January 6, 2010

IRV: The 3 Fundamental Problems

The University of Vermont conducted an assessment of Instant Runoff Voting (IRV) after the last election, which was prepared by Anthony Gierzynski, PhD, Associate Professor of Political Science.

Gierzynski identified "3 Fundamental Problems" with IRV and they are:

1. IRV discriminates against classes of voters by adding complexity to the ballot.

2. IRV has a very real potential to produce perverse outcomes or voting paradoxes that are not majoritarian.

3. IRV fails to address the real problem that arises when multiple parties compete in a two-party system.

To read the full assessment click here.

Let's face it folks. IRV sounded like a great idea and the majority of us voted for it (including myself) but the reality is IRV doesn't work.

7 comments:

Lea Terhune said...

I voted for IRV, and voted in subsequent elections. Realize it's easy to vote 1st, 2nd, 3rd, etc choices, but the run-off count is funny math, to be sure. People feel tricked; voters feel cheated; winner does not have confidence of voters who feel like they've been tricked and cheated! Then, if a crisis hits, as has happened repeatedly in Burlington since IRV was instituted, the city is vulnerable and the mayor is weakened by having been elected by IRV. Not a good situation.

Aaron said...

I'd like to address the first point, about how ranking more than one candidate doubles or triples or whatever-ples the cognitive cost.

The fact is, unless you've evaluated all the viable candidates, your choice of any one of them is meaningless noise. If we think the voters have anything meaningful to say, we're already expecting them to have opinions an all the viable candidates.

rbj said...

Lee, I understand that many people wondered what happened last March. and in fact, the tabulation method of IRV *did* screw up last March, but the screw up isn't that the Plurality winner was not elected. That's the whole point of a ranked-ballot in a multi-candidate race. It would have no reason to exist if it always elected the Plurality winner, we would just keep the old method.

but the tabulation method of IRV is screwed up and, in fact, it elected the wrong candidate. but the correct candidate was not Kurt just because he had the plurality of 1st-choice votes. that's where some of the anti-IRV crowd are just wrong.

Lea Terhune said...

"Wrong candidate," "correct candidate," no no no. Anti-IRV isn't about crystal ball visions of who would have won, should have won, or could have won. It is simply pro- "one person, one vote" and in the future one person will win with more votes than any other candidate. Whether the plurality is 39% or 40% or 51% is a phoney issue when only 27% of registered voters vote, and those pluralities represent about 12% of the registered voters!

rbj said...

"wrong candidate [to identify as elected]" --> a candidate without majority support of the electorate

"correct candidate [to identify as elected]" --> the candidate with majority support of the electorate

"one person, one vote"... is that what happens when there is a runoff and the one person returns to the polls on runoff day? maybe i'm just to simple, but that's "one person, two votes" and that was one of the intentions we had in adopting the ranked-order ballot in the first place. problem is, with IRV rules for tabulating the vote, sometimes a pathological result occurs and that happened in Burlington in 2009.

"Whether the plurality is 39% or 40% or 51% is a phony issue when only 27% of registered voters vote,..." no, it's not a phony issue because whether it's 27% or 12% or 90%, that is the franchise we have to our government. just because only 27% show up, why don't we just toss their ballots and use a dart board? or flip a coin? are you saying that the 27% that turn out are not to be taken as the representative voice of the electorate? if not, who?

if you *do* grant that 27% the authority to speak for the other 73% who didn't show, then again, how do you determine the will of that electorate (or 27% slice of electorate) from what they explicitly mark on their ballots? if a majority of that 27% agree that candidate A is better than candidate B, then should candidate B be elected, just because only 27% show up at the polls?

poor turnout is a bad thing. that is one reason why i am for Motor-Voter and against delayed runoffs or any other unnecessary hoops for voters to jump through to get to vote. election policy that unreasonably decreases convenience for voters decreases voter turnout. electing candidates with decreased voter participation cannot be as democratic as having increased turnout.

Stephan said...

There are no barriers to prevent people from voting in any election. It does not bother me if turnout is 27%, or 12%. Often people do not vote because they are happy with how things are going, and have no motivation to be involved in a decision one way or another. It's just like IRV, they might choose #1 for all candidates, they are OK with each of them so they don't vote.

That is their right. Believe me, they will stand up and be counted if they become too pissed off, but they are not. So, they leave it up to "the voters" to make the choices on election day.

You have to accept this fact.

Anonymous said...

"1. IRV discriminates against classes of voters by adding complexity to the ballot."

There is ZERO objective evidence that this is true, period.

"2. IRV has a very real potential to produce perverse outcomes or voting paradoxes that are not majoritarian."

IRV simply picks the candidate that a majority of voters feel most comfortable with, period.
---------------------------------
"but the run-off count is funny math, to be sure."

There is no "funny math" with IRV...unless one doesn't understand the concept of addition & subtraction that is...ugh...
-----------------------

"and in fact, the tabulation method of IRV *did* screw up last March"

"but the tabulation method of IRV is screwed up and, in fact, it elected the wrong candidate."

Wrong, wrong, wrong...Kiss was the candidate that the majority of Burlington voters were most comfortable with being their mayor, period end of story.
-----------------------

"It is simply pro- 'one person, one vote'"

ALL IRV voters get one chance to select their preferences for who should win a particular elected office, period.
-----------------

"poor turnout is a bad thing"

...and since IRV, Burlington voter turnout is UP, not down!