UPDATE by JAC: Dan Kahan of the Cultural Cognition Project of Yale Law School has further analyzed this survey and finds some problems with it: some data are missing in both the summary and the full report, and this makes it impossible to determine whether the pro-creationist tendencies of Republicans reflects a shift in ideology or merely a transfer of creationist Democrats into the G.O.P. or a move of evolutionist Republicans into the Democratic Party. I haven’t had time to analyze this in full, but what disturbs me is the big disparity between the Pew and the Gallup Polls. I don’t know which one gives the correct data about Americans, but one thing I’ve noticed is that Pew polls always give results more favorable to liberal religion than Gallup Polls. (In this case, Pew shows far less acceptance of both creationism and theistic evolution than does Gallup.) If Pew releases more data I’ll try to give an update.
h/t: Carl Zimmer, Matthew Cobb
by Greg Mayer
The Pew Research Center and the Gallup Poll are two large American polling operations that periodically include questions about the acceptance of evolution in their polls of the American public. Jerry reported back in April on the results of the last Gallup Poll, and two days ago the Pew, as part of its “Religion & Public Life Project“, released its latest results (press release; exact questions with answers; report). The overall result is little different from their previous survey on this question in 2009, which was 31/61 for reject/accept evolution.
As expected, religion had a large effect on evolution acceptance: white evangelical Protestants are decisively anti-evolution while white mainline Protestants even more decisively accept evolution; in fact, acceptance is slightly higher among the latter than among the “unaffiliated”. “Unaffiliated” people include the non-denominationally religious as well as the non-religious.
Also among what must be considered expected results are the following, as summarized by Pew:
Younger adults are more likely than older generations to believe that living things have evolved over time. And those with more years of formal schooling are more likely than those with less education to say that humans and animals have evolved over time.
The results by age bode well for the future (we may be able to say of creationism, “this too shall pass”), while the results by educational attainment suggest that education is not entirely powerless against superstition. [JAC: An alternative explanation is that it is largely those who accept evolution that seek or are successful in higher education.]
As you can see in the table below, Pew actually asked two different questions, one about “humans and other living things”, the other about “animals and other living things”. Each version was asked of about half of the total sample (about 4000, so 2000 for each version). The results are largely the same, although evolution acceptance is slightly higher for the “animal” version. [JAC: As Greg notes the differences are small; still, in 4 of 4 age groups, acceptance of animal evolution is higher than of human evolution. That is almost significant using the sign test, showing that people are probably less likely to think that our species evolved than did other species. ]
This shows the effect of exact phrasing of survey questions on the results obtained (an effect highlighted in a New York Times article on a different subject from Monday: see the 3rd and 4th paragraphs).
The Pew release highlights the divergence in views along political party lines:
There are sizable differences among partisan groups in beliefs about evolution. Republicans are less inclined than either Democrats or political independents to say that humans have evolved over time. Roughly two-thirds of Democrats (67%) and independents (65%) say that humans have evolved over time, compared with less than half of Republicans (43%).
The size of the gap between partisan groups has grown since 2009. Republicans are less inclined today than they were in 2009 to say that humans have evolved over time (43% today vs. 54% in 2009), while opinion among both Democrats and independents has remained about the same.
Differences in the racial and ethnic composition of Democrats and Republicans or differences in their levels of religious commitment do not wholly explain partisan differences in beliefs about evolution. Indeed, the partisan differences remain even when taking these other characteristics into account.
Back in April, Jerry noted this partisan divide in the Gallup data. Gallup had Republicans favoring creationism by a 22 point spread (58% creationism to 36 % evolution), while Democrats favored evolution by a 10 point spread (51% evolution to 41% for creationism), and independents favored evolution by a 14 point spread (53% evolution to 39% for creationism). In the Pew data, the comparable figures are Republicans with a 5% spread for creationism, while Democrats favor evolution by a 40% spread and independents favor evolution by a 37% spread. There is thus a large divergence between the Pew and Gallup data, with Pew showing Democrats, Republicans, and independents all much more favorable to evolution than do the Gallup data. Why might this be so?
To get at this question, let’s first unpack the Gallup data. Like Pew, Gallup asked about “human” evolution, and thus this part of the poll does correspond to what half the Pew sample was asked (and whose responses are the ones given in the colored graphs above). Gallup, however, gave respondents three choices: humans developed over millions of years without God guiding the process, humans developed over millions of years with God guiding the process, or humans appeared just as they are within the last 10,000 years. We may roughly call these three possibilities naturalistic evolution, theistic evolution, and creationism. Gallup has asked this same question going back to 1982:
Naturalistic evolution (the lower line) varies from 9 to 16%, with some hint of an upward movement; theistic evolution (middle line) varies from 32 to 40%, with not much hint of a trend; and creationism (the upper line) varies from 40 to 46%, again without much evidence of a trend.
Even though Pew’s first question only had two choices, we can find comparable data to Gallup in the Pew poll by looking at one of their follow-up questions. Respondents who accepted evolution were asked by Pew if they thought evolution was due to naturalistic processes or guided by a supreme being. This divides the Pew respondents into three groups based on what they accept, just like in the Gallup poll: naturalistic evolution, theistic evolution, and creationist. The results are these:
For those who were asked the “animals” version of the question, the theistic/naturalistic breakdown was 24/35, again slightly less religious than those asked the “human” form of the question. So for the most recent polls, Gallup (2012) and Pew (2013) give the following breakdown for the American public as a whole:
Naturalistic evolution: Gallup 15% ; Pew 32%
Theistic evolution: Gallup 32% ; Pew 24%
Creationism: Gallup 46% ; Pew 32%
Even allowing for a what is perhaps a random uptick of creationism in the latest Gallup poll (see Gallup graph above), there is a striking difference between the results of the two polls.
There are a number of differences in the wording of the questions that might account for this. First, Pew suggested that one of the naturalistic processes might be natural selection. Perhaps hearing the name of a familiar evolutionary mechanism encouraged more people to choose this response, as opposed to the Gallup phrasing, in which the absence of God was emphasized, and no natural mechanisms were mentioned in the naturalistic evolution choice.
Second, the time frame of the Gallup question on human evolution was “within the last 10,000 years”, while Pew’s asked about “since the beginning of time”. At first, I thought the Gallup anti-evolution response was the more extreme choice: in April, Jerry equated it to young Earth creationism (YEC), which is indeed associated with the 10,000 year figure. However, the Gallup question asked only about humans, so a respondent who is generally accepting of evolution, but thought that something special happened fairly recently in human evolution (ensoulment?), might have selected this answer. Thus, someone who put a high premium on human uniqueness, but would otherwise be a theistic evolutionist, might have chosen what on the face of it appears to be a YEC response in the Gallup poll.
Also, the Pew phrasing, “Humans and other living things have existed in their present form since the beginning of time” may have been too strong for some varieties of creationists, who believe that a certain amount of change has occurred in some animals, and that humans have ‘degenerated’ since the fall of Adam in the Garden, leading some of them to not pick the ‘creationist’ response. However, the alternative response in the Pew poll explicitly uses the phrase “Humans…evolved”, which I think few creationists, of any stripe, would have chosen.
None of these suggestions about how the wording may have shifted the responses seems fully convincing to me, and in the end I’m not really sure why the responses diverge between the two polls.
And finally, let me leave you with the full table of responses by religion to Pew’s follow up question on evolutionary processes. I would point out here that the “unaffiliated” are by far the group most strongly favoring naturalistic evolution, even though white mainline Protestants are slightly more accepting of evolution overall.
(For the latest Pew poll, the margin of error was about 3%.)