Wednesday, October 4, 2023

My Response to Terry Mahaffey re the June 2021 NYC RCV primary

Back in 2021, some NC Democrats were pushing a Ranked Choice Voting resolution.  It was tabled at the state Resolutions and Platform Committee level to wait on the results of the June 2021 NYC Primary election where RCV would be used for the first time.  

 

I've been a very vocal opponent of Ranked Choice Voting wherever it has been used - and especially here in NC - where we used Instant Runoff Voting (a form of RCV) in pilot elections in 2007 & 2009, and in 3 mandated judicial elections (including a statewide Court of Appeals race).  In the 4 instances where there was no clear majority winner at the end of the 1st round of counting, IRV never delivered a real majority winner when all ballots were exhausted, and in 3 of the  2010 Judicial races - not only was there no real majority winner, but all three "winners" flipped and beat the 1st round winner.  

 

I posted to social media many times about the dangers of RCV to election integrity and how it didn't work as advertised in NC and in NYC.  Wake County Democrat Terry Mahaffey used to post that he didn't understand my objections to RCV - like the one comment below:

 

I’ve often see you rail against RCV - I’m curious, what is your preferred voting method? Surely you can see the problems with first past the post/plurality, right?

 

It's obvious that you have not read all my posts about IRV/RCV. I've been very clear about why I don't support IRV/RCV and what voting methods I do support. And I have explained my reasons in many many posts on FB and in my own two blogs (as well as others) about why I don't like IRV/RCV – which also include what voting methods I prefer. Obviously you haven't read them. Here are the links to my two blogs: http://noirvnc.blogspot.com/ (about IRV in general); and http://statewideirvnc.blogspot.com/ (about IRV/RCV elections we used in NC).


You can read other blogs on how and why IRV/RCV endangers election integrity here:

http://irvbad4nc.blogspot.com/

http://repealirv.blogspot.com/


And here is a blog posting mentioning a member of the Wake BOE who got a letter to the editor published in opposition to IRV/RCV - http://noirvnc.blogspot.com/2008/07/former-wake-boe-member-debra-goldberg.html


But here we go again. I support our party having a full discussion and study on various voting methods. I personally support majority elections – either in the first race, or through top-two or gradual elimination runoff elections. I don't think any of the bullshit hocus-pocus promises made by IRV/RCV supporters bring anything positive to election integrity in general or will help elect progressive Democrats to public office so they can work to turn our party platform into public policy.


IRV/RCV claims to elect the candidate preferred by the majority of voters. But let's define what the word “majority” means. I feel that it means someone who wins by at least 50% plus one vote of the total number of votes cast in that entire race. Using that definition, there are no IRV/RCV races where a candidate “wins” 50% plus one vote UNLESS that is done using only the votes in the first column. Are you familiar with fractions and percentages? If you use the total number of votes cast in the first column as the denominator, and the votes cast for each candidate as the numerator, you would have to get at least 501 out of 1000 total votes cast in the race (501/1000) to get one vote greater than 50%. It doesn't matter the exact numbers – what matters is that you get one vote more than 50%. Got it so far?


Even a CA court ruled that all IRV/RCV does is give a slightly larger plurality/first past the post vote total. Exactly how they determine the “winner” depends on which IRV/RCV scheme is used, and how they arrive at the “preferential majority”. When we used IRV/RCV in NC in 2007 in Cary, it was top-two IRV. That meant that you determined the top two candidates, then stopped counting ballots where voters voted for them in the first column – those ballots were exhausted. Then you started counting the 2nd column votes for those two candidates on the remaining ballots – and exhausted those ballots. Then you counted the 3rd column votes for those two candidates on the remaining ballots – and whomever got the most votes was declared the winner.


So how did that work out in Cary in 2007? The total number of votes cast in the Cary district race that was determined by IRV was 3022 votes. That's the denominator. So anyone winning by a real 50% plus one vote majority would have been required to win at least one vote more than 1511 (which is 50% of 3022). But no one got 1512 votes – or anywhere close. Don Frantz “won” that race because after all ballots were exhausted, he got 1401 votes – which is 46.36% of 3022.


When it was used in the 13-way 2010 statewide Court of Appeals race, there was very little effective voter education, and no exit polls conducted. After the first round, we had one Democrat (Cressie Thigpen – who lived in my precinct) in first place and one Republican in second place. But no one had a 50% plus one vote majority. So they started tabulating the 2nd and 3rd column votes – and guess what happened? The Republican won – but neither candidate cracked 28%! Some majority win there – right?


And it's interesting that in 2010, there were two other IRV/RCV races for judges – in Superior Court. And in each of those races, the candidate in second place after all the first column votes were cast ended up beating the person in first place at the end of the first round. In all of IRV/RCV world – that never happened before. So why did it happen in NC? As it turns out, in the two smaller races – most all of the votes (except for ABM ballots which are paper ballots) were cast on DRE touchscreen voting machines. In the statewide IRV/RV race, in-person voting was a mix of DRE and paper ballots. From 2006-2019, we used voting machines from one vendor – ES&S – and neither ES&S nor any other voting system vendor had any election systems tested and certified by the federal EAC to run any type of IRV/RCV election. So our SBE hacked together an IRV/RCV counting method. For paper ballots, they first scanned all the ballots and counted all the votes for the top two candidates – and exhausted those ballots. All the remaining ballots were then run through scanners to tabulate the 2nd column votes for the top two candidates – and exhausted those ballots. All the remaining ballots were then run through a third time to tabular the 3rd column votes for the top two candidates.


What happened in the DRE counties? There was a multi-step process to take the DRE votes for 1st/2nd/3rd columns and port them into a spreadsheet, then use the “sort” function to duplicate what the scanners did in the paper counties. It should be noted that all the DRE counties used paper ballots for ABM and so had to have two different tabulation methods. Needless to say none of these methods were first tested and certified by the EAC – then or now. And if you talk to most people who analyze large quantities of data, none of them say that MS Excel has the accuracy to handle hundreds of thousands of numbers accurately – that's why they use customize software in their work.


So in the three IRV/RCV judicial races in 2010, all three of them flipped and the 1st place vote getter in the 1st round lost in 100% of our races. Compare that with the 95-97% rate in all the rest of IRV/RCV world.


Interestingly enough, in elections where they have real separate runoff elections, the 2nd place finisher flips and wins the runoff in 33% of the runoffs. Which is more “democratic”?


So, if IRV/RCV races favor the candidate in first place after the 1st column votes are counted in 95-97% of the races, why go through all the complex and confusing BS with IRV/RCV and just go with first past the post or plurality? If you believe in a real majority, why not have real top-two or gradual elimination runoff elections? In a runoff election, you have another chance to give voters a clear message which candidate to vote for. You also get the chance to have the candidates eliminated in the first election to endorse the remaining candidates – something that in theory can happen with IRV/RCV, but rarely ever happens.


And you would think that one single IRV/RCV election would cost less than a regular election and runoff, but that is not true if you factor in all the costs associated with IRV/RCV: much more expensive election system costs, higher costs for recounts and audits, etc.


Oh – and let's talk about the cost of recounts and auditing. With all the Republican bullshit associated with mail-in voting, and the counts and recounts and auditing of the 2020 election, what do you think that Republicans would do with three or more rounds of counting just to handle the additional columns?


How do you think Republicans would handle finding out about miss-spelled names effecting the accuracy of tabulation (like in Jeff Rose's RCV tabulation of DNC delegate votes) in the CD delegate elections?


If NC's 13-way Court of Appeals was a cluster-fuck, exactly how well do you think our primary-nomination system would have worked with 57 26-way RCV elections taking place from February through June in 2020?


I favor eliminating all these bullshit contests spread out over months and just go with a one-day national primary for President. No more individual or Super Tuesday primaries or caucuses. With a top-two or gradual elimination runoff to decide who our party Presidential nominee would be. After that, all the national party conventions would deal with would be picking a Veep running mate and preparing our national party platform. And there is no way that IRV/RCV should be used even if we had one type of voting machine and tabulating system/software used across the entire country.


Got it now Terry J. Mahaffey?