Beware the Exit Polls...

Mark Halperin lists three legitimate reasons to be skeptical of exit polls in today's contests:

1. They are not closed contests open only to Democrats.
2. Turnout is going to be huge (probably record breaking).
3. The absence of recent competitive primaries.

And I'll offer a tale of how exit polls differ from reality:

In November 2004, campaigns saw the first signs of exit polls from VNS (Voter News Service).  Called up a friend to let him know that victory was within our grasp.  We had slain the dragon, we were storming the castle, justice would be restored by way of a Democratic administration.

Then the actual returns came in.  We weren't winning Florida, in fact we were losing by some 400,000 votes.  Democrats barely defended Oregon, which was supposed to be a dark blue state.  We weren't winning Ohio - we ultimately lost it by 80,000.  And a whopping .4% is the margin by which Democrats held Wisconsin.  My friend kept the voicemail and played it back for weeks on end after that.  Good comedy.  Bad reality.

Moral of the story: exit polls are worth what you pay for them.

Comments

[Walker]

Alternately, exit polls reflect accurately what the outcome _should_ be, but do not reflect the outcome programmed into Diebold and Sequoia voting machines. I am privy to several cases involving these companies and what happened in Ohio and Florida almost certainly constitutes election-rigging (or, as my favorite band would put it, 'electioneering').

[The Grand Panjandrum]

If memory serves me correctly the exit polls usually favor Obama about +7.

I remember Al Franken telling the story of being at the Kerry HQ and everyone being so excited early on. And then thing sort of went downhill for the rest of the night.

Exit polls are always better for looking back to tell a story rather than looking forward to predict what the story will be.

[Mr Super]

In these elections, curious to know if we may see signs of the the "Bradley Effect." For more information on this political theory, click here:

http://en.wikipedia.org/wiki/Bradley_effect

[The Grand Panjandrum]

I wished I could remember where I read an excellent write about the so-called Bradley/Wilder Effect and why it was probably not a significant contributor in this primary season (or at that time it wasn't).

I'll search around and see if I can find it. But it convinced me that it was not a significant factor in the election.

[Mr Super]

Please do - would love to see it.

[Charles]

This is the best I remember, from Eugene Robinson in the WaPo just after NH: http://www.realclearpolitics.com/articles/2008/01/on_the_lookout_for_the_bradley.html

[Martin Vagues]

I am clearly for HIlary . it take s a very experienced person to run this country like what Bill Clinton did. Obama is too immature and not strong enough. We need a strong person and thats HILARY.