AAPOR
The leading association
of public opinion and
survey research professionals
American Association for Public Opinion Research

Why do polls talk about “likely voters”?
One of the values in American political culture is that “a good citizen votes.” But, as with many ideals, more people claim to vote than actually cast ballots. So one problem election pollsters face is that not all respondents who tell them they plan to vote will do so. Actual turnout (known only after the election) is generally lower than respondents’ self-reports of voting intentions in pre-election polls. So the pollster’s challenge is to try to identify those who will really vote on Election Day and which ones will stay home.

PERCENTAGE OF VOTING-ELIGIBLE POPULATION THAT VOTED

YEAR %
2012 59
2008 62
2004 61
2000 55
1996 52
1992 58
1988 53
1984 55

Source: The United States Election Project. The voting-eligible population is constructed by adjusting the voting-age population for non-citizens and ineligible felons, depending on state law. National estimates are further adjusted for overseas eligible voters.

Who are likely voters? How are they identified?
Pollsters often determine likely voters by giving each respondent a score, based on his or her answers to questions that have been shown to correlate with turning out to vote, such as interest in the election or past voting behavior. They then establish a cutoff based on what they think turnout is likely to be. For example, if a pollster believes that turnout will be 60%, they would take the 60 percent of respondents with the highest scores on their question index as “likely voters.” Although many pollsters use such a scale, the component questions that go into the scale differ. This is a reason why election polls can report different results.

Research finds no single magic bullet question or set of questions that can determine likely voters with 100 percent accuracy. Thus, different organizations have different ways of identifying who they believe likely voters are. Most polls ask a combination of questions that cover self-reported vote intention and measures of engagement. In addition to past voting behavior and interest in the election, a likely voter screen may include questions to determine if the respondent is registered to vote, how likely they think they are to vote and how closely they’re following the election.

A second challenge facing pollsters is estimating what turnout will be, which may affect estimates of candidate support. In Minnesota, for example, here's the percentage of eligible voters turning out in recent presidential elections:
 
Year Turnout %
2012 76%
2008 78%
2004 79%
2000 70%
 
Based on the 2012 data, a pollster might expect turnout to be around 76 percent in 2016. But, turnout was slightly higher in 2004 and 2008 and lower in 2000. So, what should be used as the expected turnout figure: 70 percent, 79 percent or somewhere in between? Suppose a choice of a 70 percent cutoff point gives an estimate of a Republican advantage of 4 percentage points. But when expanding the expected electorate to 79 percent, the data may show a Democratic advantage of 3 percentage points. This points to another source of possible difference between polls: whether the level of support for each candidate changes as the “likely” electorate gets smaller or larger — or if it makes no difference. This means that the definition of a likely voter can be somewhat of a moving target, depending on the turnout assumptions of each particular poll.

It can be helpful to ask a pollster what level of turnout their poll is estimating as a percentage of the voting-eligible population. For example, if a poll is estimating turnout that is much higher than recent past elections, this means that the pollster used a loose likely vote screen, and the poll may be  including many more people than will ultimately turn out to vote.

How much faith should I put in pollsters’ reports of likely voters?
There is a consensus in the polling community that it is better to report “likely” voters than “registered” voters, especially as Election Day approaches. But having acknowledged this, there are some cautions:

First, people change in their commitment to voting as the campaign unfolds. Respondents are probably better able to tell if they really are going to vote as it gets closer to Election Day. One reason the polls get in closer agreement as Election Day nears is that their “likely voter” formulations behave more similarly. It is easier to identify true “likely voters” the closer it is to Election Day. This is one reason that pollsters report results among registered voters early in the campaign and switch to likely voters closer to the Election Day.

Second, there is no magic formula for identifying likely voters. There are a number of indicators, but the combination of them will vary across polling organizations. As of now, there’s no clear right or wrong formula.

Third, no one knows what actual turnout is going to be. It’s like trying to hit a moving target. A polling organization might make a prediction based on a turnout of 70 percent of eligible voters, but there are high- and low-interest elections. The past is not a perfect guide to the future. No one knows what’s actually going to happen. That’s why we have elections.

So, pre-election polls of likely voters are “best estimates.” Good ones - as the history of numerous polling organizations over numerous years has shown — but estimates nonetheless. Which is why it’s always prudent to maintain proper caution and respect for nuance when reporting on these polls.

Download PDF Version