Sunday, June 20, 2010

I knew it was coming - calls for IRV in statewide NC races!

As soon as I knew that there would be a runoff in the Democratic Primary for US Senate, I knew that various people would be calling for Instant Runoff Voting. Here is one of those claims - from Damon Circosta of the NC Center for Voter Education:

High Cost, Low Turnout Likely for Runoff Elections

By Damon Circosta
Published: June 14, 2010

RALEIGH - Most people don’t equate summer with election season. When one conjures up visions of voting the images are typically of a crisp autumn day. Or perhaps for primary voters Election Day might involve sprouting trees and the blooms of a North Carolina spring.
Summer, for both voters and politicians, is usually a quiet time. The public’s attention is on other things like vacation plans and kids camps. Candidates are usually out of the spotlight and quietly amassing resources for the fall campaign.

But every so often, election season extends into the dog days of summer.

This year, for many voters across the state, there is an opportunity to engage in democracy. But with so much else on the minds of the electorate, most of us won’t be braving the heat to head to the polls.
Really - summer is not a season for elections? Who says so? We have always known that runoffs occur after an election. Knowing that elections require a runoff vote - why should this be a surprise for an informed electorate? Or for someone who works for the Center For Voter Education?

Shouldn't the "Center For Voter Education" be among the chief drum-beaters trying to get people out to vote, instead of lamenting why people aren't getting out to vote - thus creating a self-fulfilling prophecy?

North Carolina law provides for a runoff to be held if no candidate achieves more than 40 percent of the vote in a primary election. In the Democratic primary for U.S. Senate, as well as in Republican primary contests for congressional districts 8, 12 and 13, no clear winner prevailed. These races are headed for a June 22 runoff election between the top two candidates.
When the threshold used to be higher, we had more runoffs, and higher turnout for those runoffs. When we lowered the threshold, we didn't need as many runoffs, and we ended up having lower turnout for the runoffs we had.

Perhaps the remedy for low-turnout runoff elections is not to lower the threshold but to raise it?

Turnout projections are exceedingly low for these runoffs. In a state with about 6 million registered voters, fewer than 100,000 will likely show up to the polls. Nevertheless, the expense of holding a statewide election remains relatively constant. It doesn’t matter if two people or 2,000 people show up at a precinct. It must be opened and staffed all day.
While it is correct that there are over 6 million registered voters in NC, there are not that many registered Democrats who could vote in the runoff. According to the NC State Board of Elections as of 11:40PM on June 20, 2010, there are only 2,750,763 registered Democrats who could vote in the Statewide Primary for US Senate in the Democratic Primary.

Unaffiliated voters could vote in either the Democratic or Republican ballot in the May primary, they would have to vote the same ballot in the runoff election. I am not sure how many of the state's 1,410,324 UNA voters voted the Democratic ballot in May and thus would be eligible to vote in the June runoff.
In the Democratic Senate primary, some people expressed concern when Cal Cunningham, the second-place finisher, called for a runoff. Citing concerns about the $5 million expense of holding a statewide election and doubts about his ability to overcome frontrunner Elaine Marshall, these critics said it was an irresponsible move. While reasonable people may disagree about his prospects, it sets a dangerous precedent when we ask candidates to bow out of elections to spare the state the expense.
While I agree that some people expressed concern that Cal ran, it was his right under the law to call for a runoff because he was the second-place finisher and the first-place finisher didn't cross the threshold.

And I agree that it sets a dangerous precedent when we ask candidates to bow out of a runoff election to spare the state the expense. But I don't agree that we should endanger election integrity and public confidence in elections in our state to experiment further with Instant Runoff Voting.
Administering elections can be a costly enterprise. Accessible polls and accurately counted votes require resources. While everyone likes to see our government operate as cost-effectively as possible, scrimping on the very mechanism we use to hold our government accountable doesn’t make sense.

There are ways to achieve more citizen input in a less costly manner than holding a second primary election. Some municipalities in North Carolina and other states have experimented with something called instant runoff voting.
Yes - some municipalities have experimented with IRV. Some do not like it. Cary and Hendersonville tried it in 2007. Cary had the state's only election where IRV was used to count voter's second and third choices when no one won the election on the first choice alone. The Wake County Board of Elections couldn't follow the complicated hand sort/stack and counting procedures, and made some calculator errors that necessitated a secret count the next day with no outside observers or candidates knew about or attended. That secret count found some votes that had been missed the previous day. And out of the original 3022 first column votes, the winner of that race got 1401 votes - 111 votes short of the 1512 votes he would have needed to win in the first column of votes.

Hendersonville tried it, but had winners using just the first column votes. Hendersonville tried IRV again in 2009, and also had winners using just the first column votes. They never needed to count the additional voter choices, and there is no evidence that the Henderson County BOE could have accomplished that task using their DRE touchscreen voting machines.

Quite simply, the voting equipment that North Carolina uses is not certified to tabulate IRV ballots. That's why all the IRV experiments have either used complicated and confusing hand-counting methods like in Cary, or hybrid and jury-rigged counting methods proposed for DRE machines that involve somehow porting voting data over to Excel Spreadsheets, where the tabulation will be done all by machine with little to no possibility for outside observers to verify the tabulations.
The idea is that during the first primary election, voters are offered the opportunity to select whom they would vote for if there were to be a runoff. It’s not perfect and would require spending some money to make sure that the instant runoff system was accurate and secure. But such a system could save money in the long run and also make voting more convenient, hopefully increasing turnout.
Studies done by legislatures that take their responsibilities seriously (as our NC legislator fail to do when it comes to IRV) and real world experiences of places like Pierce County, WA and even Minneapolis MN have shown that IRV does not save money - it actually costs more money.

The MD legislature performed fiscal studies on IRV in 2006 and 2008, and costs per registered voter in MD ranged from an additional $3.08 to $3.50 per registered voter to implement IRV, and an extra $0.48 per registered voter for voter education each and every year there was an election. Applying those very reasonable costs to our state's 6 million voters - it would cost between $18 to $20 MILLION to implement IRV right up front and $3 million each and every year for voter education. Using those very reasonable costs, we'd never break even with IRV even if we needed a statewide runoff every two years!

Pierce County WA found their costs DOUBLED using IRV. IRV cost Minneapolis voters more: a primary and general election in 2005 was $1.12 million (adjusted 2009 dollars) vs. $1.46 million for one single IRV election. Furthermore, Minneapolis found that turnout for their first IRV election was the lowest since 1902 - in over 100 years! IRV has been used in San Francisco since 2004, and costs have gone up while turnout has gone down!

And another problem with continued calls for using IRV in NC is that our own State Board of Elections stated in 2007 (before the first communities decided to use IRV) was that IRV was too risky to use for statewide primary elections (like in 2008 and 2010) because it would violate state and federal election laws. There simply were no certified voting systems (machines and software) that was federally certified to do IRV elections in 2008 - nor in 2010.

Under the system used for certifying voting systems, the voting system companies have to get the whole system tested - not just the machines and software, but even the documentation and the manual procedures. Companies have to submit the whole system for federal certification which they have to pay for. And since there are many different IRV vote counting methods and each is much more complicated than single-column elections, few (if any) companies want to foot the bill. So should we lower our standards for claims of savings and increased turnout that haven't materialized in the experiments done so far?

That's the experimental side - the IRV pilot program in NC. It was originally supposed to run from 2007 through 2009 (inclusive) where only two communities used it in 2007 - but 4 communities voted "no" on IRV: Asheville, Atlantic Beach, Raleigh and Rocky Mount.

So even though no one could use IRV in 2008 because it was deemed "too risky", some of the same advocacy groups pushing IRV now came out right after the June 2008 Democratic Labor Commissioner Runoff to call for extending the IRV pilot - citing mainly the need to save money. They got the pilot program extended until 2011, but only one community - Hendersonville where they never really put IRV to the full test - used it in 2009.

Where IRV was mandated as an election method that must be used, it has also fallen short on promises. IRV was dumped in Burlington, VT by a larger majority and a larger turnout of voters than in the referendum that voted to use IRV. 67% of Pierce County WA voters voted to dump IRV after only one try. Aspen CO voters gave IRV a no--confidence vote after only one try, and now the Aspen DA is investigating whether or not the IRV election violated state election laws.
Short of implementing instant runoff voting, there are other changes we could make, such as rethinking the requirement that a candidate must get 40 percent of the vote to avoid a runoff. Our election system is not set in stone. Using the democratic process, we are free to alter the system to make it more effective.
Not really sure that lowering the threshold is the right way to go, since turnout has only gone down in runoffs since the threshold has been lowered. Perhaps set a sliding threshold based on where the top vote getter placed compared with where the second-place finisher did - and factoring in how many other candidates there were? And not sure we want to say that it's more effective to us a confusing and complex vote counting method that we claim saves money but really doesn't?
The fraction of registered voters who will carve out some time on June 22 to vote, or who cast a ballot using the early voting system, hold a considerable amount of sway in this election. It’s time to consider ways of changing the election process so more of us will get involved.
Yes it is true that we need to get more people involved. So let's raise the thereshold for winning a primary election, so that more elections go to runoff and we get more bang for our runoff buck!

And by all means, let's not have our local Boards of Elections do things to discourage people from voting. They should be encouraging people to wait in line to be the first to vote like people wait in line to buy concert tickets - instead of trying to run people off!

But we in North Carolina are fortunate in many ways that our election administration systems are better than in many other states. After passage of the Public Confidence in Elections Act in 2005, NC was ranked #1 in election audit accuracy in 2006 by the non-profit Brennan Center. The same group ranked NC as being one of the 8 states best able to run the 2008 general election. Election integrity advocates have worked hard to get NC where we are today, and we have to be vigilant to make sure that we know enough about so-called "electoral reforms" like IRV before we decide whether or not we want to implement them.

North Carolina has better elections than South Carolina. SC has open primaries (where you can cross party lines and vote for candidates in other parties even if you are not an Unaffiliated voter). Their elections are run on paperless DRE touchscreen machines that were decertified for us in other states. They can't even tell whether or not there was any election fraud in their Democratic US Senate primary, because to check for fraud in that one race might challenge the integrity of ALL SC elections. Those same machines are used in the SC general election - including for US President - so how could we possibly trust them to correctly record and count any election. And add to that mess the fact that some folks are pushing for National Popular Vote for President to abolish the Electoral College and you can see why we shouldn't be pushing for ANYTHING that will further complicate election integrity.

No comments: