Avoiding the Instant Runoff Virus in the Virginia Gubenatorial Primary

I read David Swanson's blog on the VA Gubenatorial primary over at DailyKos. He mentioned that it's essentially a three way race, and asked how to deal with that.

This dilemma could also be solved, in a way, with Instant Runoff Voting (IRV). Back when Deeds was trailing, Rob Richie was arguing that both McAuliffe and Moran supporters would choose Deeds as their second choice and thus give him the victory if IRV were used. But so might supporters of Deeds favor Moran as their second choice. In primaries conducted on paper and counted locally (as in a recent Charlottesville City Council Democratic primary at a single polling place that used IRV) the integrity of an election can be protected while considering second choices and ensuring majority support for a winner. But in a state-wide race, votes could only be counted at a central location if IRV were used. If you can't ensure the results by having them counted publicly where they are cast, what good is improving the method of calculation?

There is a better way that takes into consideration the weaknesses of winner-take-all. We should figure out who the best winner would be and back that candidate. Claims about viability do not in this case even enter into it. The race is a three-way tie.

I did not know that the Charlottesville firehouse primary votes were counted in the single location where cast. But David Swanson hits on a big problem with IRV - you will need central counting for anything other than an IRV race that excompasses only a single precinct polling place. He recognizes that counting votes were cast is crucial to election integrity, and that central counting is problematic.

IRV seems to be a "we can't get people really interested in the political process, so let's throw in the towel". Swanson isn't buying into that strategy. He seems to recognize that the way to elect the best candidate in the primary is to motivate people to get out and vote for the best candidate. It's about what it will take to inspire more people to get out and vote, not about figuring out who is the more viable candidate or running ENRON accounting tricks with IRV.

One other consideration, beyond who's best to win the primary, is who's most likely to win the general election if nominated. But I've heard passionate declarations that only Deeds and only Moran and only McAuliffe can win the general election. It comes down to whether you buy the conventional wisdom that the best way for Democrats to win is to steal Republican votes, or you accept the alternative view that Democrats have a better chance if they inspire more people to vote and allow the Republicans to keep their voters. Given that huge numbers of Virginians registered to vote last year precisely in order to vote for Obama for president, the inspiring-more-people approach has greater potential than usual. Can first-time voters in 2008 be persuaded to vote in a general election in 2009? What about in a primary?

Obama or DNC Chair Tom Kaine haven't done anything to build the Democratic Party so that they can take advantage of the "inspiring-more-people" approach for the 2009 primary election. The "Obama For America" groups have morphed into "Organizing For America" that is part of the DNC. So far they appear to only be trying to organize people to support specific Obama policies and not party buiding. That is a mistake, and I feel that will come back to bite them in the ass in the 2009 and 2010 elections.

But even Swanson recognizes that IRV is not the answer to what ails lower voter interest in the political process. It will take plenty of good old-fashioned hard work to get people out there to vote.

Think Hoboken is "hot" now? The voters will be steaming if they catch the Instant Runoff Virus!

Back in the mid 80's, I used to work in NYC as a photographer's assistant. My college room-mate Victor Ongkingco introduced me to all the charming places to go eating and drinking in NJ - and we tended to end up in Hoboken at the end of the night. Parking was a pain, but I always enjoyed visiting the place and I developed quite a fondness for one bar (which is no longer there) where I had my first wheat beer.

Hoboken showed up in the news the other day, when an article in the Hudson Reporter claimed that the Hoboken City Council was going to be considering IRV. I thought they must be joking of course. IRV is hardly a "new idea" - maybe just new for Hoboken or NJ in general?

And unlike the merde that FairyTaleVote is pushing in NC - that IRV gives better pluralities - they are still pushing the story that you can keep counting IRV ballots until someone gets a majority. But at least the reporter admits that it could be confusing.

Turns out they weren't really voting to institute IRV in their town. I went to the Hoboken City Council website, and found that there was merely one of many resolutions on the agenda asking the NJ legislature to pass a bill creating an IRV study commission. The Hoboken resolution merely asked the legislature to pass the bill and expedite their study so that Hoboken could possibly rush to do IRV for their November 2009 general elections. Of course the rationale for using IRV was a slightly different variation of the usual Fairytale vote propaganda. Here is the resolution with my comments under each relevant section:

A RESOLUTION TO CUT AN ESTIMATED $75,000 IN TAXPAYER EXPENSE WHILE INCREASING VOTER PARTICIPATION THROUGH USE OF INSTANT RUNOFF VOTING (IRV)

WHEREAS, Hoboken runoff elections cost the taxpayers of the City an estimated $75,000 to $100,000 every two (2) years; and

TELESCA COMMENTS: IRV advocates claim that IRV saves money if you buy the simple assumption that one election is cheaper than two. That is not true if you factor in all the costs of the more complicated IRV method: voter and candidate education, pollworker and election administrator training, documentation, and voting system upgrade or replacement. And there are no voting systems that are federally certified to handle IRV tabulations.

You can see the high cost of IRV both from governmental studies done in jurisdictions considering IRV, and from the jurisdictions already doing it.

The Maryland State Legislature considered doing IRV three times - in 2001, 2006 and 2008 - and did fiscal studies in 2006 and 2008.
http://mlis.state.md.us/2008rs/fnotes/bil_0002/hb1502.pdf
http://mlis.state.md.us/2006rs/fnotes/bil_0002/sb0292.pdf

Their costs for voter education alone were estimated to be $0.48 per registered voter - the cost of a 1st class stamp. Think that's enough? San Francisco has spent $1.87 per registered voter per year in the IRV elections they have done since 2004 - and a recent San Francisco civil grand jury report indicates that might not be enough, because voters still don't know enough about IRV after 4 IRV elections.
http://www.sfgov.org/site/uploadedfiles/courts/divisions/Civil_Grand_Jury/year-of-five-elections-for-sf.pdf

That's just the cost of voter education. The same MD fiscal studies estimated that it would cost an additional $3.50 per registered voter to implement IRV in 2006 when they were using paperless DRE touchscreen voting machines. In 2008, they estimated it would cost an additional $3.08 per registered voter if they switched over to using op-scan paper ballots. But the 2008 study didn't include the cost of federally certified IRV voting machines and software that didn’t exist then and still does not exist!

Actual costs of implementation in places that have used it is even scarier! Pierce County, WA used IRV in 2008. It cost them $2 million to implement an uncertified system for 375,589 votes - or $5.33 per registered voter! That is on top of the regular costs of their election system. And in two of the three races that used IRV to decided the "winner", the "winner" didn't get a majority of the first column votes cast! Now 2 out of 3 voters in Pierce County want to ditch IRV after their first election!
www.aocweb.org/aoc/Portals/0/Content_Managers/feb0909govcommrev.pdf

WHEREAS, multiple elections annually foster the disenfranchisement of voters, reducing voter turnout and public confidence in the process; and

TELESCA COMMENTS: I don’t know of any formal studies that show traditional elections and runoffs reduce voter turnout and public confidence in elections.

As a precinct chair and an officer in my county’s Get Out The Vote program, I do know that voter turnout in runoff elections can be lower than for the initial election. But that could be due to many factors including lower voter interest, weather, burnout, or just not liking any of the other candidates.

San Francisco first used IRV in 2004. They have used IRV in every subsequent election since then. From 2004 to 2007, voter turnout has dropped along with the number of registered voters, so IRV does not increase voter turnout. And the percentage of people who showed up at the polls who didn’t know they were supposed to rank their choices increased from roughly 33% in 2004 to almost 50% in 2005.

One problem with IRV is that it very rarely ensures an authentic majority winner in a single election. In the majority of elections where there is no winner in the first column and IRV is used to tabulate votes in subsequent columns, the winner rarely wins by a majority of number of 1st column ballots. A winner is manufactured using IRV vote tabulation methods that seem more like ENRON accounting methods.

There is a movement across the country not to trust the results of “black box” elections. IRV is such a complex tabulation method that few people understand it – including election administrators. If trained election administrators don’t understand it, what chance does the average voter have of understanding and trusting it? Many verified voting and election integrity advocates feel IRV is a step backwards, not forwards.

WHEREAS, the Council is committed to the democratic process and wishes to encourage voter participation while simultaneously reducing the cost to the taxpayers; and

TELESCA COMMENTS: There is always going to be a trade-off between voter participation and election costs. You could cut costs by having one place in a municipality to cast your vote in person in order to reduce costs, but you would end up disenfranchising voters who live further from the location. Perhaps you really can't do both. Maybe try finding a savings someplace else?

WHEREAS, Instant Runoff Voting has proven to be successful nationwide, in such diverse places as Aspen, Colorado, San Francisco, California, and the States of Louisiana and South Carolina;

TELESCA COMMENTS: This clause is somewhat misleading. IRV has been used in Aspen and San Francisco, but it could hardly be called successful. They voted to use IRV in elections without knowing how they would implement it. As such, they were forced to use the method under threat of lawsuits from IRV advocacy organizations like FairVote.

Costs have gone up, and election transparency has done down. Aspen had to hire an outside consulting company to run their complicated IRV elections – which cost more than holding a regular runoff election. And there are almost daily reports coming from Aspen about election irregularities – and this is from an election that took place in early May. It was not certified by the Aspen Election Commission because they wouldn’t sign off on a method they didn’t understand.

South Carolina and Louisiana passed laws to allow for the use of IRV for overseas absentee by mail voters, but they do not use IRV for any other elections.

North Carolina passed a law creating an IRV pilot program for 2007-2008, and extending it from 2009 to 2011 (inclusive). Even after a full-court press by the State Board of Elections and many IRV advocacy groups like FairVote, they could only get two communities to use IRV in 2007 – Hendersonville and Cary.

Even before Cary voted in May 2007 to pilot IRV, the State Board knew it was too risky to use in 2008 elections because state law and federal regulations require using only certified voting systems to tabulate IRV.

Only one NC community – Cary - needed IRV to tabulate votes beyond the first column. Due to my work in verified voting, I was appointed an official observer to the IRV pilot by the Chair of the Wake County Democratic Party. IRV did not do well in Cary.

The 2007 Cary IRV pilot program was largely managed by IRV advocacy groups, with no advance guidelines. Some voter education volunteers admit deviating from Election Board instructions to create a more positive outcome on the exit poll surveys — also conducted by IRV advocates.

The Wake Board of Elections couldn’t follow simple IRV hand tabulation procedures. Ballots were mis-sorted, simple calculator mistakes were made and a non-public recount turned up missing votes. The winner did not receive the 50 percent plus one vote majority advocates claimed IRV would ensure in a single election. He got 1401 out of 3022 first-column votes.

There has been no analysis of the 2007 pilot. The proffered reason given for extending the pilot beyond 2008 was cost savings, even though fiscal studies done by other jurisdictions show IRV elections cost more than traditional election methods.

The original IRV pilot extension bill had the same flaws as the first pilot program. Election integrity groups requested an improvement which required “… the pilot program shall be conducted according to … standards consistent with general election law …” Unfortunately, this legislative requirement has not been met.

After passage of the pilot, election integrity advocates (including myself) pointed out how IRV conflicts with general election law not written with IRV in mind, and recommended ways to make IRV comply with general election law. The State Board ignored those recommendations and approved IRV guidelines that conflict with general election laws.

North Carolina and other states have laws requiring that votes be counted where cast until the count is completed to prevent ballot tampering. But State Board IRV guidelines call for partial ballot counting at polling places, then moving the ballots to a central location for further counting. The federal Help America Vote Act requires voters be notified of over-votes before a ballot is cast. Our voting system can’t notify voters of second and third column over-votes on IRV ballots.

From early 2007 through January 2009, State Board members and staff claimed we needed federally certified software to automate IRV tabulations. The State Board recently developed automated procedures they now claim need no federal certification. Those procedures were developed with no input from election equipment vendor ES&S. We still do not know if the new IRV procedures violate any contracts, warranties or other agreements with ES&S? Will NC voters be required to foot the bill in the event of election problems?

Very few NC communities considered taking part in the 2009 IRV pilot. Cary – the only NC municipality that used IRV to tabulate an election winner in 2007 – voted not to participate in the 2009 IRV pilot. The consensus of the Town Council was that IRV didn’t work as advertised in 2007, and they didn’t want to be an election lab rat again. Don Frantz – the most vocal opponent for IRV on the Cary Town Council – was the elected with the method. He didn’t like it in 2007 and he doesn’t like it now. Councilperson Julie Robison – who voted to participate in the IRV pilot in 2007 – doesn’t support the IRV election method because she doesn’t trust the tabulation procedures. On April 30, 2009 – Cary voted to stick with traditional majority non-partisan majority elections with runoffs if needed because they are more transparent than IRV.

NOW, THEREFORE, BE IT RESOLVED, by the Council of the City of Hoboken, County of Hudson, State of New Jersey, that it fully supports Senate Joint Resolution No. 43, sponsored by Senator Bill Baroni of District 14, creating a commission to study instant runoff voting and the implications of IRV within the State of New Jersey and to encourage the commission to act promptly so that the City can introduce a referendum for voter consideration establishing IRV during the next general election on November 3, 2009.

TELESCA COMMENTS: It appears according to SJR 41 that they will actually study IRV - something that was not done in North Carolina before the pilot passed in 2006. If they don't rush the study, they will find out more information about IRV than they ever wanted to know – including all the extra costs and perhaps even the many ways that IRV conflicts with existing elections laws in NJ. It will take a while – possibly years – to resolve just the conflicts in their election laws if they decide to use it.

Based on the experience of other jurisdictions that are using IRV, NJ should not rush into using the method until they can take the time to weigh all the evidence. Or heaven forbid, be required to use it and then realize just how problematic it will be.

From the way the resolution is written, Hoboken wants to rush the actions of the state IRV Study Commission so they can have a special referendum on IRV sometime this summer enabling them to use IRV during their November 2009 general election? That is not a good idea.

Does it make sense to go to the trouble and expense of holding a low-turnout special referendum on IRV just to use IRV in November and supposedly save money not having to hold a runoff election? That’s like robbing Peter to pay Paul. Furthermore, the turnout in special elections for IRV tend to have even less turnout – and therefore are less democratic – than even the runoff elections they are using to replace. That is what happened in Aspen, CO.

BE IT FURTHER RESOLVED, that a certified copy of this resolution be transmitted by the City Clerk to Senator Baroni and our 33rd Legislative District Representatives.
Meeting Date: June 3, 2009

A friend of mine attended this meeting and told me that - unfortunately - the resolution passed: 6 voting yes, 2 voting no and 1 abstention. There was no discussion of how IRV might effect election integrity. But that is to be expected. Cost cutting at all levels of government tends to be the biggest concern on elected leader's minds these days.

But I am really sure that this IRV study bill will go anywhere. Doesn't their legislature have a research staff that can study this issue for them and make a report? Or is this gonna turn into a Rob Richie "dog and pony" show (or is it "chili cook offs" and "ice cream socials") where FairyTaleVote will control the agenda for the meeting?

Rest assured that election integrity and verified voting activists will be paying attention to what goes on in the Garden State. My aunt lives in NE Philly not too far from Trenton, and I got plenty of places to stay near Hoboken - from a futon in Tribeca to a very nice couch in Belleville.

NJ residents are practical and pragmatic. If they can joke about being able to see the air they breathe, they will want to see an actual majority they are being promised. There is a good chance they won't buy into the hype once they find out how bad the Instant Runoff Virus really is!

Minneapolis gets ready to return to primary elections in case they lose their IRV suit!

Minneapolis is one of those places that IRV advocates point to as a place where IRV was chosen by voters in landslide referendum wins.

But it's not popular with everyone. A local election group took the city to court challenging the constitutionality of IRV. IRV advocates are confident they will win in court. But not everyone is so confident. So Minneapolis is getting ready to let them do runoff elections in case they lose in court.

Return of primary requested, just in case

The City Council unanimously approved a measure that puts the city on track to make a September primary election possible, just in case Minneapolis needs one.

Currently, there’s expected to be no primary. That’s because this year’s municipal election is set to use ranked-choice, or instant-runoff, voting.

RCV lets voters rank their top three candidates in each race. In single-seat elections, any candidate wins by getting 50 percent of the votes plus one right off the bat; if no one reaches that threshold, second- and third-choice votes could get weighed. The process eliminates the need for
a primary.

While voters approved RCV for use in this year’s election back in 2006, a lawsuit has thrown a potential wrench in the city’s plans. The Minnesota Voters Alliance, a citizens’ group, is questioning RCV’s constitutionality, arguing the system doesn’t equate to one person, one vote.

The case has traveled to the state Supreme Court, where arguments will be heard May 13. A ruling is expected in early June, according to city documents.

It’s important for the city to get as quick an answer as possible — if the court ruled against RCV, the city would need to bring back a primary, something that’s easier said than done. Currently there is no language in the city’s charter directing how to hold a primary that Minneapolis voters are used to; that was eliminated along with the 2006 approval of RCV.

That’s why the City Council is requesting to re-amend the charter. In other words, were the Supreme Court to deliver an unfavorable ruling, Minneapolis would be prepared.

On June 9, the city’s Intergovernmental Relations Committee also will weigh whether it should be able to reinstate a primary for another reason: if RCV turns out to be just plain too difficult.

Minneapolis is set to make history by being the first municipality in the world to hand-count a ranked-choice multiple-seat election, elections Director Cynthia Reichert has said. That process was tested May 6-7 by elections staff. An official report has yet to be given on the experience, but Council Member Paul Ostrow (4th Ward) has some concerns.

“What I’ve heard is that hand-counting of the at-large seats is extraordinarily challenging,” he said.

But other council members already have said they don’t feel comfortable making that a reason to altogether abandon the new system. Council Member Cam Gordon (2nd Ward) said he would only support a primary if RCV were found to be unconstitutional.

Extraordinarily challenging? You ain't seen nothing yet - they say it could take weeks to count!

We figured out here in NC that if we had one statewide race that went to IRV in our 2008 May primary, it would have taken 7 weeks to count. And that counting could only start AFTER the State Board of Elections certified who came in second out of 4 candidates - since NC did top-two IRV in the 2007 election pilot.

In other words, they'd still be hand tabulating IRV ballots well after the late June primary runoff election already gave the results the night of the election. Some savings of time there!