Ignore Those Exit Polls!

Sample sizes and enthusiasm gaps can skew results
By Nick McMaster,  Newser Staff
Posted Nov 4, 2008 1:25 PM CST
Krystal Sipp votes at Cherry Creek Town Hall in Cherry Creek, NY, Tuesday, Nov. 4, 2008.   (AP Photo/David Duprey)
camera-icon View 2 more images

(Newser) – Exit polls can be addictive for those who can't take Election Day suspense, but poll analyst Nate Silver recommends you resist the temptation. Here's why:

  1. Exit polls have high margins of error because they are taken at only a few precincts in a state.
  2. Exit polls overstated Democratic support in the 2000 and 2004 elections.
  3. Exit polls overstated Obama’s support by an average of 7 points in the primary season.

  1. Theoretically, exit polls are a random sampling of the precinct because the pollster approaches every nth voter to leave, but in practice this is hard to stick to.
  2. Democrats are more likely to agree to take exit polls, especially when enthusiasm in the party is high.
  3. It is difficult for exit polls to adjust their results to reflect early voters.
  4. It is common for exit polls to miss those who voted late in the day.
  5. “Leaked” exit polls may be fabricated or may be so early that they have small sample sizes.
  6. High turnout, especially among groups who usually vote in low numbers, may not conform to the exit polls' turnout predictions and skew their precinct sampling.
  7. You can just wait a few hours and get the real results.
(Read more Election 2008 stories.)