Brow Beat

The Fascinating Things You Learn From Reading Very Old Polls

The Twitter account @HistOpinion is utterly simple and endlessly fascinating. The brainchild of Case Western history professor Peter Shulman, the account tweets findings from public opinion surveys taken between 1935 and 1946, inserting results from Depression- and WWII-era opinion polls into your feed at the rate of three tweets a day. (You can see larger versions of all the charts in this post by clicking on them.)

The volume that supplies source material for the tweets is Public Opinion, 1935-1946, by Princeton psychologist Hadley Cantril. Cantril was a pioneer in the field of public opinion research, which took off in the mid-1930s after pollsters George Gallup, Elmo Roper, and Archibald Crossley successfully predicted FDR’s victory using statistical sampling in 1936.

From his perch at Princeton, Cantril adapted these new methods for academic purposes, and advised presidents including FDR and Eisenhower. (Cantril also authored the first study of the Orson Welles War of the Worldspanic” of 1938.)

Public Opinion compiles data from 23 polling organizations around the world, with results coming from Hungary, France, Germany, Denmark, Sweden, Australia, and Britain, as well as the U.S. Shulman is working his way through the volume, selecting the most surprising, intriguing, or unusual responses to share on the Twitter feed.

Some of the most interesting @HistOpinion tweets have to do with World War II. Contemporary responses to questions about the purposes and direction of the war remind us of the uncertainty people felt while the conflict was ongoing. “No matter how hard I try to remember that people didn’t know the war was going to end in 1945,” Shulman says, “it’s a different thing to see the poll.”

Some polls expose the underbelly of public opinion: overwhelming support for the sterilization of “habitual criminals and the hopelessly insane,” for instance, or disproportionately harsh feelings toward the Japanese.

Some @HistOpinion tweets are intriguing not because of the response data, but because of the questions themselves. Many, like this query about weather forecasting, expose now-vanished assumptions about daily practices.

A few months after starting the feed, Shulman hit upon the idea of creating graphics based on the poll results. The charts make the tweets pop on your timeline, and they also give Shulman a way to pack more information on the source of the poll, and its methods, into a tweet.

“My biggest hesitation in [curating this feed] is that you might be suggesting an age of certainty,” Shulman says. Even with the newer, more “scientific” approach to polling, he points out, sampling methods for 1930s and 1940s polls weren’t always representative. For example, “pollsters asking political questions often pulled back from asking African-Americans in the South to answer, because, they argued, black people didn’t vote.”

The Cantril volume is hefty, and fodder for the Twitter feed is practically endless. Shulman also hopes to expand @HistOpinion into a blog or a Tumblr, where he could have space to address questions about the pollsters’ methods, as well as ponder the possible implications of their results. 

Previous Follow Fridays
Florida Man, the World’s Worst Superhero
Every U.S. Drone Attack
Philly Sports Jerks
The U.S. Interior on Instagram
Where Debit-Card Numbers Get Stolen
A Twitter Feed That Teaches You About Language
A Crowd-Sourced Lab for Collective Nouns
A Twitter Feed Douglas Adams Would Have Loved
Terrorists Who Tweet