In our interminable run-up to the next U.S. Presidential election, we are regularly bombarded with information from a variety of sources about how each of the candidates from both parties are doing in the polls. Often these poll results, whether from Pew Research, Gallup, or another organization, include information about how a particular candidate is faring in a particular religious demographic.
There are some who question how those religious profiles are constructed and whether they are, indeed, accurate.
In the newly released book, Inventing American Religion: Polls, Surveys, and the Tenuous Quest for a Nation’s Faith, Robert Wuthnow outlines the rise of scientific polling, the increasing influence of the religious questions in them, and the more recent decline in support for the published poll results.
Wuthnow’s thesis is “that the polling industry has influenced—and at times distorted—how religion is understood and portrayed, particularly in the media but also to some extent by religious leaders, practitioners, and scholars.” He argues this thesis is eight chapters.
The introduction outlines the early history of public polling and surveys the breadth of the history that Wuthnow goes on to unpack and interpret in the remaining seven chapters. Chapter Two covers the early attempts to do comprehensive surveys to assess public opinion. Such surveys were accurate for local issues, but they were time consuming, expensive, and unable to establish broader public opinion. Still, they were a common tool used by social organizers like Du Bois. In the third chapter, Wuthnow outlines the rise of George Gallup, who pioneered the use of the scientific poll to assay public opinion on a broader scale; since Gallup was a self-professed Christian, he asked religious questions in his polls, which began the process of examining the impact of religion on social and political positions.
Chapter Four highlights the differences between scientific studies, which are usually carried out by scholars, and public opinion polls. Wuthnow explains that polls are designed as quick hit diagnostics, based on an attempt to gain a rough idea of a person's opinion with as few questions as possible. In contrast, scientific studies ask more probing questions. As a result, scientific studies tend to be more narrowly focused, but they also tend to have more precise explanations for the results. Scientific studies go after the “why” not merely the “what.” In the fifth chapter, the evolution of the pollster as pundit is discussed. In 1976, the so-called year of the evangelical, the religion question become more important. Suddenly Gallup’s years of asking about religion began to pay off. Additionally, the people doing the polls began to interpret them for the media audiences. It’s easy to see how possible misinterpretations can result. This trend to question the polls has grown since that point; for some, the promise of punditry undermines the possibility of objectivity in the polls.
In the sixth chapter, Wuthnow describes the falling confidence in polls. This was due to the conflation of pundits and pollsters. It also has to do with the changing demographic of respondents. Initially people would answer the phone and respond to polls, but that began to change. Response rate became an issue and the questions about the demographics of those actually responding to polls arose. Additionally, external observers (though not the pollsters themselves) began to question some of the ways conclusions were drawn. These observers began to notice fluctuations in some of the responses, such that the percentage of church goers varied widely across a six-month span in some cases. Chapter Seven discusses the breaking down of the fourth wall, when pollsters began to take polls about polls. The answers began to show a growing distrust in the accuracy and usefulness of polling. However, Wuthnow argues that the influence of polling is far from gone. He notes, “Polling studies demonstrate that polling rarely has discernible effects on election outcomes, but it offers background information that draws attention to how candidates are doing and reinforces implicit perceptions that some issues are more important than others.”
Wuthnow concludes the book by surveying the state of polling in Chapter Eight. Polls are still important, but they aren’t the trusted sources of information they once were. They are now more likely to be used as sermon illustrations or points for beginning a more in depth process of investigation. Polls continue to suffer due to lowering response rates. People's lives have become saturated with polls, opinion questions, and other calls for feedback. In a world of big data, politicians and corporations are turning away from using polls as ultimate grounds for decisions; better information is available in usage statistics from Facebook, Twitter, and other sources. Additionally, the history of polls continues to show that the categories being used to define religion are no longer adequate (if they ever were). The future of polling, particularly related to religion, is indeterminate. It is unlikely that polls will disappear, but criticisms of polls may continue to reduce their importance. Time will tell.
Analysis and Conclusion
Wuthnow’s book is timely. Polls are regularly published; their results are lauded as sure truths by the 24-hour media cycle by pollsters and the talking heads. Real people, on the other hand, are asking more and more whether the results are trustworthy. After all, we think, when is the last time I was asked to respond to a poll? Most of us don’t even answer the phone when we don’t know the number. Our experience drives us to question the validity of polls, whether that is just or not.
Inventing American Religion is part history and part critique. His history shows what has happened and, it seems to me, explains is very clearly. His critique is a telling warning about how polls have been abused and how to avoid being mislead by them. At least it provides grounds for asking further questions, something that is nearly always a worthwhile endeavor.
The weakness of this volume is that it highlights a problem--the potential unreliability of polls--but it fails to provide a solution. This, of course, was not part of his thesis. However, if Wuthnow had any suggestions about how to improve the use of polls or interpret them better, it would have been good to include them.
In the balance, this is an important book for academics and pastors who want to use polls in their papers, books, and sermons. Wuthnow's point is well made: polls may not be trustworthy and misreading the data may well lead the consumer astray.
Note: Oxford University Press provided a gratis copy of this volume with no expectation of a positive review.