This past November 2014, we at The Center for Election Science (with generous help from some Maine residents) conducted a exit poll for the Maine gubernatorial election. Our goal was to compare ordinary vote-for-one “plurality” voting to alternatives. We asked voters:
- Who would you have voted for using approval voting, where you can vote for as many candidates as you wish?
- Who did you actually vote for?
- How would you have ranked the three candidates in order of preference (1st, 2nd, 3rd)?
In the real election, Republican Paul LePage was re-elected. With approval voting and instant runoff voting, independent Eliot Cutler won. LePage actually came in last place with approval voting.
When we initially mentioned this to FairVote executive director Rob Richie, he replied via email (on Nov 14, 2014):
Interesting — thanks. Do you describe the reweighting process somewhere and walk through the actual results in the ballots you had and how they were adjusted into this final result?
Then, on Nov 20, he randomly replied:
This exit poll isn’t close to science. It truly is embarrassing for CES.
He cited not one iota of evidence for that claim. No criticism of any of our statistical model. Nothing. This is interesting, given that our statistical analysis was performed by a Princeton math PhD with over a decade of experience focusing on elections, as well as a computer scientist who formerly worked in the D.C. office of Maine’s independent senator Angus King. Not exactly a fly-by-night operation.
I stressed to Richie that the poll also was a positive for his favored instant runoff voting system, but he replied, “We won’t touch it with a 10-foot pole.”
Apparently, some of his colleagues in the voting reform world have different standards. On Dec 8, the Rank Your Vote Duluth campaign posted on Twitter:
Such a terrific article from our friends who are gaining steam in Maine
And on Dec 9, FairVote Minnesota posted:
“Independent Eliot Cutler Would Have Won Maine Governor’s Race under Approval Voting”
That link was later deleted. Interesting.