Abstract
Among sociologists of religion, the rhetorical query about the Pope’s affiliation could easily be replaced by the question ‘Is Robert Wuthnow productive?’ Wuthnow has written a book more or less every year for the past three decades. And these books are not slim SPCK paperbacks; they are monster monographs. The offering from 2014 (How Texas Became America’s Most Powerful Bible-Belt State) came in at 664 pages. The first book of 2015 was a mere pamphlet at 240 pages (In the Blood: Understanding America’s Farm Families), so there was plenty of time to crank out another in the same year: Inventing American Religion: Polls, Surveys, and the Tenuous Quest for a Nation’s Faith. Of course Wuthnow is not a lone scholar, ensconced in the library at Princeton (where he is professor of sociology) or out conducting interviews. Along with Robert Putnam, Christian Smith and others, he is an exemplar of the ‘lab director’ model of American social science, which has proved so productive in studies of religion and society. Still, it is hard not to feel both impressed and exhausted.
Inventing American Religion provides a history of how opinion polls and surveys came to be used to measure the religious involvement of the American public. I now know far more than I did before about my own field, though I doubt that the volume will be flying off the shelves in anyone’s local bookshop. Wuthnow isn’t content to provide a descriptive overview, however: he wants to argue that these measures of faith are failing to live up to their billing. In this respect his theme is the maxim ‘Not everything that counts can be counted, and not everything that can be counted counts.’
I wholly agree. My concern is that non-quantitative people (and Wuthnow is primarily a qualitative researcher) tend to have one or both of two contrasting responses to numbers. The first is naïve confidence: if a survey says that 72 per cent of the population is Christian, that must be right. The second is a radical scepticism about the very idea of quantification. On this view, faith defies measurement and so measurement necessarily traduces faith.
The title of this book points in that direction: the implication seems to be that polls and surveys have somehow fabricated a kind a religion that wouldn’t otherwise exist. One might argue that we are only wedded to denominational identities, for example, because someone out there has persuaded us that these categories are important. Wuthnow does not really try to push a postmodernist critique, however. Instead he highlights multiple issues around validity, reliability, variations in wording, framing effects, latent constructs versus observable indicators, dimensionality, subpopulations, sampling, non-response and so on. But these problems are the bread-and-butter concerns of survey researchers; while ordinary people may indeed be too ready to accept numbers at face value, the failure is in the public understanding rather than in the science.
Wuthnow points to the mediocre performance of Gallup polls in the 2012 American election and the lack of confidence that might inspire in the organization’s output on religion. What I find more striking is that Nate Silver, the statistician who analyses political polls, correctly forecast the 2012 outcome in every one of the 50 states. The useful conclusion is not that polls are flawed (which they are) and that we should take top-line findings with large doses of salt (though that is also true), but that we need expert interpreters. The more data we have, the better our understanding is likely to be.