Good Polls, Bad Polls - And Tom Barrett's Polls
Before we move on from Governor Walker's smashing victory, I think it's appropriate to look back on the pre-election polling. In general, the pollsters got it right.
The final liberal leaning PPP poll had the margin at 3%. Close but no cigar. The fact that they missed it as they did was probably a combination of wishful thinking - and a peculiar decision to drop the third candidate from their last poll. This seems to have had the effect of artificially increasing Barrett's support and garnering him headlines showing the race tightening (when it really wasn't).
In general though, I think it's fair to say that pollsters distinguished themselves.
Except, that is, for Tom Barrett's pollsters.
You remember Tom Barrett? The guy who claimed to have polling three weeks before the election showing the race was "a dead heat". Or who claimed to have polling one week before the election showing the race was "a dead heat".
So, what happened? Was Barrett lying about polling results in an effort to try to inspire false confidence in his supporters - or did Barrett simply hire hack pollsters?
My guess is that this was a classic example of how campaigns manipulate statistics.
When polling for candidates, it's not uncommon for pollsters to make estimates based on voter turnout. For example, candidates will get estimates showing the results if 45% of the electorate shows up, if 50% of the electorate shows up, etc. I don't know for sure but my guess is that Barrett's claims were based on ridiculously high turnout estimates and therefore had little basis in reality.
On the other hand, Barrett's pollsters could just suck!
Still, all in all, Tuesday was a pretty good day for the pre-election pollsters.
Now, the exit polls ... well, that's another story!