Skip to main content icon/video/no-internet

Undecideds, in election campaign lingo, are voters who have yet to decide which candidate—or side of an issue or referendum—they will support on Election Day. In pre-election polls, there are several ways of measuring support for a candidate or issue and, con-comitantly, the number of undecided voters. A few researchers argue for an open-ended question format, in which the respondent volunteers the name of his or her preferred candidate. Answering such open-ended questions requires respondents to recall candidate names on their own, which means that these questions generally have larger percentages of “don't know” responses than closed-ended questions in which respondents hear or read candidates' names.

More frequently, however, researchers use a closed-ended “trial heat” question because the closed-ended measure has proven over time to be most accurate if comparison to Election Day results is the yardstick. These “read-list” questions in a telephone survey typically list all the candidates—or at least the major party candidates in a race—and ask whom the respondent would vote for “if the election were held today.” Some research suggests that, especially in state and local races, it is important to list all candidates and their parties, rather than asking only about major-party candidates or about major-party candidates and the nonspecific ideal candidate, “someone else.” For those who were initially undecided or refused to express a candidate preference, the question would attempt to ascertain whom the respondent leaned toward supporting. By combining these “leaners” with the original “choosers,” the question determines overall support.

At least three other types of respondents are interviewed in pre-election polls: (1) those who say they will vote in the election but won't cast a vote in one race or another, (2) those who refuse to say whom they will support, and (3) those who are truly undecided. Often election pollsters combine the three types for reporting and analysis purposes into a category called “no opinion.”

Proportions of Undecided Voters

Generally, the percentage of undecided voters in preelection polls dwindles the closer to the election that the polls are taken. This certainly is the case in presidential elections. In local elections, however, the proportion of undecided voters often is quite large right up until Election Day. Any number of factors could explain the relatively large or changing size of undecided voters: low candidate name recognition, low salience of the election or issue, or events that cause a dramatic shift in support late in the campaign.

One such event is the death of the incumbent candidate during the campaign, especially when the death occurs late in the campaign, such as the 2002 plane crash that killed U.S. Senator Paul Wellstone (D-Minn.). Another such event would be the withdrawal of a candidate from the race because of ethical revelations late in the campaign. Some researchers also suggest that a large proportion of undecided voters late in the campaign may indicate a faulty likely voter model.

Examining demographic, geographic, political, and attitudinal profiles of undecided voters can provide useful information to campaign and party pollsters, who want to know what it might take to persuade undecided voters to support their candidate. Pollsters for news organizations do similar analyses because understanding undecided voters is an important part of the news story, particularly close to Election Day.

...

  • Loading...
locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles

Sage Recommends

We found other relevant content for you on other Sage platforms.

Loading