Back to top Jump to featured resources
Article filed under The Public Good

American Voice 2004: Is Announcing Exit Polling During the Election a Good Idea?

| Written by John Bailey | No Comments | Updated on Jun 1, 2004 The content that follows was originally published on the Institute for Local Self-Reliance website at https://ilsr.org/is-announcing-exit-polling-during-the-election-a-good-idea/

Q.  My friend says he’s going to wait to vote until late in the day on November 2nd. By then the media will have started reporting exit poll results and he can decide whether or not to cast his vote for a third party candidate. Is this a good idea?

Answer:

Putting aside the possibility that 10,000 other people in his state plan to do the same thing, your friend’s strategy is still questionable.

As in 2000, this is likely to be a very close presidential election. Four years ago, 537 votes decided the outcome in Florida, but other states were close as well: 7211 votes in New Hampshire, about 4000 in Iowa, 5700 in Wisconsin, 6765 in Oregon and just 366 votes in New Mexico.[1]

All of these numbers are well within the margin of error of even the most accurate poll, including exit polls. That’s a statistical term referring to the reliability of the poll numbers. This is often reported in these words, “accurate within plus or minus x percentage points”.

Exit polls are questionnaires filled in by voters after leaving their polling places. If an exit poll with a margin of error of two percentage points shows Candidate A leading Candidate B by 49-47 percent, Candidate A may have as little as 47 percent or as much as 51 percent of the vote. Candidate B’s vote total will be between 45 and 49 percent. Thus the poll, in effect, reflects the possibility that Candidate B could lose by as much as six points or win by two.

To put this in perspective, the margin of victory in New Hampshire in 2000 was one-tenth of one percent of ballots cast. In Florida, Iowa, New Mexico, Oregon and Wisconsin the margins were closer to one-hundredth of one percent.[2]

Exit polls, and polls of any kind, are statistical samples of the population. Though they are designed to represent the population as accurately as possible, there are always unknown factors. For example, this year election officials are expecting a larger than usual number of people to vote by absentee ballot. There has been a general trend toward more absentee voting, but in some states voters are choosing to vote on paper by absentee ballot rather than with voting technology they don’t trust. There’s no way to predict whether the previous guiding principle – that many more Republicans than Democrats cast absentee ballots – will be an accurate guide.

Exit polls can provide useful post-election information because voters are asked not only for whom they cast their ballot, but also what issues concerned them and several demographic questions. The data from exit polls, says David Iverson of Best Practices in Journalism, “allowed us to understand whether a candidate was favored by younger voters, union members or those who attended church. It told us whether the people who voted for a particular candidate cared more about education or tax reform.”[3]

Michael Delli Carpini of the Pew Charitable Trusts maintains that the predictive power of exit polls adds to the drama of election night coverage, but is the least important reason why exit polls are valuable. More important is the fact that they can “provide scholars with data that can be used for more systematic, theory-driven research.”[4]

A bigger question than whether your friend should wait to vote till the end of the election day is whether the media should wait to report exit poll results until the polls have closed nationwide. Many believe they should, but there’s no law that requires them to do so.

Exit polls have been used since 1964. In 1980, for the first time, NBC used exit poll results and some poll returns to call the election for Ronald Reagan at 8:15 Eastern time. President Carter actually conceded more than an hour before the polls had closed in California.

By 1982 all the networks were projecting results, but with the stated policy that they would not project elections in a state until the polls in that state have closed.[5] Four years later the networks again declared Reagan the winner before polls had closed in the West.

In response, at the urging of Western legislators concerned about the impact of exit polling on local and state races and referenda, the House and Senate passed a joint “sense of Congress” resolution asking broadcasters to voluntarily refrain from reporting results based on exit polls until all polls for the office have closed.

Broadcasters have largely accepted this recommendation. But a new twist was added to the exit poll debate in 2000, when online publications disclosed numbers hours before the polls had closed. In response the Voter News Service delayed the release of its findings by two hours to 4 p.m. Eastern time.[6]

During the 2004 primaries a debate broke out among some bloggers over whether it is ethical for online publications to reveal exit poll information before the networks.[7]

After the 2000 election in which inaccurate results based on exit polls were broadcast widely, a commission brought together by CNN concluded that it and other news networks had emphasized speed over accuracy. Election night coverage is highly competitive and exit polls can give a network a competitive edge

The National Commission on Election Reform, chaired by former presidents Gerald Ford and Jimmy Carter, condemned the practice of reporting exit poll results before the polls have closed in the Continental U.S.

The Commission argued that exit poll projections harm democracy by reducing voter turnout.

Others disagree. Slate editor Michael Kinsley says this thinking means, “essentially, that people should be tricked into voting by keeping them in the dark.”[8]

Jack Shafer, deputy editor of Slate, argues that the tales of would-be-voters on the West Coast leaving the polls in 1980 were never substantiated, even though CBS News assigned reporters to document the vanishing voters created by NBC’s projection.

Some in the media believe that exit polling is a free speech issue. “If Congress had its way,” writes Lawrence Grossman, former president of NBC News and PBS, “thousands of insiders in the press and political campaigns who have access to exit poll results would know who won the elections in each state, while only the public would be kept in the dark”.[9] News organizations would be deceiving people, he argues, “if we encourage them to cast their ballot by hiding the fact that the election had already been decided. Why should anyone’s vote for president be based on the false belief that their vote will count, if, in truth, one candidate has already received enough electoral college votes to win the presidency…?”

My position? The same as that of former Federal Communications Commissioner Newton Minnow. “[W]e’ve got to get beyond talking about whether we have a right to do something and, in the words of Supreme Court Justice Potter Stewart, ask whether it is the right thing to do.”[10]


NOTES

[1] Charlie Cook, “The Next Nader Effect”, New York Times, March 9, 2004.

[2] Footnote link no longer available

[3] David Iverson, “Meaty Vote Coverage Beats Hasty Results”, Newsday, January 16, 2003.

[4] Michael X. Delli Carpini, “The Functions of Exit Polls”, Political Communication Report, 13(2) Spring 2003.

[5] This policy has frequently been ignored in states with two time zones. Uniform poll times, both statewide and nationally, have been proposed as a solution to this. In 1982 Senator Samuel Hayakawa proposed a bill that would impose criminal penalties on anyone who releases results of the presidential election before polls are closed in all 50 states. The bill never made it beyond committee hearings. Since then several “sense of Congress” resolutions have been proposed reinforcing, without criminal penalty, the same idea.

[6] New York Times, March 8, 2000.

[7] For a review of the discussion see Jay Rosens’ PressThink

[8] Michael Kinsley, “Electoral Hypochondria”, Washington Post, August 3, 2001.

[9] Lawrence K. Grossman, “Exit Polls, Academy Awards and Presidential Elections”, Columbia Journalism Review, May/June 2000.

[10] Newton Minnow, “Jumping the Gun”, Issues in Ethics, Santa Clara University, Spring 1997.