Democracy Wins

Paul Bass PhotoWith the help of a spare cardboard box and some higher-order math, one neighborhood tested a new way of choosing candidates—when the choice is between more than two.

The test took place in the Edgewood School cafeteria Tuesday night. Members of the Democratic committee in Westville’s Ward 25—often the city’s highest-voting district—gathered to endorse a candidate for mayor in a new way. All 30 ward committees around town are holding these straw polls this week and next, usually a quick, straightforward process. Then each ward’s two co-chairs go to a city convention to cast votes for the town committee’s official mayoral endorsement.

Since a divisive 2011 election, Ward 25 has worked to unite the neighborhood by trying to include all factions in decisions, inviting everyone to participate, and conducting business in the open. A few months ago, members noticed that almost all seven of the original 2013 Democratic mayoral candidates had visible support in the neighborhood. Members decided they didn’t want to make an endorsement based on the wishes of a minority of the committee. Neighbor Scott McLean, a Quinnipiac University political scientist, offered a solution: Imitate Portland, Maine; Oakland, Calif.; and the Australian House of Representatives. Try “instant run-off voting” or “IRV.”

That means having people fill out a ballot by ranking the candidates in their order of preference. Then ballot-counters tally up people’s first choices in a first round. If no candidate wins an outright majority, the candidate with the least first-choice votes is eliminated. Then ballot-counters start a second round. They examine the votes of people who had voted for that losing candidate; the ballot-counters look at those people’s second choices, then add those second-choice votes to the other candidates’ first-choice vote totals. That process continues until a candidate gets a clear majority of the vote.

The idea: You vote for people you actually want, rather than strategically voting for the lesser of evils. And no candidate wins simply because two other candidates split a majority of votes.

Ward co-chairs Janis Underwood and Mike Slattery (pictured at the top of the story) agreed to abide by the ward’s wishes at the town committee convention. They agreed to vote in the first round for the winner of the IRV poll—even though Slattery is publicly supporting Justin Elicker for mayor. (Underwood has not declared support for a candidate.) “I was really concerned by how the committee was going to express itself. I really wanted to hear what the ward was thinking” and reach consensus, he said.

In advance of the candidate-selection vote, the ward committee also organized a lively May 20 mayoral debate in the neighborhood so people could hear all candidates at once.

Then members met in June for a dry run to practice IRV in a mock election—for food rather than for candidates. IRV turned out more complicated than it sounds. Members had everyone vote for their favorite dinner entree; seven choices appeared on the menu (as on the then-mayoral menu). No clear winner emerged intitially. It took five rounds for beef to edge out “vegan” to win. The committee also voted for three forms of creamer (see tally sheet at the top of the story) to practice for choosing among the three candidates for city/town clerk. The practice round also taught organizers that manually counting and recounting the ballots in one pile got too confusing. So in preparation for the real vote this Tuesday night, Slattery brought in the box from a recently-purchased Ikea bedside table to use as a grid on which to place multiple piles of ballots to be recounted in multiple rounds.

Thirty-eight of the ward committee’s 46 members gathered around elementary-school-scale lunch tables to listen to a refresher before voting.

They also watched this three-minute Youtube video that explains IRV. Somehow the cartoon characters made it sound simpler.

Then members wrote down all their choices for mayor and city/town clerk, passed by list-checker Laurel Underwood-Price ...

... and dropped their ballots in an old-fashioned ballot box.

Westville’s unofficial mayor, Gabe DaSilva, tabulated the clerk race results ...

... and found that no one had a clear majority. Sergio Rodriguez (who lives in upper Westville’s neighborhood Ward 26) had the most votes, 16, but only two more than Mike Smart, and less than half of the 36 votes cast for all candidates. (Two people voted for no one, who didn’t count.) On to round two. Ron Smith, who received the fewest votes, was eliminated, his voters’ second choices added to the other candidates’ totals. It turned five out of his six voters had picked Rodriguez as their second choice. That put Rodriguez comfortably over the top with 21 votes, a clear majority. If Smart had been Smith’s voters’ second choice, then he would have won the nod.

On to the main event. Professor McLean and City Hall staffer Rebecca Bombero re-examined the piles in Slattery’s IKEA box to focus on the mayoral results.

The results: Twenty votes for state Sen. Toni Harp. Ten for Alderman Justin Elicker. Five for former city economic development chief Henry Fernandez. Two for Hillhouse Principal Kermit Carolina. One for plumber Sundiata Keitazulu. No need for a second round: Harp had pulled out a clear majority.

That came as an anticlimax to some of those assembled—not because of their feelings toward any of the candidates, but because the idea of multiple rounds sounded kind of ... fun. “We really enjoy democracy in this ward. We really enjoy voting and the whole process,” said Jessica Feinleib (pictured with state Rep. Pat Dillon).

“Mike and I will now go to the town committee and cast your votes as you have indicated,” Underwood announced. To which Slattery added: “Thank you. That was fun.”

The same night, the Democratic committee in another high-voting district, Morris Cove’s Ward 18, made its endorsement. State Sen. Martin Looney called for a voice vote. It went unaninmously for Toni Harp.

Demcoratic Town Chairwoman Jackie James said no other ward committee is trying out IRV this year. But she said change is in the air. She said the town committee has formed a working group to look at amending its bylaws for the first time in decades. Including rules for how ward committees endorse candidates. “We’re going to consider all possibilities,” she said. “We want the process to be transparent and inclusive of everyone.”

Post a Comment

Commenting has closed for this entry


posted by: anonymous on July 10, 2013  12:18pm

Quite a few people on this Ward Committee (including the Alderman) are directly employed by the Yale Unions / UNITE HERE. 

It’s interesting to note that if you assume that those paid staffers all voted for Harp and then subtract those votes, Harp did not receive the majority of the total vote in the first round of voting.

Ward Committee votes are not particularly important given the above situation where Committees are “stacked” with paid staffers who happen to live in the Ward, but hopefully Jackie James’ move towards transparency will also allow residents to determine this type of information.

posted by: THREEFIFTHS on July 10, 2013  12:31pm

Get rid of the crooked Town committees and let the voters do the electing with the use of IRV.

posted by: Curious on July 10, 2013  12:43pm

I would completely support IRV in the general election.

posted by: robn on July 10, 2013  1:07pm


It is contrary to the by laws of the Democratic National Committee, and the Democratic State Committee to hold anonymous votes. This vote was contrary to party rules and must be redone with votes attributed to each committee member.

posted by: Jjunior on July 10, 2013  1:21pm

Very well done article, and I like Threefifth’s suggestion.

posted by: Noteworthy on July 10, 2013  2:48pm

One other Transparent Note:

Did Ward 18 hold a candidate forum? Interview anyone? When was the last time somebody unconnected to Looney came on the committee? There should be no voice votes. There should be ballots. Well, that assumes the vote wasn’t cooked.

posted by: New Haven Taxpayer on July 10, 2013  3:05pm

3/5ths is making a difference!
baby steps 3/5ths, I never thought I would see this day.

posted by: Dale Sheldon-Hess on July 10, 2013  3:26pm

“You vote for people you actually want, rather than strategically voting for the lesser of evils.”

Unfortunately, this isn’t entirely true. It’s easiest to show with an example. With 3 candidates, A, B, and C:

45%: A > B > C
10%: B > A > C
15%: B > C > A
30%: C > B > A

If you look at just the relative ranks of A and B, you’ll see that B wins 55:45. And if you look at just the relative ranks of B and C, you’ll see that B wins again, 70:30. So, against either one of these opponents, B would be the one the voters actually want.

Who wins by IRV? A.

The C-preferring voters have a choice though; they could avoid having A (their least-favorite) win if they rank B above C. This is the exact same “lesser evil” choice faced by voters with plurality voting. (Equivalently, candidate C could drop out of the race.)

This is precisely why IRV was rejected, after two elections, in Burlington, VT, because it failed to live up to these promises.

There are voting methods where you actually never have a reason to vote against your favorite; score voting and (somewhat simpler) approval voting. More information available at

posted by: robn on July 10, 2013  3:53pm

Not 3/5. Alderwoman Erin Sturgis-Pascal made a ward level IRV effort in 2007.

Also, DHS;;; im pretty sure your math is flawed. Check it again.

posted by: Righteous Cyclist on July 10, 2013  4:15pm

DHS, you fail to understand the basic concepts of IRV. I sincerely hope your comment doesn’t troll anyone into thinking IRVs don’t work.

posted by: Gretchen Pritchard on July 10, 2013  4:32pm

This method was used years ago for student government offices at the colleges my husband and I went to—Quaker-affiliated colleges Bryn Mawr and Haverford, which attempted to do things by consensus where practical and used preferential voting when it was not.  I wouldn’t be surprised if it is still in use there, and for the same reasons.  It’s fair, and it works.  All it requires is some training for the people who count the votes.

posted by: Hieronymous on July 10, 2013  5:00pm

Assuming that the purpose of IRV is to allow a voter to support his preferred candidate without fearing that his vote is wasted, it works. Seems like DSH’s point is that it does not prevent strategic, insincere voting, which seems pretty hard to dispute. Forget DSH’s hypothetical; this very election illustrates the potential for strategic voting. If I’m one of the 38(!) members of this ward committee, which is apparently one of 30(!!) in the City, and I strongly support Elicker, I give him my #1 vote and Keitazulu my #2. If I’m afraid of Harp getting this silly endorsement, I give her my #5, even if she’s actually my second favorite candidate. IRV certainly doesn’t prevent this kind of gamesmanship, but because this vote didn’t go to subsequent rounds, it would appear that the tally does represent the sincere votes of this committee, meaning that Harp is the most preferred candidate of 20 out of 38 random people in Westville. Stop the presses!

posted by: THREEFIFTHS on July 10, 2013  6:01pm

posted by: robn on July 10, 2013 3:53pm
Not 3/5. Alderwoman Erin Sturgis-Pascal made a ward level IRV effort in 2007.

Sorry But she just made a a ward level IRV effort.I have be working with FairVote,The Center for Voting and Democracy for years to get not just IRV,But also Term Limits and the system of proportional Representation for city state and federal levels.

posted by: Mike Slattery on July 10, 2013  6:08pm

@robn, we were sensitive to the fact that this is not official procedure, nor an official result.  As I understand it, delegations must operate by public vote.  A New Haven Ward Committee is a proper delegation for the purpose of the ward’s Aldermanic endorsement.  That was performed by public vote.  Adam Marchand was approved by that process.  Not operating as an official delegation for the citywide recommendation, we came to this process together for voluntary guidance.  It was clear to the 25th Ward Committee in previous meetings that this experiment was technically non-binding.

With that said, this was an exercise in legitimacy and consensus building that came from the Committee members, not handed down, and I am moved by the insight, attention and effort the whole Committee gave to it.

It is true that in one tally, a candidate got exactly half plus one, so we didn’t need more rounds.  That doesn’t mean there was no benefit.  I saw a reorientation toward the primary contest that sponsored more goodwill.  The conversations that arose from this process were interesting, and possibly unlike those we’d have if we were faced with a plurality vote.  People were more eager to talk with obvious supporters of Candidate X about Candidate Y in terms of acceptance, how much common ground existed, etc. in contrast to seeing a given bumpersticker and moving on in every case.  So we had plenty of politicking but it largely positive in nature.  I was very happy to be a part of it, and thankful for my Co-Chair and this Committee.

posted by: robn on July 10, 2013  6:33pm


Congratulations. I support the cause also, but point of fact, ESP actually did this exact thing in New Haven a half decade ago (not just advocated it).


Party rules forbid anonymous ballots. I don’t have a problem with the IRV ( I support it) but if you did it with written anonymous ballots its illegitimate by party rules.

posted by: robn on July 10, 2013  6:49pm


Just to re-emphasize, no matter what your intentions and whether you considered the vote to be “official” or “non-official”, secret ballots at any Democratic party meeting is strictly verboten by the Democratic Party. I’ll quote the national rule (state rule is the same, as should be the local).

Article 9, Section 12. All meetings of the Democratic National Committee, the Executive Committee, and all other official Party committees, commissions and bodies shall be open to the public, and votes shall not be
taken by secret ballot.

posted by: cedarhillresident! on July 10, 2013  7:21pm

I like this idea as well and bravo for giving it a shot. I think it should happen in the mayoral elections.

BUT on a ward level hmmm remembering that the co chairs (many appointed by susie voite at the time and or jackie james…and these co chairs get to pick the 50 members that get to vote on the ward committee. So does the vote represent the ward?? Or that circle of people? It could if you have people on it that can actually vote the way their community want and set aside their personal views. But most of these folks are from the same circle of people and remembering the number of independents in this city defiantly does not represent the ward as a whole or even the democrats of these wards. All this has become is a horse and pony show.

But again if we can do it in the general I would be a happy camper.

posted by: Eddie on July 10, 2013  7:31pm

I absolutely love that the ward committee used IRV.  I agree with Threefifths that it would be ideal for all elections to use a similar method for aggregating votes.  A big thanks to Janis and Mike for undertaking the work to institute this method for our ward!!  It sounds like it was administered smoothly and generated an excellent discussion. 

Anon, I find it offensive that you claim union members should not get to vote on ward committees or that Ward 25 is “stacked” with union members.  You participated in this meeting right?  As a committee member, surely you understand all of the work the ward has undertaken to find engaged and active community members for the committee.  Surely, you appreciate all of the work that was undertaken to produce a meaningful ward vote.  Who exactly are you claiming stacked the ward committee?  Who exactly are you claiming shouldn’t have voted?

Dale in your example the outcome of A winning still doesn’t seem so bad.  It is the first preference of 45% of the voters, and for 55% of the voters it is the first and second preference.  B is only the first preference of 25% of the voters and again for 55% of the voters it is the first and second preference.  Your right that there could be an incentive for strategic voting but in situations where preference orderings aren’t clear the opportunity for strategic voting is diminished and risky.  Even if runoff voting has some odd outcomes for certain preference configurations, I think it desirable to first past the post. 

Hieronymous your suggested strategy isn’t rational.  In your example, the only only way that Harp would be endorsed is if Elicker was eliminated.  In this case, you would presumably want your second choice to be endorsed.

posted by: Dale Sheldon-Hess on July 10, 2013  7:33pm

Thank you, Hieronymous, for an on-the-ground corroboration of how IRV can fail, from your own point of view.

robn, RC: My math is correct. This is a very well-known failure mode of IRV, undisputed in the professional literature.

If you’d like to explain how you think I’m wrong, rather than just asserting it, I’m willing to listen. But I think, as you work through the example, that you’ll see it’s just as I’ve described.

(Or you try looking through FairVote’s manual for “core support.” That’s the best counter argument they’ve been able to come up with, but it doesn’t directly refute my argument. Rather, it argues that A _deserves_ to win in this example; I don’t buy that.)

posted by: THREEFIFTHS on July 10, 2013  8:46pm

Every voting system has problems.But plurality voting supports 2 party domination
and can easily give rise to one party control.

posted by: abg22 on July 10, 2013  10:01pm

Mike and Janis should be commended for the exemplary way they run the ward committee. I only wish every ward committee in the city were as well-run, transparent, and willing to try new things (like IRV). Bravo!

posted by: chapillsbury on July 10, 2013  10:04pm

The Green Party uses IRV in all its elections. It’s great to see that some Democrats are catching on. This primary election and the general election that follows both cry out for IRV. Too bad that democracy is winning in only one ward, and not citywide.

The rest of us will be losers if the next Mayor of New Haven is elected to office by less than 50% of votes cast. The genius of IRV is that it mandates that the winner have the support of at least 50.1% of the voters. It would be great if one day we could add New Haven to the list of cities below, which use IRV.

Cities using instant runoff voting as of May 2012 -

  Berkeley, California: Adopted in 2004 and first used 2010 (for mayor, city council and other city offices)
  Hendersonville, North Carolina Adopted and used as part of a pilot program in 2007, 2009 and 2011 (mayor and multi-seat variation for city council)
  Minneapolis, Minnesota: Adopted in 2006 and first used in 2009 (for mayor, city council and other city offices)
  Oakland, California: Adopted in 2006 and first used in 2010 (for mayor, city council and other city offices)
  Portland, Maine: Adopted in 2010 and used in 2011 (for electing mayor only)
  San Francisco, California: Adopted in 2002 and first used in 2004 (for mayor, city attorney,  Board of Supervisors and most other city offices)
  San Leandro, California: Adopted as option in 2000 charter amendment and first used in 2010 (for mayor and city council)
  St. Paul, Minnesota: Adopted in 2009 and first used in 2011 (mayor and city council)
  Takoma Park, Maryland: Adopted in 2006 and first used in 2007; (for mayor and city council)
  Telluride, Colorado: Adopted in 2008 and first used in 2011 (for mayoral elections)

Upcoming implementations

  Memphis, TN (adopted 2008; scheduled for 2013 for electing city council and other offices)

posted by: swatty on July 10, 2013  10:35pm

Ghosts in the MACHINE.

posted by: anonymous on July 10, 2013  10:55pm

Eddie, there is a difference between “union members” and “paid staff.”  A moderate share of New Haven consists of union households, which is great, but a very small share consists of paid staff from organizations that, for example, say publicly that Yale listens to them only because they purchased the entire Board of Aldermen.

Before evaluating endorsements, the public should have access to information on which paid campaign staff, city employees, and contractors are also voting, as well as which politicians are voting for themselves.  It wouldn’t take more than a day or two for Jackie James to do a quick inventory - isn’t this information collected of campaign donors?

posted by: THREEFIFTHS on July 10, 2013  11:06pm

The only way IRV will work is We the people must do the voting.Not just the people of Ward Committees.

posted by: HhE on July 11, 2013  12:30am

Eddie, discribing Hieronymous’s gaming of IRV as “rational” is rather off, and thus offensive.  The critique is lodgical, and is grounded in reason.

Please consider this, is your second choice a close second, positive choice that you would not object to?  In my case, Gary.  Then of course I would put him as my #2.  But what if your second choice is a distant second, that you rather not have, but might well win?  Then putting a long shot you rather not have starts to look good.  Hello Zulu Plumber for my number two. 

Clearly, IRV is not perfict, but is probably far better than what we have.  Bravo Ward 25 for having a go, and being a positive example.

posted by: robn on July 11, 2013  1:03am


I follow your logic of B winning over A 55:45 (which is what IRV intends…aggregating preference upward from the poorest perfroming candidate) but I don’t really follow your second logic that B wins over C 70:30. If IRV eliminates the candidate with fewest ballots and redistributes them to the next, C disappears and you wind up with your first logic; B wins over A 55:45.

posted by: robn on July 11, 2013  1:04am


Secret ballots aren’t transparent and they aren’t democratic (big or little D). This vote is illegitimate.

posted by: Bill Saunders on July 11, 2013  2:51am

New Haven Taxpayer,

I agree.  Though ‘The crooked two party system’ is his favorite tagline, 3/5’s has some serious juice when it comes to issues like IRV and Charter Schools.

posted by: THREEFIFTHS on July 11, 2013  11:02am

And by the way there are other systems.Not jusy IRV.

Borda Count Approval Voting Range Voting

Election Selection
Are we using the worst voting procedure?
Erica Klarreich Science News 162,18 (2 Nov 2002) 280.

posted by: Eddie on July 11, 2013  12:43pm


I didn’t mean for my comment to be offensive.  I just mean the suggested strategy is not rational in the sense that Dale is portraying.  Say that your choice is E>H>K, as I think you suggest.  Then you vote E as your first choice, K as your second choice and H as your third choice.  Then say that E is eliminated and your vote causes K to win.  Either you are unhappy, because you really preferred H to K, in which case you acted irrationally.  Or you are happy because K beat H.  If the second case is true then your preference ordering for that particular election really is E>K>H.  That is to say that you truly would prefer the ward to elect K over H.  Perhaps this preference ordering is the product of a broader strategic environment, in which each ward will subsequently vote.  But I don’t think any set of electoral rules at the ward level can correct for that type of strategic voting.  For a particular election IRV seems like it would do a pretty good job of diminishing strategic voting. 

Even Dale’s example requires those with the top C preference to decipher the secondary preferences of a lot of voters.  For example, if only 5.000001% more of the voters that had B as a first preference also preferred C to A, the C voters could win their top choice.  It is probably a difficult calculation to make and acting strategically risks losing your first choice.

posted by: abg22 on July 11, 2013  2:51pm

Some local Democratic Party organizations follow DNC by-laws and some do not (it is encouraged but not required). It is not uncommon for local party organizations to use an Australian ballot exclusively for endorsements, for the same reason that the United States has used it for elections since the late 1800s. But this is all a moot point in the case of an unofficial and nonbinding straw poll, which is what we are talking about here. The fact that the co-chairs had any kind of ward committee consultation is commendable; the fact that they used it as a civics lesson about voting systems goes well beyond the call of duty.

My view has always been that elected officials at any level are not entitled to secret ballot. Obviously you would never have an elected body like the board of aldermen, or party officials who are elected, taking a secret vote on a matter of public policy, or to fill a vacancy for a public official, since knowing how elected officials vote is essential to holding them accountable. But a ward committee is not elected, they do not vote on matters of public policy, and when they conduct non-binding straw polls I don’t object to them having the same privacy as you or I when we go to the polls in November. For whatever it’s worth, I suspect you would have seen an even larger margin for Harp with an open roll-call vote.

posted by: robn on July 11, 2013  3:30pm


Fwiw, all Democrats are supposed to follow the rules of the Democratic National Committee and their rules say that votes cannot be done by secret ballot. This vote is illegitimate.

posted by: HhE on July 11, 2013  4:58pm

Eddie, I am glad to hear you did not mean to be offensive (I am also glad you were able to take my mean in spite of a typo).

Any system of voting, and the particulars of any election, invites people to game the system.  For example, my Uncle voted “against” President Obama:  he wanted him to win, knew that he would (especially in Oregon), but did not want him to win by too large a margin.

I think this a brilliant use of IRV; a great lesson for us all, and a better way of voting—even if the outcome is not what I wanted. 

Gaming the system, or voting “strategically,” is not irrational.  Is not grounded in emotions, nor is it insane.  To be successful, it requires rational, and typically well thought-out, choices. 

I think it is important for all of us of the NHI Comentariat to consider the meaning of the words we use when objecting to someone else’s argument.  Or to use the extreme example I used with my friends at the Newhallville CMT, you are welcome to say, “I take great issue with you.” One just does not need to add, “and your Mother is a whore.” 

(Now excuse me while I go take issue with you on another thread.)

posted by: Hieronymous on July 11, 2013  5:19pm

For what it’s worth, I understood Eddie to be using a term of art, and I take his point.

posted by: Dale Sheldon-Hess on July 11, 2013  6:14pm

Robn: “I don’t really follow your second logic that B wins over C 70:30.”

Sure. What I’m showing here is what would happen if the election were just B and C; what would happen if A weren’t running. So if you just ignore all the places A shows up in the ballot, you see that 70% of them list B first, and 30% list C first.

With 3 candidates (A, B, C) there are 3 possible pairings of candidates (A&B, B&C, A&C). B beats A 55:45, B beats C 70:30, and A beats C 55:45. Since B never loses any of these one-on-one challenges, B is probably a pretty good choice for a winner. Election researchers call this a Condorcet winner (named for the Marquis de Condorcet, a political philosopher during the French Revolution.)

There are a host of election methods which will always pick the Condorcet winner, if one exists (and voters vote honestly.) But IRV is not one of them. As you point out, “B vs. C” never comes up with IRV, because B is eliminated first. And that’s what causes the problem. Candidates, like B, are compromises. They may not be a lot of people’s first choice (they may not be *anyone’s* first choice!) but they represent a consensus or centrist stance. A vs. C is a battle of extremes, which IRV brings to you by first eliminating the middle ground.

(Not to say that “somewhere in the middle” is always automatically the right answer. But when a quality candidate credibly can be the compromise choice, I think that’s a good thing.)

posted by: Dale Sheldon-Hess on July 11, 2013  6:25pm


Two points I’d like to make.

One, the Green party doesn’t use IRV for all of it’s elections. The Green party of Texas is actually using approval voting in its primaries now(which is one of the voting methods I strongly support over IRV.) A few state Libertarian parties are using approval as well, as is the Harvey Milk club in San Francisco. (I also think, for reasons like my example above, that Green party will not ultimate be helped by advocating for IRV. But something like approval WILL help.)

Two, IRV doesn’t guarantee a winner who’s supported by more than half of the voters. The reason why not, is exhausted ballots. They saw this happen in Oakland, CA. Even though the final total was reported as 51:49, that ignores the fact that about 13,000 valid ballots were not accounted for in that tally. The actual result was 45:44, with 11% no longer being counted, because all the candidates listed on those ballots had been eliminated. I mean, you may as well eliminate the 2nd place candidate at that point, and claim it was a unanimous decision. That would be just as meaningful of a claim. And I think Minneapolis, with 5 Democrats (DFL) running for mayor, could see a similar problem.

posted by: Dale Sheldon-Hess on July 11, 2013  6:51pm


I should have read Hieronymous’s example more closely. You’re right, listing a candidate later on your IRV ballot can never cause one of your higher-ranked candidates to lose. And while that sounds pretty good, it’s not as good as what you actually want, which is that listing a candidate higher would never cause you to experience a worse outcome. That’s not the same thing, and that’s also where IRV fails.

Consider my first A, B, C example. The C-first voters aren’t hurt by listing B 2nd; the election never even examines that rank on their ballots. The way they are hurt is by not listing B _first_. Because if they would do that, they’d get their 2nd choice, B, instead of A. That would be better for them.

Yes, knowing when to pick “the lesser evil” is a hard strategic decision to make, and it’s the kind we’re all familiar with from plurality voting. But that’s my entire point: IRV doesn’t eliminate having to make those kinds of decisions, it just puts them off for awhile.

With approval voting (and with score voting as well) it is never to a voter’s advantage to rate their true favorite candidate as anything other than the best. Ever. That’s not a promise IRV can make. Someday, with IRV, you will find yourself again facing the choice of when to pick the lesser evil.

posted by: Eddie on July 11, 2013  9:48pm


I have no bones to pick with approval voting.  Steven Brams persuaded me that it is an excellent electoral rule. 

Still, I don’t believe that strategic voting is just as pervasive with IRV as it is with plurality.  Strategic voting with IRV requires much more information. I really can’t imagine trying to infer voters’ secondary preferences with such precision that strategic voting would ever be a good idea.  Perhaps it is more common than I think.  I would need to see some empirical findings to suggest that this situation actually arises in practice.

posted by: Claudia Herrera on July 12, 2013  6:32am

Question for who would be interested.

Between all the different opinions and obviously many of the commentators and the people who is working on this IRV idea,  seem to have a lot of understanding of the political process for elections.

Are you are counting too that there are lot of new generations of immigrants voters that in their home have no passion “education” of the importance of vote? And when they vote it is most of the time because the candidate is “like-able” or because the parents (who can not vote) give and shared their own frustrations.
There is thousands of young Latinos in New Haven that can vote now and I just wonder what do you think about this?

posted by: cedarhillresident! on July 12, 2013  10:48am

Claudia Herrera

I can not really answer your question but what I can say is city’s in California do use this very method. And California has a high Latino population. And it works there. And Latino leaders there do approve of this method.

posted by: Curious on July 12, 2013  1:08pm

Claudia, what are you asking?

It sounds like you are saying that these immigrant Latinos were not raised to be political and care about voting…and that when they do, it is because they just “like” a candidate for simple reasons, or because of the influence of their parents.

I would say that is no different from most citizens in the USA, regardless of background or length of time in this country.

posted by: HhE on July 12, 2013  2:03pm

Good one on Hieronymous for taking Eddie’s meaning.

Good one on Eddie for his thoughtful analysis.

Shame on me for not understanding how Eddie used “irrational.”

posted by: Gabriel Da silva on July 13, 2013  8:08pm

I would like to commend the amazing job that the co-chairs for Ward 25 have done in this very contested election year!. The results of the endorsement are what they are, many don’t like it but we all have to live with it (50% of the country voted for the OTHER candidate and we all needed to accept the results). The vote was fair and anonymous (as it should), no one should feel pressured to raise their hand, we don’t do it in the presidential race why would we change it for this election? The insinuation that the process is rigged REALLY rubs me the wrong way and I would invite anyone to prove me wrong. Janis and MIke keep the good work

posted by: robn on July 14, 2013  8:12am


One would think with all this talk of democracy and transparency, the chairs would at least be familiar with their own party rules. Secret ballots violate Democratic Party rules. Plain and simple the vote is illigitimate.