For those watching the election results tonight, I have a word of warning: beware of news reports based on exit polls. Here is how ABC News describes the traditional methodology:
Interviewers stand outside polling places in randomly selected precincts and attempt to talk to voters on their way out at specific intervals — every third or fifth voter, for example.
Voters who agree to participate in the poll fill out a short questionnaire and place it in a ballot box.
Interviewers phone in results three times during the day. When a voter refuses to participate, interviewers note the gender and approximate age and race of that voter. In this way, the exit poll can be statistically corrected to make sure all voters are fairly represented in the final results.
That is how exit polls have been conducted by Edison Research since 1972. In the past, all major news outlets have participated in reporting on their results. The case for doing exit polling this way is largely because it captures people who have definitely voted. But several problems have emerged:
- In 2014, more than a third of Americans cast their votes before election day; in 2016, it was 40 percent.
- Exit polls have traditionally over-counted Democrats, which is what happened in 2016.
- While the random sample of polling locations does an adequate job of capturing vote totals, it’s reliance on geographic locations reduces the accuracy of demographic data about voters. That has been the issue highlighted by groups like Latino Decisions in the inaccuracy of reporting the Latino vote.
Due to some of these concerns, this year the Associated Press and Fox News will join several other major news organizations, like the Los Angeles Times and the Washington Post, in substituting a new strategy conducted by NORC at the University of Chicago called VoteCast.
Under AP’s system, postcards are mailed to a random sample of registered voters in 25 states, inviting them to take a survey either online or by phone. At the same time, a random-sample survey of registered voters nationwide is conducted using a panel designed to be representative of the U.S. population. Finally, VoteCast surveys self-identified registered voters in all 50 states using opt-in online panels.
Those who take the survey are asked if they have, in fact, voted or are sure to.
When you hear reporting based on exit polls, it will be important to know whether it is based on the traditional method or VoteCast. In the months ahead, the two systems will be researched and compared, but at this point, we’ll be getting results from the old inaccurate system and an untested new system. It’s probably best to take all of them with a grain of salt.