Political polling as we know it began more than 80 years ago. Soon after its introduction, the polling industry gathered in Central City, Colorado, in 1947 to establish itself as a professional association and define its set of rules. The American Association for Public Opinion Research (AAPOR) will return to Colorado for the first time since 1947 with its national convention in Denver next May.
The 1936 presidential election became the first election to include a statistically based presidential poll. The polling began as a curiosity, but today it dominates much of the national media coverage and political conversation.
In 1936, for the first time in a presidential election, polling pioneers Archibald Crossley, Elmo Roper and George Gallup put polls in the field for dozens of media clients, including the Hearst Publication and Henry Luce’s Fortune Magazine. The risk was high. A miss could doom their credibility and stifle a nascent industry – or at least delay it for several years. They predicted a Franklin D. Roosevelt win over Alf Landon, which refuted a projected Landon win by the heralded presidential poll of the era (non-random) sponsored by a national magazine, the Literary Digest.
After their winning result, the three pollsters became known as the “Trio of ‘36” and went on to conduct accurate polls in Roosevelt’s next two elections, 1940 and 1944. They gained personal renown and established the legitimacy of polling as an accurate gage of public opinion in high stake elections.
|“Trio of ‘36”: Archibald Crossley,
George Gallup and Elmo Roper
Life Magazine, 1944
All three men, and especially Gallup, became proselytizers for frequent polling, arguing that it helped counter special and well-off interests from dominating government. They saw polling as a way of bringing the public to the table when policy decisions were made, especially during the periods between elections when specific issues arose that might not have dominated the most recent election.
Gallup posed: “Shall the common people be free to express their basic needs and purposes, or shall they be dominated by a small ruling clique? In other words, how does one make those holding high public office responsive to the needs and wishes of the public?”
But the 1948 election shook the new industry to its core. All three pollsters predicted Thomas Dewey, the Republican, would win the election over incumbent Harry Truman. And, as Truman famously quipped, “That ain’t how I heard it.”
The leaders of the profession realized that mistakes were made and that changes were in order. They acknowledged that they had prematurely quit fieldwork weeks ahead of the election in 1948, believing it was over. Most importantly, they recognized flaws in their sampling procedures and shifted to improve their random selection methods.
AAPOR was created after the first organizing meeting in Central City. It became the professional association that established the rules and ethics of the polling industry. The rules required that published polls provide basic information to readers, such as the date of the poll, sponsor, population polled, sample selected, questions asked and margin of error.
Gallup Inc. became an international company with polls conducted in many countries. Crossley’s specialty was surveys of radio audiences, but he continued to poll for political parties, leaders and public policy. His family, including daughter Helen Crossley, also a renowned public opinion researcher, established the Crossley Center for Public Opinion Research at the Josef Korbel School of International Studies at the University of Denver. Roper operated a national market research firm and created a polling archive that became the Roper Center, first at Williams College, then the University of Connecticut and now Cornell University. Each made a lasting mark on a fledgling field that is now universally used to understand public preferences toward both domestic and international policies and leaders.