Group Size and the Trajectory of Religious Identification


  • Acknowledgments: Collection of the New Zealand Attitudes and Values Study 2009 (NZAVS-09) data analyzed in this article was funded by University of Auckland FRDF (#3624435/9853) and ECREA (#3626075) grants awarded to Chris Sibley. William Hoverd is funded by a DFAIT postdoctoral fellowship to the SSHRC- funded MCRI “The Religion and Diversity Project” at the University of Ottawa. We thank Joe Bulbulia for his comments on an earlier version of this aricle.

William James Hoverd, Department of Classics and Religious Studies, University of Ottawa, 70 Laurier Avenue, #102, Ottawa, ON KIN 6N5. E-mail:


The relative size of religious groups or denominations within societies or nations influences variation in the extent to which group members psychologically identify with their religion. National-level census data measuring the proportional size of religious groups in New Zealand are merged with nationally representative data on self-reported psychological identification drawn from the New Zealand Attitudes and Values Study. A multilevel random coefficient model shows a logarithmic function for the relationship between religious group size and average group-level religious identification. Members of smaller religious groups (less than 1.5 percent of the population) tend to strongly identify with their religion, whereas members of groups that are larger in size (over 6 percent of the population) tend to be less identified, on average. Religious group cohesion may be a dynamic process. Larger religious groups are less cohesive and experience more contested identities and ideological positions (average group identification is lower).