The Blackwell Encyclopedia of Political Science

Authors: Bogdanor
Summary: In 920 words, this entry describes the origins of opinions polls, discusses the polling methodology, and analyzes the political significance of opinion polls.
Opinion polling as we know it today originated in the 1930s. In fact, as early as 1936 both the Gallup and Roper polls correctly predicted the presidential winner. The 1960s marked the start of widespread polling by the mass media and the political parties for election purposes. Over the last thirty years the whole market and opinion research industry and the use of polls has grown tremendously.
According to the entry, the typical opinion poll is based on interviews with a randomly drawn sample of approximately one thousand members of the electorate. Most polls claim accuracy within three percent, and most polls are reasonably accurate. However, there have been some spectacular mispredictions. For example, Dewey was mistakenly predicted the winner in the 1948 US presidential election, and there were a number of mistaken predictions during the 1970s in Britain.
Opinion polls are both frequently attacked and praised. Polls and the predictions based on them are often critiqued as influencing election results. Similarly, they are often used by government and party leaders strategically to time elections and policies--behavior seemingly in conflict with democracy. When used responsibly and carefully, opinion polls provide vital feedback between citizens and decision makers. Such feedback is particularly important in a vast, representative democracy like the United States.