by Matthew Bennett, University of Oxford
This paper uses multilevel modeling techniques to test individual-level, contextual-level and cross-level hypotheses predicting volunteering across counties in England and Wales using the British Social Attitudes Surveys. I test hypotheses that focus on individual-level religiosity, the religious context, and the interplay between these levels while controlling for other known individual and contextual-level predictors of volunteering.The literature suggests that the religious volunteer more than the non-religious, however, I only find that Anglicans and members of other protestant denominations volunteer more than the secular. There are no differences between Catholics and minority religious groups compared to secular respondents. Service attendance accounts for a significant degree of this finding, as any initial differences in voluntarism disappear when I include this in a subsequent model. The self-reported importance of religion also increases the likelihood of volunteering. A more diverse religious context at the county level does increase the likelihood of volunteering, supporting the idea of an indirect link between religious competition and commitment (Borgonovi 2008). Contrary to network theories, a devout context does not have an additional positive effect on volunteering (Ruiter and De Graaf 2006). Similarly, being a member of a minority religion in a county is not associated with volunteering. I also fail to find a cross-level interaction between service attendance at the individual-level and devoutness at the county level: secular people do not have an increased likelihood of volunteering when they live in a more devout county.BiographyMatthew is a DPhil Sociology student at Oxford University. His main research interests are religion and pro‐social behaviour.