Low voter turnout renews interest in IRV

It would probably have been cheaper to buy each candidate a fancy racecar and have them do 300 laps:

By mid-morning Tuesday, 10 people had voted in Tyrell County and 14 in Camden County. The largest precinct in Wake County, the seat of state government, had 23 voters. Tuesday’s runoff will cost N.C. taxpayers at least $6 million.

It cost a hell of a lot more than that; we could end up with Tea Party Tedesco superintending all the schools in the state. Here's more via our pals at NC Policywatch:

North Carolina has had several IRV elections, and three exit polls show voters overwhelmingly preferred it to returning to the polls for a runoff. Unfortunately, the state’s voting equipment currently requires “workarounds” that delay the count. Once North Carolina has optical scan equipment like others have, it would have an IRV tally to share on election night along with other results.

Methinks that $6 million would have been better spent upgrading our machines, especially considering the idiots in the General Assembly refused millions in Federal voting funds.


Comparing runoff turnout to IRV is apples to oranges

While I join in deploring abysmal voter participation rates, I remain unconvinced that IRV is an answer to the turnout question.

If some consider IRV the answer, shouldn't it follow that the IRV turnout rates should be the goal?

I do not consider 43% voter turnout an acceptable goal. 43% was the overall turnout in the Nov 2010 election when IRV was used for a NC Courts of Appeals race.

I do not deny that holding more elections costs more money -- both for election infrastructure and candidates. However, recent NC experience does not indicate that IRV is an answer to increasing turnout.


Did I do that?

I've reread what I wrote several times, and I don't see anything about IRV increasing voter turnout.

My problem is, the turnout for the runoff (compared to the Primary itself) amounts to a handful of voters deciding races that tens of thousands voted on the first time around.

I don't know. It just seems both pointless and expensive.

IRV does not increase turnout

In fact in locations where IRV is used (like San Francisco), turnout is down since the time they started using it.

Turnout in Minneapolis was the lowest in over 100 years when they used IRV in 2009. And the single IRV election in 2009 cost $365K MORE (adjusted for inflation) than the combined cost of the 2005 primary and general elections.

Oh - and voter confusion was way up!

The same thing happened in 2008 and 2010 after the 2nd runoffs. Yet the two times that IRV was used to calculate "winners", it was too complex to count. And there were no majority winners.

Former Wake BOE member and verified voting advocate Debra Goldberg (not the Wake Board of Education member Debra Goldman) wrote how IRV threatened election integrity.

Not really sure how much interest there is in the GA for IRV. There were one or two bills to kill judicial IRV outright, and I think only 5 legislators out of 170 favored it. At least one of them isn't coming back.

Oh - the same old IRV advocates will trot it out and claim it's the greatest thing since sliced white bread. But it could cost upwards of $20 million to implement in NC, and then another $3-4 million a year for additional voter education. That's hardly saving any money. You could have a 2nd primary ever two years and never catch up with the cost of implementing IRV in our state.

Chris Telesca
Wake County Verified Voting

A few more sites a verified voting perspective of IRV

http://statewideirvnc.blogspot.com/ - it has an even catchier title once you get to the page
http://instantrunoff.blogspot.com & http://irvbad4nc.blogspot.com - by Joyce McCloy of NC Verified Voting
http://aspenelectionreview.blogspot.com/ - by Marilyn Marks who killed IRV in Aspen CO
http://leastevil.blogspot.com/ - by the Approval Voting folks!
http://repealirv.blogspot.com/ - by the folks who killed IRV in Burlington, VT!
Racial and Ethnic Disparities in Voting Behavior under Oakland's Ranked Choice Voting Program which won a Firestone Medal for Excellence in Undergraduate Research at Stanford University. I think it has a little more going for it than the usual pro-IRV puffery pushed by FairytaleVote!

Chris Telesca
Wake County Verified Voting

Suppose they gave an election and nobody came...

What if they held an election and no one showed up?

At four polling places in the state - one each in Guilford, Currituck and Tyrell counties and at Washington GT Magnet Elementary School near downtown Raleigh - no one even showed up to cast a ballot in Tuesday's runoff elections.

Watch the video for the story by Laura Lesilie of WRAL.com, and consider why IRV must be on the table for consideration.

Martha Brock

Did any of you rocket scientist check the precinct voting data?

A Tale of Two Precincts

Washington GT Magnet School is the polling place for precinct 01-25. Do you know where precinct 01-25 is in Wake County? 1000 Fayetteville Street, Raleigh NC 27601

Martha knows where it is - right next to where she lives (or by now used to live) down off South Saunders Street.

The precinct contains 728 voters. Of those, 315 are active voters (43.27%).
47 of those voters cast a vote in the May primary. 14.9% of active voters, and 6.45% of all voters.

34 of those voters were Democrats - and how many of them do you really think cared about the Labor Commissioner race in the 2nd primary?

6 were Republican, and 7 were UNA.

28 voters cast in-person ballots (59.57% of all votes cast). 16 voted Early or Absentee (34%). 2 voted a provisional ballot, and one voted a transfer ballot.

We won't know how many of those folks voted early in the 2nd primary - won't know till the 60 day sort.

How many folks in that precinct voted in the May 2008 primary? 119 - 37% of active voters. But only 3 people voted in the June 2008 3nd primary - also for the NC Labor Commissioner race. 332 voted in the November 2008 race. But that's more than the number of currently active voters in that precinct now.

14 voters voted in the May 2008 primary that didn't vote in the November 2008 race - one person even voted in both the 1st and 2nd primary but didn't vote in November. Think there might be a lot of transient people in that neighborhood? Think also that some freak natural thing (like a tornado) might have touched down and wiped out many houses, forcing people to relocate?

I am Chair of Precinct 01-42 - Millbrook Exchange Club Center. We have 2963 voters registered, with 2748 active. 1520 people voted in the May primary (51.3% of all voters, and 55.31% of active voters). 1194 voted in-person. 323 were Absentee or Early Voting, 2 provisional ballots.

In my precinct for the 2nd primary, 247 people voted in person - with two of those being provisional ballots. That's 8.34% of total voters, and 8.99% of active voters).

13 folks voted early or Absentee. 54 voters were Democrats - with 32 votes for Brooks and 22 for Foster. No overvotes or undervotes. I am the precinct Chair, and my voters get great voter education and they don't cast hardly any under-votes and nearly ZERO over-votes (unless there is an IRV race).

The Republicans had undervotes in every one of the 4 statewide races on their ballots, and one over-vote in the Lt. Governor race. You can draw your own conclusions from that.

What conclusions can you draw from one precinct nearing record turnout for the 2nd runoff while one precinct has ZERO turnout on Election Day? Could other demographics be involved? Like income, education levels, and work/transportation issues? People living on the edge who can't find the time (or a ride) to vote?

So you are going to claim that NC should switch to a little used voting method that no voting machines or software has been tested and certified for use in NC because of what happened in precinct 01-25? How well do you think the

And given those things, do you really think that IRV would have helped wring anymore votes out of 01-25 in the May primary?

Chris Telesca
Wake County Verified Voting

The heat is addling my brain

Here's the title of Rob's diary over at the Pulse:

Dismal runoff numbers show need to make instant runoffs work

And here's mine:

Dismal voter participation renews interest in IRV

Sorry, Rob. I really didn't mean to steal such a dismal word...

Forgot the Thigpen and McCullough Fiasco already?

IRV is sort of a hydraulic fracturing of ballots. It is complex, non transparent and has unintended consequences. IF you don't like runoff elections, don't have them. Most states do not have them.

How soon do you forget?

With the statewide contest for NC Superior Court, between Thigpen and McCullough, we got the full IRV experience:

  1. 48 days after the election, instant runoff voting produced a “winner” for the NC Court of Appeals, for the “Wynn”seat.
  2. Thanks to IRV, an experimental tallying method was used,
  3. the election was almost a tie,
  4. a recount was called for, and
  5. we had a plurality result, not a majority win.

Oh and about buying new optical scanners to tally IRV.
Another bad and expensive idea:

USA Today ran an article about those particular scanners, the are called the ES&S DS 200 optical scanner. The Federal Election Assistance Commission found several major problems with those scanners:

  • Random screen-freezes that prevent ballots from being fed.
  • Failure to log errors in a file that would let election officials know of problems.
  • Skewing of ballots as they’re fed into the machine, making votes cast in some parts of the ballot unreadable.

In Ohio, 10% of the machines failed the pre election tests.

There are other problems not mentioned in the USA Today article, but basically what happens is that the machines flat out don’t count all of the votes, due to these malfunctions.

See “Federal agency finds defects in ballot scanners”
By Gregory Korte, USA TODAY Updated 12/23/2011

IRV violates a core principle of elections - KISS - keep it simple....

Why are statewide elections increasing in cost?

In 2008, IRV supporters claimed that a statewide election runoff had costs ranging from $3 to $4 million

In 2010, IRV supporters claimed that a statewide election runoff cost $5 million!

In 2012, IRV supporters claimed that a statewide election runoff cost $7 million - an increase of $2 million over 2010!

Why has the cost of a statewide election more than doubled (a 200% - 230% increase) in cost from 2008 to 2012?

The number of precincts is roughly the same - 6 extra precincts in Wake County from 2008. We have the machines already paid for. Ballot prices have gone down not up as more counties find their own certified printers. Poll workers salaries are roughly the same. State and county BOE employee salaries haven't really gone up that much. Gas prices are even down from 2008. We even have fewer days for Early Voting in 2012 than from 2008 and 2010.

So why a 230% increase in cost?

Chris Telesca
Wake County Verified Voting

Another factor in the low turnout....

...could be the late date of the 2nd runoff.

In 2008 and 2010, the election date was in late June. This year it's three weeks later.

That's due to a change in election law last year - to allow for more time to send ABM ballots to overseas voters and our troops.

So I wonder what the tradeoff is between opening up the democratic process to overseas voters and our troops, but losing more voters at home?

Is the gain of overseas voters and troops worth the lower turnout?

Chris Telesca
Wake County Verified Voting

I'm not aware of very many

aspects of human behavior and culture that can be reduced down to a single correlation or trade-off. In fact, I'm aware of none.

The variables at play are far too numerous and complex to conduct a reasonable analysis. The economy, the candidates themselves, the miasma of negative advertising, the top of the ticket, the bottom of the ticket, the presence of a referendum, local issues, etc., etc. Asking whether X is THE trade-off to compensate for Y is like a guy wondering if he didn't get laid because of bad breath or crappy table manners.

What's clear is that the current system is (1) not cost-effective and more importantly (2) allows important elections to be decided by an otherwise negligibly small number of voters. The system must and will be changed eventually ... it's just a question of "changed to what?"

I believe it would be impossible to argue that IRV suppresses turnout, either in the short term or over the long term. That said, it most assuredly creates confusion among an electorate that often lacks the brains god gave an animal cracker. Simple is better. And IRV is by no means simple.

I like the plurality idea. Or maybe runoffs could be limited to general elections?

I don't know ... just throwing out ideas.

I read most of your links but didn't find any clarity about what to do instead of IRV.

IRV is complicated

I'm sure there's some engineers and actuaries reading this who are thinking, "It's not complicated at all!" But it is.

Frankly, a simple plurality makes more sense to me than anything else.

Mecklenburg Democratic Turnout

These numbers are only for registered Democrats in each precinct. Our range was 0.07% - 1.17%. Countywide it was 0.60%. Only two precincts (out of 195) had more than 50 people vote (early and election day), 33 had less than 10 people vote with one precinct only having one Democrat vote and that was an early vote.

Well most of the links are there to

debunk the idea that IRV is the greatest thing since sliced whitebread. Once those claims are debunked and IRV comes down off the pedestal and is placed alongside of all the other possibilities (such as lowering the threshold for all primaries, or just lower threshold for the congressional and legislative races and go with straight plurality for the statewide races), then we have a fair comparison.

I get incensed when I go down to the Legislature or to a county commission meeting or city/town council meeting and I hear IRV advocates tell them things that A) have already been disproved other places and B) that the elected public leaders end up find out themselves the hard way.

This reminds me of Julie Robison from Cary, who originally voted to allow Cary to participate in the IRV pilot mainly because she was told that IRV ensures a 50% plus one vote majority win. She never actually checked the results of other IRV races to see if that was correct. Julie worked with RTI in the area of governance and elections - so when she saw the IRV sort/stack/tally process in operation, she was shocked! (as witnessed by myself, Perry Woods, Don Hyatt, Don Frantz and Bryan Lerschall). We witnessed so many irregularities during the count. Don Hyatt and myself caught the numbers not adding up that was later blamed on a "calculator error".

Two days later - after a non-public recount of the IRV votes that no candidates or outside observers were notified of - some missing votes popped up. Not enough to change the results - Don Frantz was declared the winner. Out of the original 3022 votes cast in the Cary District B race, a candidate would have needed 1512 votes or better to win outright. No one met or exceeded that threshold, so they had to look at the 2nd and 3rd column votes. But Don Frantz had the higher total at that point. And at no time during the IRV tabulation did that change. But when all the votes were exhausted, Don only had 1401 votes - not quite the 1512 he would have needed to win in the 1st round. In fact had not one single extra vote been cast for either candidate, Don would have still won. One of the many peculiarities of IRV. It's happened in other races a little further down the rankings - which is even harder to understand. 1401 out of 3022 is 46.36% of the 1st column votes.

And the 2010 statewide judicial IRV race wasn't a majority win either. Judicial IRV was implemented because folks felt that a 24% plurality win for the 8-way 2004 NC Supreme Court race wasn't fair or democratic. So they came up with this mandatory judicial IRV scheme in the same bill as the IRV pilot, but then didn't fund a way to implement it for 4 years. So for the 13-way IRV race in 2010, the winner had just a little under 28% after all the confusion and 45-days of tabulation (including 1 recount). So it's worthwhile asking if the extra cost, confusion and delay were worth it. Those questions need to be asked and answered in the debate over IRV.

IRV most certainly does not increase turnout, but the added confusion for voters (having to consider and rank more than one candidate) can and has turned off voters. Turnout in San Francisco is down since they implemented IRV. And turnout for the 2009 Minneapolis RCV race was the lowest in over 100 years. Costs were $365K higher (adjusted for inflation) for that single RCV race than for the 2005 primary and general election. Costs are higher in San Francisco as well. IRV doubled the costs of elections in Peirce County, WA when they implemented it. When you honestly and accurately include all the costs of doing IRV, it's not cheap.

There has never really been a good and thorough study of IRV done by the SBOE or the General Assembly. Other states have done them - Maryland (home of FairyTaleVote) has done three. I got copies of them, and they show that Maryland is a state with similar election administration and issues that we have. It would cost them around $3.50 per registered voter to implement IRV up front, and about $0.50 per registered voter for voter education every year since it would be used for both state and local elections. And based on a study by a San Francisco grand jury, they found that even the $1.87 per registered voter isn't enough to adequately educate voters.

But let's pretend that it is enough. Transferring those costs to the 6 million NC voters, we'd be looking at $21 million to implement and between $3 million and $11.2 million every year to educate voters. Not just the voters that move into the state every year, but even voters who do IRV only once or twice a year. When you make the ballot more complicated, you have to spend more money on voter education.

Even at $24 million for the first year and $3 million every year beyond that - how many statewide runoffs would you have to have every two years at $7 million to break even with IRV costs? And I don't believe that statewide runoffs cost $7 million in 2012 if they only cost $3.5 to $4 million in 2008. Even if the equipment was available (and it's not available), you wouldn't break even after 6 presidential elections.

All these questions need to be answered fully before you make any decisions about IRV. Lots of research is involved. You gotta dive in deep to get many of the answers, and they won't be glaring out at you on any link provided by me or by FairVote.

Chris Telesca
Wake County Verified Voting

Maybe draw straws, cut a deck of cards, or even arm wrestle...

...to resolve a race without using a runoff - it seems to work for a tie.

But here's my big problem even with lowering the threshold. In traditional top-two runoff elections, the 2nd place finisher in the 1st round can flip the election and win in the runoff 33% of the time. Many things can cause that: new information about the candidate; endorsements from the 1st round losers can go to the 2nd place finisher based on deals or negotiations; etc.

With IRV, the top vote getting in the 1st round tends to win the election in the end about 95% of the time or better with very few races actually being a real majority win (of the first round votes). And I've seen some really screwy IRV races where someone doesn't get a single extra vote and yet you have to go through the extra time and expense of counting the next round and the threshold shrinks and - bingo - the guy wins the election without getting a single extra vote!

Look - we have enough trouble getting folks out to vote or even to run for office these days. Do you really want to ask voters to use a voting method that's confusing and they can't follow it without a degree in math, or have a candidate work their asses off in the campaign (hitting up their friends and family for money) and then not be able to explain just why in the HELL they lost?

In the Aspen IRV election the company their Board of Elections had to pay to come in and administer IRV actually programmed the machines to sort backwards. Luckily the test was such a small one that it bothered someone who stayed behind to look at the results - and they caught the problem before the actual election. But even the pros hired to come in and do the counting got it wrong. Why make it so complicated?

Chris Telesca
Wake County Verified Voting