Landon Schnabel, Sean Bock
Sociological Science, November 15, 2018
10.15195/v5.a30
Abstract
In their comment on our article about the persistence of intense religion in the United States, David Voas and Mark Chaves (2018) claimed that “the intensely religious segment of the American population is shrinking.” In this response, we show that intense religion has persisted from the 1970s to the present, with a temporary uptick during the exceptional Reagan years. Voas and Chaves concluded otherwise because their analytical strategy was not sufficiently sensitive to nonlinear patterns. In addition to demonstrating the continuing persistence of intense religion, we also discuss criteria for measuring intense religion over time and the importance of avoiding unfounded assumptions in age–period–cohort analysis. We conclude that aspects of the classic secularization thesis championed by Voas, Chaves, and others are not supported by the data, and we suggest that scholars should look for better ways of thinking about religious change.
In their comment on our article about the persistence of intense religion in the United States, David Voas and Mark Chaves (2018) claimed that “the intensely religious segment of the American population is shrinking.” In this response, we show that intense religion has persisted from the 1970s to the present, with a temporary uptick during the exceptional Reagan years. Voas and Chaves concluded otherwise because their analytical strategy was not sufficiently sensitive to nonlinear patterns. In addition to demonstrating the continuing persistence of intense religion, we also discuss criteria for measuring intense religion over time and the importance of avoiding unfounded assumptions in age–period–cohort analysis. We conclude that aspects of the classic secularization thesis championed by Voas, Chaves, and others are not supported by the data, and we suggest that scholars should look for better ways of thinking about religious change.
 | This work is licensed under a Creative Commons Attribution 4.0 International License. |
- Citation: Schnabel, Landon, and Sean Bock. 2018. “The Continuing Persistence of Intense Religion in the United States.” Sociological Science 5: 711-721.
- Received: October 25, 2018
- Accepted: October 29, 2018
- Editors: Jesper Sørensen
- DOI: 10.15195/v5.a30
We encourage readers to pay close attention to the full range of evidence presented in both our article and in Schnabel’s and Bock’s rejoinder. The evidence for decline in “intense” religiosity is clear in both pieces.
On the core issue of whether intense religiosity is declining or stable, we encourage readers to pay particular attention to Table 3 in our article. The early period for two of the key items – considering oneself strongly religiously affiliated and attending services more than weekly – predate 1980. The earlier period for two other items – knowing God exists and considering oneself a very religious person – start in or after 1988. Slow but clear decline is evident in all four, showing that the decline in intense religiosity is not an artifact of higher levels of religion during the Reagan years. These declines also are evident in the results reported in Table 1 of Schnabel’s and Bock’s rejoinder, and they are visually evident in their Figure 1. The negative slopes of the regression lines are not always statistically significant because the declines are too slow to reliably discern in that way. But the declines are clear and statistically significant in our Table 3. Recalculating all these numbers using the entire GSS sample, including immigrants, makes no substantive difference to any of them.
Regarding cohort change, we encourage readers to look carefully at Figure 2 in this rejoinder. In this figure, cohort differences in the secular direction would appear as downwardly sloping lines for each age group tracked over time. The lines in their figure 2 all slope downward, even if the Reagan uptick is ignored, just as one would expect when younger cohorts are steadily less religious than older cohorts. The most sensible interpretation of this pattern is the cohort change interpretation we originally offered.
Overall, even the evidence Schnabel and Bock present in their rejoinder shows that intense religiosity has declined in recent decades. They do not interpret it that way because, just like in their original article, they do not appreciate that very slow decline is still decline. We encourage readers to consider all the evidence contained in both our article and their rejoinder (including the supporting evidence from Pew surveys that we cited in our article), and ask themselves whether, taking all the evidence into account, a clear signal of declining intense religiosity emerges from the noise.
Mark Chaves and David Voas
A clarification: In the second sentence of the second paragraph of our previous comment, we should have said “predate or include 1980” instead of “predate 1980.” In our table 3, the early period for attending more than weekly is 1973-1978; the early period for strong affiliation is 1974-1980.
But it doesn’t matter. Excluding 1980 from the early period for strong affiliation makes no difference. Excluding 1988 from the early period for knowing God exists also doesn’t matter. Decline is evident either way. Ignoring the Reagan years does not change the basic picture of slow decline.
Thank you for the responses, Mark and David.
We clearly disagree on quite a few things, but we completely agree that it’s important to look at both the comment and the rejoinder.
We strongly encourage people to read through both papers with a focus on the analyses, and to judge the empirical evidence for themselves.