Creationism appears to be annoyingly common in the USA (even if gradually decreasing). However, things may not be as hopeless as they seem; as Pew Research finds they’ve been consistently over-estimating the number of creationists in the country.

It all comes down to how the question is asked. Prior research has found that giving people more options (like “undecided”) can dramatically reduce how many claim to be creationist. Pew Research is taking things a step further, and adding even more nuance into their options.

When doing so, the number of creationists they found nearly halved; dropping from 31% to 18%!

Where did Pew Rearch go wrong?

Previously, they’d been asking about evolution first; with a follow-up question about God’s role given to the people in agreement with evolution.

However, there were concerns that the first question might come across as “science versus religion”, scaring people away from the more secular answer. Certainly, other studies show that religious folk are more accepting of evolution when they can clarify God was still involved.

So Pew Research came up with a single-question format, mushing everything together into a single statement with more answers; including some featuring God’s involvement.

Their typical question (right) and the new format they tried out (right)

Obviously, the next stage was to see if this change did make a difference. So they gave an updated questionnaire, featuring the new format, to some people that had previously responded to surveys with the old format.

This revealed a dramatic drop in creationism under the single-question format. Either a lot of people changed their mind between surveys, or question format made a big difference.

As the results below show, there was an overall decrease in creationism in the general population, going from 31% to 18%. This stems in large part from a dramatic change in evangelical Protestants and Catholics; with more mainline Protestants (and the unaffiliated) remaining consistent between the two question types.

Hooray for science?

So great, the prevalence of creationism is far lower than we thought. Don’t get me wrong, this is good news. However, there are also some downsides to this discovery as well. Notably, it means we have no real historical context for creationist belief.

One of the big issues with creationism in America is how stubborn it was. Although its acceptance was going down, it was doing so at a snail’s pace.

The results of Gallup surveys over time

However, all that prior data is now suspect. How many creationists were there in 2009? 1985? Turns out we don’t know, as the survey technique back then was flawed; over-estimating their prevalence. Maybe it’s actually been on the rise! Or maybe it has also held steady over the years, putting us right back at square one (albeit a smaller, less depressing square).

The fact is we’re going to have to wait years to get a new, more accurate idea of creationist trends over time. Until then, it might be a bit early to pop the champagne.


  1. Funk, C. (2019). How highly religious Americans view evolution depends on how they’re asked about it. [online] Pew Research Center.

Related posts


Joe · 14th February 2019 at 2:15 pm

They should ask how many people agree that minds came from the mindless via blind and mindless processes? And how many of those people think that the claim is scientific.

The bigger issue than Creationism is people who believe minds came from the mindless via blind and mindless processes. Those people don’t know anything about science nor reality.

    Adam Benton · 20th February 2019 at 2:02 am

    The data is there in the post. ~40% of US adults think humans evolved via natural processes, the single largest group.

Leave a Reply

Your e-mail address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.