Polls, Pre-Election

PRE-ELECTION POLLING IS a scientific effort to estimate what the entire electorate thinks about candidates and/or issues by asking some members of the electorate (a sample) for their opinions. The presumption is that such a sample, if properly drawn, will yield findings that are representative of the big picture. Pre-election polling has become more important in American politics, because the news media often report on polls (and use such polls to determine the extent of their coverage of campaigns and candidates) and political campaigns rely on their pre-election polling to develop campaign messages and political strategy.

Scientific pre-election polls did not exist until the 1930s. Prior to that time, pre-election polling took the form of non-scientific straw polls. Unlike scientific polling, there is no effort to obtain a representative sample of the larger population in a straw poll. Such straw polls date back to the 1820s. In 1824, a straw poll taken in Wilmington, Pennsylvania, found Andrew Jackson to be the people's choice. Straw polls would become increasingly popular throughout the remainder of the 19th, and first few decades of the 20th, centuries.

In 1888, the term dark horse was used for the first time by the Boston Journal newspaper to describe a candidate other than one of the leading contenders who emerges to win an election. This would eventually become known as horse race journalism. During the early 1900s, a number of newspapers started conducting [Page 537]straw polls. These included the New York Herald, Cincinnati Enquirer, Columbus Dispatch, Chicago Tribune, Omaha World Herald, and Des Moines Tribune.

In 1932, George Gallup, an executive with an advertising agency, conducted a public opinion poll for his mother-in-law, who was running as a Democrat for the secretary of state in Iowa. Gallup used techniques he had developed while working on his Ph.D., and his mother-in-law was elected. The following year, Gallup decided to experiment with scientific polling by applying many of the methods of market research to politics. Gallup had examined voting records in the United States dating back to 1836, and sent out questionnaires to a sample of voters in each state. When he came within one percent of the actual results of the 1934 mid-term congressional elections, Gallup decided to establish the American Institute of Public Opinion to poll for the 1936 presidential election. In 1935, the Institute started a syndicated newspaper column, “America Speaks,” which reported on how Americans felt about the issues of the day by conducting polls.

As Gallup continued with his experiment, the Literary Digest, a general interest magazine that had published the result of their presidential straw polls since 1916 (when they correctly predicted that Woodrow Wilson would be re-elected), carried out their straw poll. This poll, based on responses from more than two million people, predicted that Republican Alf Landon would defeat President Franklin D. Roosevelt in a landslide. The straw poll predicted that Landon would receive 57 percent of the popular vote. The actual result was the complete opposite: Roosevelt carried 46 of the 48 states in the union (Landon won Maine and Vermont). This poll was inaccurate because Literary Digest drew its sample by taking names from telephone directories and automobile registration lists. Given that the nation was in the throes of the Great Depression, and that people who could afford cars or phones were very wealthy, the Literary Digest sample was not representative of the larger electorate. As a result of this failure, the Literary Digest lost credibility and, by 1938, the Funk and Wag-nails Company had sold the magazine to Time Magazine, which absorbed it.

...