principal output of the religion sector,and we view church attendance as one of the inputs to this sector. Thus,if we hold fixed the beliefs—as we do in the regressions that we highlight—an increase