The Man Who Made The Public Opinion Poll
In 1948, TIME called George Gallup the “Babe Ruth of the polling profession.” Gallup is the father of the now famous Gallup Poll which uses statistics and science to drive polling. Lily Rothman at TIME would say that Gallup was, at the end of the day, “driven…by his faith in numbers and a desire to measure the world.”
Gallup was born on November 18, 1901, and grew up raised by dairy farmers in Iowa. He was known as “Ted” and would deliver milk as his job as a teenager, and would later start a high school newspaper. Along with being an active member of the newspaper, Gallup played football. Gallup went to college at the University of Iowa and was a part of the Sigma Alpha Epsilon (SAE) fraternity. There, he also played football and edited for the school newspaper.
Gallup would earn a Bachelor’s, Master’s, and Ph. D. from the University of Iowa from 1923 to 1929. He then served for two years as the head of the Department of Journalism at Drake University until 1931. He would serve in academia teaching journalism and advertising for a couple of years until he formed the American Institute of Public Opinion (Gallup Poll) in 1935.
According to Lily Rothman at TIME, Gallup liked to predict who and figure out who would actually read the newspaper. To do that, Gallup started taking out the crossword puzzles in certain weeks, and seeing how many people complained about it. When someone confronted Gallup, Gallup started asking people what they liked about the newspaper and what they didn’t like about the newspaper.
It’s important to note the Gallup poll wasn’t only used for elections — it predicted so much more. Gallup wanted to see if he could use his strategy for predicting the 1932 Presidential Election. He also wanted to see if his strategy would work by incorporating science into political surveys for his mother-in-law, who was running as Secretary of State in Iowa.
His mother-in-law won, as Gallup predicted correctly. According to a TIME report in 1932, Gallup became convinced predicting toothpaste and politics were the same, and the existing poll at the time was run by a newspaper called The Literary Digest. Gallup would soon challenge it. In 1935, Gallup started the American Institute of Public Opinion, which started the Gallup Poll.
The Digest had a very biased way of conducting its polls — it mailed out millions of postcards based on telephone and auto registration lists. As a result, it failed to take into account many low-income voters who weren’t on those lists, and who had bought into the hype surrounding the New Deal. In the 1936 election between Alf Landon and Franklin D. Roosevelt, the Literary Digest predicted Landon would receive 56% of the vote —but FDR won by a landslide (62% of the vote), leaving a 19% sampling error for the Digest, the largest ever sampling error for a public opinion poll.
The Literary Digest’s mistake would make a lesson for the University of Pennsylvania math department: avoiding selection bias and nonresponse bias. The Digest had a list slanted towards middle and upper-class voters. Gallup correctly predicted, through polling, that FDR would win the 1936 Presidential Election.
“By 1948, the Gallup Poll organization — officially called the American Institute of Public Opinion — operated in a dozen countries, influencing everything from the titles of Hollywood movies to Book-of-the-Month Club picks, in addition to its political predictions. By its sheer omnipresence (the Gallup Poll released data to newspapers a whopping four days a week) Gallup became synonymous with polling,” Rothman said.
Of course, the Gallup poll was not perfect. In 1948, the Gallup poll was among many to make the worst polling failure in American history when it predicted Thomas Dewey would win over Harry Truman in a landslide.
It turns out predicting voter turnout and behavior is significantly more difficult than predicting what kind of toothpaste people like.
In 1958, Gallup combined his organization, the American Institute of Public Opinion with the British Institute of Public Opinion. It would be called The Gallup Organization, which still obviously operates today. Gallup also worked with Elmo Roper, another pollster, to establish the Roper Center for Public Opinion Research at Cornell University. He archived the data for the institution to make sure people could study the history of public opinion.
Over time, Gallup gained more credence and influence in the public opinion polling industry. According to the Roper Center, the Gallup Poll had surveyed 1 in 9 Americans over the age of 19 by 1975. They also were able to ask questions beyond just policy issues, asking about religious beliefs, consumer preferences, and political ideology to make more comprehensive data.
Prominent politicians saw Gallup as an asset. In the 1960s, George Romney, John F. Kennedy, and Nelson Rockefeller consulted Gallup for their presidential campaigns. He started doing private polls for candidates he was working with and became criticized for doing so. In the words of the Roper Center:
“Some charg[ed] that Gallup’s organization was using these private polls to influence candidates.”
Gallup then stopped doing private polls for politicians and has tried to take a nonpartisan stance.
“In the years since, Gallup’s polls have had a significant impact in both the public and private sphere. Gallup Polls have given politicians data on public support of policy measures, helped corporations decide how to best market their products, and provided sociologists with insight into what America does with its leisure time,” the Roper Center said.
Gallup died in 1984 of a heart attack, but he achieved his life mission. He always wanted to capture public opinion on a variety of topics. He thought that by knowing how Americans thought and believed, many of society’s problems can be solved. I think he achieved his life mission.
I have been a huge skeptic about polls for the past week or so, but also the past four years. But what would life be like without polls? I can’t deny that reading public opinion polls gives a sense of connectedness. If 65% of Americans agree with me on my favorite ice cream flavor, I feel like I’m in good company. At least I know I’m not crazy for thinking and believing what I do on a variety of issues.
But polling certainly is a touchy profession. Predicting human behavior will always come with a heavy amount of uncertainty. And it’s always adapting to account for its biases, as it did after the Literary Digest poll in 1936 or the widespread polling error in the 1948 Presidential Election, or its miscalculations in the 2016 Presidential Election and our most recent 2020 Presidential Election.
For me, polling is a great tool and indicator of opinion and attitudes. But it is just an indicator — should never dictate opinion and behavior. I wrote previously that polling showing your side is ahead in an election by a landslide should never lead to complacency. Public opinion isn’t going away any time soon — society relies on it too much, but it
So no matter what you might think of Gallup’s polling and his organization, you can’t deny that George Gallup changed the world, and brought us one step closer to measuring it.