Saturday, May 23, 2009

Votes aren't the only things being pulled out of someone's "Asspen"!

Rob Richie wrote a "huff piece" for the Huff Post about IRV. As usual, he put a positive spin on the Aspen IRV election (which I don't think has even been certified yet). And he got stuff wrong.

In a Nutshell
Instant runoff voting is a ranked choice voting system that allows voters to rank candidates in order of preference. Recommended by Robert's Rules of Order for postal elections and used in a rapidly growing number of elections here and abroad, it represents a major improvement over the usual plurality-based and two-round systems of voting. It protects majority rule, eliminates the need for costly extra elections and all but eradicates the potential chaos of "spoiler" candidacies. But beyond its clearly established benefits, we are seeing anecdotal evidence that suggests that IRV has a positive effect on the influence of big money on elections, and mitigating the temptation for campaigns to "go negative."

It used to be that FairVote claimed that Robert’s Rules of Order recommended IRV over all other election methods – now it’s just for postal elections. But I am not sure that IRV is being used in a rapidly growing number of elections here and abroad. Those governments that already have a parliament use IRV/RCV or STV. I don't think that any new overseas governments are clammoring to use IRV. In fact, 61% of British Columbia voters recently gave STV (a version of IRV) a crushing defeat - in the second defeat for ranked choice voting in BC.

I am sure FairVote is spending a lot of money trying to push IRV in communities all around the country and the world. But here in NC, IRV is failing to catch on.

In 2007, only 2 out of over 500 municipalities chose to take part in an IRV pilot program. Only one election went to IRV, and that was a disaster! There were no IRV elections in NC in 2008 (probably because our own State Board of Elections felt IRV was too risky to use in the 2008 federal elections with the expected heavy turnout). A bill to allow the Wilmington City Council to have the option to have IRV elections was pulled at the request of the City Council when the language of the bill would have REQUIRED the use of IRV. And even though an IRV pilot program extension bill was passed in 2008, that bill required guidelines that our State Board of Elections could not meet (IRV conflicts with general election laws) so that only one community opted to participate in the 2009 pilot.

That one single community out of over 500 in North Carolina was Hendersonville. They participated in the 2007 pilot only in the front end of the election (the ranking of candidates). As the election in Hendersonville went the same way as two IRV elections in Takoma Park (they had a first round majority and didn’t need IRV to determine the winner by tabulating subsequent rounds), Hendersonville really had no rational basis for assuming that IRV elections would go smoothly. In fact, since they use a largely untested and certainly uncertified workaround for tabulation of the DRE votes on Excel spreadsheets, the IRV method using DRE machines is an untested one. One City Council member claimed that there was a paper trail for the IRV vote, but there has never to the best of my knowledge been a testing of doing a full-scale recount of IRV votes using the thermal paper trail created by DRE machines.

In all well-intentioned attempts to reform our electoral system, the primary goal is fairness: finding mechanisms allowing all eligible voters to have a better chance to participate and be represented. When those criteria are satisfied, we think that government becomes more accountable and more honestly reflects the will of the voters. But sometimes we can be pleasantly surprised when a change designed to improve the political system in a broad sense also turns out to have other desirable effects beyond the initial intentions.
I am not really sure that all attempts to reform our electoral system are well intentioned and have fairness in mind. I am certainly not sure that those are the intentions of FairVote. And I certainly don’t think that IRV elections satisfy those criteria.

I do not believe that IRV is a fair electoral system because it is too complex not only in the front-end voting part of the election, but especially in the back-end of the counting. Candidates in Cary, NC admit to being confused on how to deal with campaigning in IRV elections. Voters in Cary and in Hendersonville were also complaining about the cofusion.

You must have a simple and easy to understand method for counting votes and explaining how the winners will be determined. IRV in almost any form is “black box” voting – hard to explain, hard to understand for educated people and “just trust us – we’re well-intentioned reformers” for everyone else.

This is exactly what we're seeing with the growing implementation of instant runoff voting (IRV) in municipalities across the country. With IRV, voters have one vote, but are allowed to indicate their backup choices in the event that their favorite candidate lacks enough support to win. After voters rank candidates on a ranked choice ballot, the first-choice rankings are tabulated. If no candidate wins a majority (50% plus one), a series of "instant runoffs" take place. The weakest candidates are eliminated and ballots for that candidate are added to the totals of the remaining candidates until one candidate earns more than half the votes. The winner is the majority, consensus choice. (Minnesota Public Radio recently did a charming video demonstration of IRV in action using Post-It notes, which you can watch here.)
One candidate earns more than half of what votes? The total number of votes cast in the first column of each race, or more than half the votes of the last two candidates standing?

A successful IRV election was held in Aspen, Colorado last week (the city's first IRV election), in which incumbent mayor Mick Ireland defeated three challengers in a contest with a record-breaking turnout; 45% versus the usual 37-38%. Analysis of the election by TrueBallot showed that every single vote cast for mayor was valid, meaning 100% of those who opted to vote for mayor had their vote count. There were more voter errors in the novel use of IRV to elect two at-large city council seats, but still less than 1% of those at the polls.
Who says it was successful? The election hasn't even been certified yet! Comments by many people suggest that the reason for the turnout was that people wanted a change in administration. Ireland was pushing IRV because of the advantage it gives incumbents. Ireland and other pro-IRV candidates got themselves on the IRV commission to figure out how to run IRV in Aspen. Were other candidates on this commission? If not - why not?

Also notable were the fundraising figures. Challenger Marilyn Marks outspent Ireland, breaking Aspen records with almost $40,000 in funds. Ireland mustered less than half of Marks' total, with less than $18,000 raised. Despite this disparity in resources, Ireland emerged victorious. The biggest spender in the city council race also was defeated in an election in which the two incumbents were upended.
Where did Rob Richie get these figures from – did he pull them from his "Asspen"? According to the public spending reports, Ireland had spent $14,513 up to 4/29/09 vs. $10,149 by Marks during the same period. The final reports won’t be out until June, so where did Rob Richie get his “almost $40,000 in funds” figure from? Did he make that up?

We knew IRV helped level the campaign finance playing field when avoiding costly runoffs, as would have happened previously in Aspen. We didn't anticipate an impact within single elections, but here's why there might be a connection. In a typical campaign, campaign money is often spent attacking one's major opponent through ads. That tactic assumes the "zero sum" logic of a two-person race in which every vote lost by an opponent helps you by default. But with IRV, voters are more likely to have more than two choices. Candidates have a greater motivation to make an affirmative case to earn support because negative attacks may hurt another candidate without helping you.
There "might be a connection"? This is quite a stretch! How much would a runoff election cost in Aspen's 4 precincts (5 including ABM and in-person Early Voting)? Would it have cost the $7,500 paid to True Ballot to run the election - including the fix of that pesky "inverse" problem that declared the lowest vote-getter in a pre-election test to be the "winner"?

That is not what folks in Aspen say about the campaigns. Most everyone agrees that the lack of the run-off and having 9 candidates in the field allowed the candidates to run “motherhood and apple-pie campaigns.” There was no real substance to the answers, and positions. They were able to run popularity contests without having to take positions. There was too much noise to pick out a real message. That is one of the big problems with IRV!

One of the reasons why Robert’s Rules of Order favor traditional election and runoffs over preferential voting is the real lack of choices. In Aspen and elsewhere, the month of run off among only 6 candidates (4 council and 2 mayor) would have allowed for real issues to be debated and discussed. From my own experience in the 2008 primary campaigns in NC, with so many candidates running, no one got a real chance to tell other voters what the real differences were between the candidates. Both of the runoff candidates for Labor Commissioner – John Brooks and Mary Fant Donnan (disclosure – I know John Brooks and supported him in the runoff) – felt that the runoff gave them a chance to explain the differences between each other but also between them and the current Republican Labor Commissioner Cherie Berry. Had we had IRV in that primary election, I doubt few voters would have been able to make responsible and informed choices in the Labor Commissioner race that had a 33% falloff from the Presidential and other top-ticket races.

I am convinced that in the future, if there are more IRV elections, candidates will hire math consultants to learn how to game the system, and run campaigns for a month that focus on popularity (lots of bbq’s and pizza parties) and not on ideas and substance. Is that really how we want to elect our officials?

Because voters get the option to rank their preferences, candidates also have a new incentive to make their case to backers of other candidates. Negative attacks perceived to be unfair are particularly counter-productive if the candidate on the receiving end loses early in the counting and that candidate's backers punish the attacker by ranking other candidates higher on their ballot. Attacks will still be leveled at opponents in IRV elections, particularly when there is a clear frontrunner as was the case in Aspen, but overall IRV encourages more positive, substantive campaigns in which candidates try to earn first-choice support from opponents while remaining attractive to other candidates' supporters. The Aspen Times weighed in after the election, writing, "[We] have been impressed with the professionalism displayed...[C]andidates have treated each other respectfully during these stressful times."
That is not what some observers who chose to remain anonymous have claimed about the Aspen elections. Some supporters of Marks were threatened by people in the Ireland campaign.

Let's take a look at another example. Earlier this year in Burlington, Vermont, the Progressive Party's Bob Kiss was re-elected as mayor, vaulting from second place after the first count in an IRV election to win with 51% against Republican state legislator Kurt Wright. Just as in his initial upset win in 2006, when he was outspent by approximately four to one by a Democratic state senator, Kiss was heavily outspent by his three main opponents. All three wielded larger war chests, including Wright who spent twice as much as Kiss.
So what? Progressive candidates have won election for Mayor since the 70s or 80s WITHOUT IRV. IRV was not necessary for Progressives to win elections in Burlington, VT.

But once again money seemed to mean less when negative attacks aren't useful. Burlington's candidates participated in forums across the city, and, in part due to IRV, spent little time debasing each other. The positive, substantive tenor of the campaign even won IRV some new converts of past skeptics such as Democratic city councilor Bill Keogh who told the Burlington Free Press, "This campaign has been very, very good," and that the four leading candidates had been "as forthright as they can be with their views. This is the most respectful and informative campaign in Burlington in a long time."
Other observers have found that IRV tends to drive the negative campaigning underground.

Obviously, in the Aspen and Burlington cases the mayoral victors had the benefit of incumbency despite their deficits in cash. But they also showcase a trend that is emerging in IRV races across the country. Similar results have been seen in San Francisco, which has used IRV for city elections every November since 2004. Numerous highly competitive races have gone to candidates who were outspent, including several neighborhood-based candidates targeted by downtown business in the 2008 elections. The editor of the San Francisco Bay Guardian in 2008 wrote that in the highly contested open seat race for District 9 on the Board of Supervisors:
Not really sure these examples are accurate – since Richie got his numbers wrong in the Ireland v. Marks race.

"[The winner] will probably be the one who gets the most second-place [rankings]. So it's in everyone's interest not to go negative. If Sanchez, say, started to attack Quezada, the Quezada backers would get mad and leave Sanchez off their ballots -- and that would hurt Sanchez when the second-place votes are counted. So everyone has been pretty well behaved in [District 9]. I've heard a few whispers here and there, and a few people have tossed off a few nasty comments, but overall the candidates and their supporters recognize that it's better to stay positive."
So that results in bland campaigns that have little focus on issues and the differences between candidates. That is not very democratic!

Indeed, the winner was in fact one of those candidates that embraced the idea of forging alliances over burning bridges. One result of this is that with every member now elected through IRV, the city's Board is far more diverse and community-based than ever in its history. Its 11 members include three Asians, two Latinos, one African-American and one Iranian-American.
Does this have more to do with who runs for office in those communities vs. who votes for them? If IRV is supposed to bring diversity – why doesn’t Takoma Park MD (home of FairVote) have a diverse elected government?

This is not a definitive, scientific case study proving beyond a doubt that IRV will always negate the advantages of money or unfailingly produce smiley-faced campaigns. But what is certain is that because candidates must appeal beyond their die-hard supporters in order to rank highly on as many ballots as possible, the efficacy of negativity becomes at best highly questionable, while reasoned, substantive debate and coalition-building become far more attractive. And when discussion is valued over destruction, the relentless raising and spending of campaign funds can be less decisive.
It it not very scientific or accurate. But if you have read as much of this pro-IRV drivel as I have, you would not wonder why I call this organization “FairyTaleVote”.

In our view, IRV is already a significant improvement simply on its technical merits alone. But if it can also produce such positive byproducts -- even only occasionally -- it only serves to make a good idea even better.
Many people would not agree that IRV is a significant improvement based on technical merits alone. Many people feel that IRV threatens election integrity and verified voting. And now what do we make out of Rob Richie’s claims of “gee whiz” Pollyanna election wonderfulness when he pulls numbers out of his "Asspen" to cast the incumbent mayor as a little guy who got elected over big-spending people who, by default, must be evil?

No comments: