Measuring aggregate religiosity in the United States, 1952-2005

I present a method for measuring aggregate religiosity. The method allows one to combine many indicators of religiosity, even if the indicators measure only a portion of religiosity, or if the indicators are not measured every year of the series. I demonstrate the method by measuring aggregate religiosity of the US population from 1952 through 2005. The results show the sharp rise in religiosity during the 1950s, the decline beginning in the 1960s, and the slower decline since the 1970s. I present evidence of content and convergent validity of this new index.