Does maths explain religious decline?

Last week three American academics got widespread publicity for predicting the continued decline of religion. A specific prediction "nearly 70% of the Netherlands will be non-a liated [to religion] by midcentury" was widely quoted.

So what, you might think. But the paper is more interesting than the press reports suggest.

The authors show that the decline in religious affiliation (as measured in national censuses) in four European countries follows a single mathematical form. They derive this form by assuming that religion and non-religion are two competing social groups - a view that seems ridiculously simple-minded - and explain the growth in non-religion by a drop in the perceived utility of religion. It seems odd, even offensive, to use the term 'utility' for social groups whose key differences are their opinions about truth. More significantly the authors do not explain why the utility of religion should have declined. They therefore provide no insight into the trends they document.

The interesting bit comes next.

The authors extended their mathematical model to cases where there was limited communication between the competing social groups. These models predicted the same processes of change but over a longer period. Even quite small amounts of communication between groups was sufficient to facilitate the decline in religious affiliation IF the utility of affiliation was falling.

This kind of modeling can contribute to our understanding of secularisation - but it will be essential to understand changes in the 'perceived utility of affiliation'.

The paper

A mathematical model of social group competition with application to the growth of religious non-afilliation.
Daniel M. Abrams and Haley A. Yaple of Northwestern University and Richard J. Wiener of the University of Arizona.