If I were crazy, this is how I would count an instant runoff election. And guess what, there are election officials in North Carolina willing to get a little crazy with our votes!
In 2007, 2 cities in North Carolina were allowed to have an Instant Runoff experiment. Since there is no federally approved software to count instant runoff voting, the NC State Board of Elections set up a "workaround" to help out Henderson County, NC, a touchscreen jurisdiction in the test. Luckily there was no "runoff" so the work around was not used. The pilot program expired but hungering for more disaster, our lawmakers voted to experiment again! In this case, a boutique-style election fad trumps election transparency.
Keep in mind that North Carolina's 20 touchscreen counties have the toilet paper ballot, the VVPAT does NOT print a summary but prints selections and deselections, and during early voting one machine may have 80-100 different ballot styles on it. Manual recounts or audits will be impossible.
Read the NC State Board of Elections' instructions for counting IRV in touchscreen counties if they had incurred an "instant runoff" . One single error would change the outcome of the election. (at the bottom of the page we state how this workaround likely is illegal):
To tabulate a runoff election follow these procedures: 1. Announce the two (2) candidates that are in the Instant Runoff. 2. Print document - Hendersonville IRV-Ballot Position Numbers.pdf to determine the voting positions for each candidate in the Instant Runoff. 3. Remove the "Compact Flash Cards" from the iVotronic voting devices in the City of Hendersonville precincts. 4. Capture the Election Data in ERM; a. Clear Audit Data in ERM. b. Import Compact Flash Audit Data c. Collect Audit Data - From Specified Drive – c:\elecdata\7GNCHEND\GNGFLASH\ADT. d. Select machines from Armory Precinct only. e. Consolidate Audit Data. f. Create Vote Image Log. g. Print Vote Image Log – Select Contest/Precinct, Numbers Only, Printer. h. Select EL155 from the Report File Utility and click on Copy – Name the file IRV_Armory.txt and copy to a location that you can retrieve from (desktop, portable flash drive, etc). i. Clear Audit Data in ERM. j. Collect Audit Data - From Specified Drive –c:\elecdata\7GNCHEND\GNGFLASH\ADT. k. Select machines from Southwest Precinct only. l. Consolidate Audit Data. m. Create Vote Image Log. n. Print Vote Image Log – Select Contest/Precinct, Numbers Only, Printer. o. Select EL155 from the Report File Utility and click on Copy – Name the file IRV_Southwest.txt and copy to a location that you can retrieve from (desktop, portable flash drive, etc). p. Clear Audit Data in ERM. q. Collect Audit Data - From Specified Drive –c:\elecdata\7GNCHEND\GNGFLASH\ADT. r. Select machines from all remaining Hendersonville precinct machines. s. Consolidate Audit Data. t. Create Vote Image Log. u. Print Vote Image Log – Select Contest/Precinct, Numbers Only, Printer. v. Select EL155 from the Report File Utility and click on Copy – Name the file IRV_All and copy to a location that you can retrieve from (desktop, portable flash drive, etc). 5. Open a blank Excel Spreadsheet and import. a. Change "files of type" to All Files - Select file IRV_Armory.txt to import – Click on OK. b. Select Fixed width and click on Next. c. Place separators at 10, 15, 20, 25, 30, 35, 40, 45 and 50. d. Press Next and then Finish. Save the Excel File with the name IRV_Armory.xls and a location that you can retrieve. 6. Open a blank Excel Spreadsheet and import. a. Change "files of type" to All Files - Select file IRV_Southwest.txt to import – Click on OK. b. Select Fixed width and click on Next. c. Place separators at 10, 15, 20, 25, 30, 35, 40, 45 and 50. d. Press Next and then Finish. Save the Excel File with the name IRV_Southwest.xls and a location that you can retrieve. 7. Open a blank Excel Spreadsheet and import. a. Change "files of type" to All Files - Select file IRV_All.txt to import – Click on OK. b. Select Fixed width and click on Next. c. Place separators at 10, 15, 20, 25, 30, 35, 40, 45 and 50. d. Press Next and then Finish. Save the Excel File with the name IRV_All.xls and a location that you can retrieve. 8. Open file Hendersonville IRV Tabulation Form.xls. 9. Open the Excel file that you created in Step 5d. a. Delete Column B b. Sort data on Column A c. Delete all rows without machine numbers. d. Highlight all the data for Armory Precinct and copy. e. Copy the data into Hendersonville IRV Tabulation Form.xls in the RED Tab 1st-2nd Choice – click in Cell A9 and then copy. f. Highlight the imported data (A9 to the end) and sort on Column B g. Using the file created in Step 5d, highlight all the data for Southwest Precinct and copy. h. Copy the data into Hendersonville IRV Tabulation Form.xls in the BLUE Tab 1st-2nd Choice – click in Cell A9 and then copy. i. Highlight the imported data (A9 to the end) and sort on Column B j. Using the file created in Step 5d, highlight all the data for all the other precincts in the City of Hendersonville (excluding Armory and Southwest) and copy. k. Copy the data into Hendersonville IRV Tabulation Form.xls in the BLACK Tab 1st-2nd Choice – click in Cell A9 and then copy. l. Highlight the imported data (A9 to the end) and sort on Column B. 10. Verify that the vote totals for the candidates match the ERM Report. a. Click on Yellow Tab Grand Totals – Totals for each candidate should match the report on ERM. b. Print copy of YELLOW Tab Grand Totals. 11. Click on BLACK Tab 3rd Choice – Remove all votes for Runoff candidates from 1st & 2nd Choice. a. Column B should already be sorted. b. Highlight all rows that have a vote for the runoff candidates from 1st & 2nd Choice (Voting positions 3 thru 7 – example: Caldwell=3, Caraker=4, etc.). c. Delete the selected rows. d. The votes for the runoff candidates should now reflect the votes cast for the runoff candidates that were a 3rd Choice but not a 1st or 2nd choice (Voting positions 13 thru 17). e. Highlight the remaining data (A8 to the end) and copy. 12. Click on BLACK Tab 4th Choice – Remove all votes for Runoff candidates from 3rd Choice. a. Paste the data from Step 9e into cell 9a. b. Highlight all rows that have a vote for the runoff candidates from 3rd Choice (Voting positions 13 thru 17 – example: Caldwell=13, Caraker=14, etc.). c. Delete the selected rows. d. The votes for the runoff candidates should now reflect the votes cast for the runoff candidates that were a 4th Choice but not a 1st, 2nd, or 3rd choice. e. Highlight the remaining data (A8 to the end) and copy. 13. Click on BLACK Tab 5th Choice – Remove all votes for Runoff candidates from 4th Choice. a. Paste the data from Step 10e into cell 9a. b. Highlight all rows that have a vote for the runoff candidates from 4th Choice (Voting positions 20 thru 24 – example: Caldwell=20, Caraker=21, etc.). c. Delete the selected rows. d. The votes for the runoff candidates should now reflect the votes cast for the runoff candidates that were a 5th Choice but not a 1st, 2nd, 3rd or 4th choice. 14. Click on RED Tab 3rd Choice – Remove all votes for Runoff candidates from 1st & 2nd Choice in Armory Precinct. a. Column B should already be sorted. b. Highlight all rows that have a vote for the runoff candidates from 1st & 2nd Choice (Voting positions 16 thru 20 – example: Caldwell=16, Caraker=17, etc.). c. Delete the selected rows. d. The votes for the runoff candidates should now reflect the votes cast for the runoff candidates that were a 3rd Choice but not a 1st or 2nd choice (Voting positions 26 thru 30). e. Highlight the remaining data (A8 to the end) and copy. 15. Click on RED Tab 4th Choice – Remove all votes for Runoff candidates from 3rd Choice. a. Paste the data from Step 12e into cell 9a. b. Highlight all rows that have a vote for the runoff candidates from 3rd Choice (Voting positions 26 thru 30 – example: Caldwell=26, Caraker=27, etc.). c. Delete the selected rows. d. The votes for the runoff candidates should now reflect the votes cast for the runoff candidates that were a 4th Choice but not a 1st, 2nd, or 3rd choice. e. Highlight the remaining data (A8 to the end) and copy. 16. Click on RED Tab 5th Choice – Remove all votes for Runoff candidates from 4th Choice. a. Paste the data from Step 13e into cell 9a. b. Highlight all rows that have a vote for the runoff candidates from 4th Choice (Voting positions 33 thru 37 – example: Caldwell=33, Caraker=34, etc.). c. Delete the selected rows. d. The votes for the runoff candidates should now reflect the votes cast for the runoff candidates that were a 5th Choice but not a 1st, 2nd, 3rd or 4th choice. 17. Click on BLUE Tab 3rd Choice – Remove all votes for Runoff candidates from 1st & 2nd Choice in Armory Precinct. a. Column B should already be sorted. b. Highlight all rows that have a vote for the runoff candidates from 1st & 2nd Choice (Voting positions 11 thru 15 – example: Caldwell=11, Caraker=12, etc.). c. Delete the selected rows. d. The votes for the runoff candidates should now reflect the votes cast for the runoff candidates that were a 3rd Choice but not a 1st or 2nd choice (Voting positions 21 thru 25). e. Highlight the remaining data (A8 to the end) and copy. 18. Click on BLUE Tab 4th Choice – Remove all votes for Runoff candidates from 3rd Choice. a. Paste the data from Step 15e into cell 9a. b. Highlight all rows that have a vote for the runoff candidates from 3rd Choice (Voting positions 21 thru 25 – example: Caldwell=21, Caraker=22, etc.). c. Delete the selected rows. d. The votes for the runoff candidates should now reflect the votes cast for the runoff candidates that were a 4th Choice but not a 1st, 2nd, or 3rd choice. e. Highlight the remaining data (A8 to the end) and copy. 19. Click on BLUE Tab 5th Choice – Remove all votes for Runoff candidates from 4th Choice. a. Paste the data from Step 16e into cell 9a. b. Highlight all rows that have a vote for the runoff candidates from 4th Choice (Voting positions 28 thru 32 – example: Caldwell=28, Caraker=29, etc.). c. Delete the selected rows. d. The votes for the runoff candidates should now reflect the votes cast for the runoff candidates that were a 5th Choice but not a 1st, 2nd, 3rd or 4th choice. 20. Click on YELLOW Tab Grand Totals – The votes displayed in the grand totals for the Runoff Candidates should be the final results. http://www.ncvoter.net/downloads/Henderson_County_IRV%20Tabulation.pdf
If you got this far, here are the problems with this:
1. This work around removes vote data from the ES&S Unity system to a system not tested with it - exporting data first to notepad/wordpad and then excell to tabulate the votes.
a. neither word pad, note pad or excell have been tested for their vote tabulation ability. b. this process erases audit data as it progresses, excell doesn't have an audit trail, and some versions of excell have bugs. c. it is not known what happens to the data as it is moved from the ES&S vote tabulation system to a non ES&S vote tabulation system.
2. All parts of the vote tabulation system must be federally tested together, to ensure they work together.
3. There are over 100 steps in the process, with instructions like "click on the red tab, or click on the blue tab".
One single keystroke error would change the outcome of the election, and there is no audit trail for this process. Audit data is deleted as steps are performed.
www.ncvoter.net
Founder of the NC Coalition for Verified Voting. We passed a law to require VVPB on August 2005 after years of work. NC Coalition for Verified Voting is an all volunteer organization that does not solicit or accept donations.
STV is a voting system which empowers people by reconciling proportionality and preference voting, so it has much to recommend it - though not unless it can be harnessed to smaller wards/constituencies not bigger and bigger ones to secure more and more propostionality [which it can be through Participagtory Weighted Voting].
But I suspect that one reason why some interest groups support the Single Tranferrable Vote is because its supposedly (too complex for quick results) counting system means that STV is seen as a convenient cover for the E-voting and E-counting corporate wide boys to come in and get rich and more to the point serve the CIA or MI5 or Mossad or etc whoever set up these companines in the first place.
Actually with citizen 'jury' type involvement in monitoring the counting LOCALLY and sending up of the results at each stage publicly to the centre and public tabulation and calculations at the centre and then instructions downwards to e.g. eliminate candiate F and redistribute according to F's second preferences ,,,, then I believe that the results of even quite complex single transferrable vote elections could be delivered in a working day, starting fresh in the morning.
IN Scotland we were told that the STV counting programme had been rigourously tested, but no audit trail was really offered re the tested programme being the one installed in all the machines which counted the votes in the Scottihs election on May3rd 2007. Quite apart from all the other secutiy vulnerabilities of all the myriad other bits of software/malware? in all the machines.
by
Keith Mothersson (5 articles, 0 quicklinks, 0 diaries, 39 comments)
on Monday, July 21, 2008 at 7:15:52 PM
all states and nations w/RV use optical scan & hand counts
Joyce -- As you know, but often seem to forget, every single jurisdiction in all of the states and nations with IRV use paper ballots -- with optical scan counting or hand-counting. The leader of the Irish effort to block touchscreens is a big fan of IRV in her country, for example.
You like to say IRV is a Trojan Horse for touchscreens, but don't you find it odd that none of these other jurisdictions are moving to DRE's to run IRV? And might you feel just a tad remorseful that it was your vigorous oppostion in 2005 to North Carolina requiring voting equipment to come ready to run IRV elections that now keeps the state from being able to run IRV elections on current optical scan systems without some degree of hand-sorting and hand-counting? First you create the problem, then you point fingers saying "see, there's a problem."
Meanwhile, there are very real problems with our electoral process that IRV addresses well. Others in your state see that, fortunately -- that's why the NC state legislature just extended the IRV pilot program and why the NC League of Women Voters overwhelmingly endorsed IRV at its recent annual meeting. See www.ncvotes123.com and www.instantrunoff.com
by
Rob Richie (4 articles, 0 quicklinks, 0 diaries, 16 comments)
on Thursday, July 24, 2008 at 5:33:10 PM
How San Francisco almost got touchscreens, Fair Vote's part
Would Fair Vote lobby to prevent touchscreen counties in North Carolina from implementing IRV?
Will you make a statement to the North Carolina State Board of Elections and the State Legislature opposing ANY use of touchscreens in North Carolina elections? So far you haven't.
Check out Fair Vote's furious reaction when San Francisco's election director John Arntz would not switch to touch screens. Fair Vote used to be called The Center for Voting and Democracy. Screen shot saved:
October 7 "Elections Director John Arntz goes against advice from his vendor, members of the Elections Commission, and staff fromthe Center for Voting and Democracy, (Fair Vote) and makes a decision to have ES&S upgrade the Optech Eagles (precinct-based scanner) for counting IRV ballots, rather than using the Optech IV-C (centralized scanner) or touch screens. "
It turns out that John Arntz was a hero, not a fumbler and bumbler as painted by Fair Vote. People should visit the webpage and read the entire thing.
Thanks to North Carolina's crazy 100 plus step "work around", IRV can be "automated" in the touchscreen counties, whereas it has to be manually counted in the state's optical scan counties.
Can anyone say IRV will incentivize touchscreens in North Carolina and also make manual counts and audits prohibitive?
And Cary North Carolina had trouble just counting 3,000 IRV ballots by hand:
A Knightdale resident said. “If the best board of elections in North Carolina had this much trouble counting 3,000 votes, this is too dangerous to try statewide.” (January 22, 2008 Opinion mixed on Cary's instant-runoff trial http://www.carynews.com/news/story/8057.html )
Does Fair Vote have a recommended manual audit protocol for the touch screen toilet paper ballots?
And why is it that jurisdictions like Scotland switched from hand counted paper ballots to a form of computerized voting?So what if it was optical scan, Scotland was much better off with hand counted paper ballots. Look at the mess, which was caused by a multitude of events including IRV/STV Single Transferrable Vote. Would Scotland have purchased voting machines if not for this ranked choice voting mess?
"SCOTLAND'S status as a modern democracy was dealt a grievous blow yesterday by a scandal in which up to one in ten votes in the Holyrood election were thrown in the bin uncounted.
San Francisco, Basalt Colorado, Burlington Vermont, Pierce County Washington, Takoma Park Maryland, and Aspen Colorado have implemented IRV or will implement it for November 2008. * if a reader has updates or new info please email us.
As for San Francisco:
San Francisco, California -They purchased Seqouioa optical scanners and touchscreens, (more money again) will manually count the IRV contests unless the Sequoia software obtains federal certification by Nov 2008. In 2007 they found that the equipment, ES&S Optech Eagles that were conditionally approved to count IRV, were reported by the Secretary of State as having flawed algorithm. So they used systems with a flawed IRV algorithm for 3 years before they found out it had problems.
Pierce County officials said it was too hard to count just 14 ballots by hand, so they are conditionally certifying the Sequoia software (will they find problems with the algorithm 3 years later):
Pierce County will implement IRV in November 2008. Pierce County needed "emergency" certification for their voting equipment because it had not received federal certification yet. Flaws were uncovered in the software when interacting with the Sequoia Insight Optical scanners. As a result, these scanners, used for precinct voting - will not be allowed. The Sequoia Edge DRE/touchscreens and the Sequoia Central Count Scanner will be used. Pierce is mostly vote by mail, so the touchscreens might not get alot of use. This Sequoia Equipment is not federally certified. See the report at Voters Unite.
Pierce County renamed IRV and now calls it Ranked Choice Voting.
IRV is a lot of trouble to get the same results as a plurality election:
Out of 20 RCV elections that have been held since the referendum establishing it passed, when IRV was used, it elected a plurality winner.
4 elections were uncontested. 9 elections were won with a majority in the first round.
7 elections went into RCV vote transfer. For every one of the 7 elections using RCV, the candidate with a plurality in the first round went on to win the election after transfers were complete. In every one of these elections, the winner did not obtain a majority of ballots cast. In every one of these elections, the runner-up in the first round remained the runner-up in the last round.
One election had 22 candidates; the winner had 33.65% of the vote.
by
ncvoter (10 articles, 1 quicklinks, 1 diaries, 103 comments)
on Thursday, July 24, 2008 at 8:15:36 PM
... that dialogues with you on this can go on endlessly. But there are reasons that people like the NC state legislature and the NC League of Women Voters have heard all that you have to say and this year gone foward with support for IRV.
As one of many examples that could be challenged, you know very well that Scotland used optical scan ballots last year and that reviews of that election show that voter handled the ranked choice ballots just as expected -- the problems were in the pluralty vote elections due to bad ballot design.
There will be a good opportunity to work out procedures for running IRV elections in North Carolina. We hope and expect this process will develop secure procedures.
As to FairVote's position on DRE's, we didn't recognize their problems as much as we should have back in the 2001-3 era, as was true of many others. Thanks to you and others for making it clear how problematic they are. Our position on DRE's is here: www.fairvote.org/machines
Looking forward, future implementation efforts or IRV are all focused on optical scan machines, as will be the case in Pierce County (I believe only people with disabilities will use the DRE system, as they will for the non-IRV election there., San Francisco and so on.
You can write some other screed, but I'm not going to respond to it. Onward.
by
Rob Richie (4 articles, 0 quicklinks, 0 diaries, 16 comments)
on Friday, July 25, 2008 at 1:21:39 PM
If IRV actually helped ANYONE, it might be worth the trouble
STV/ranked voting pushed Scotland into giving up their hand counted paper ballots. WHY? The Director of the Electoral Reform Society is also on the BoD of Fair Vote, and clearly you guys communicate.
Clearly Scottish voters had trouble with all of the chaos and hot mess that STV, computerized voting (which didn't work well by the way) and the mixed ballots (did the non profits advise that or urge against this mass confusion?)
In spite of over 100,000 ballots being spoiled, Fair Vote said the election went well.
And here comes Rob Richie saying that if the League of Women Voters supports IRV, it must be good.
Well, lets see, the League of Women Voters opposed paper ballots for years, out right opposed them.
So people may be taken in by Fair Vote's talking points, but we see that IRV doesn't work well in the US, hasn't helped third parties, doesn't help minorities. IRV makes casting and counting ballots harder.
IRV creates new costs, shifting them from the obvious cost of a one on one runoff, to the back end - new machines, new software, more complex programming, more laborious recounts or audits.
IRV DOES incentivize more complex technology (special software to count IRV's complex algorithm, ballot images etc).
It makes fraud and miscounts harder to detect, and easier to hide.
Officials in Pierce County Washington are scared to try to manually count IRV - they said just counting 14 ballots was too hard.
If IRV actually helped ANYONE, it might be worth spending the time and money to try to make it transparent.
by
ncvoter (10 articles, 1 quicklinks, 1 diaries, 103 comments)
on Friday, July 25, 2008 at 1:33:04 PM
The key to spreading IRV is the massive PR campaign as outlined in Fair Vote's 2005 990 filings.
"Making a strong case and neutralizing any opposition: By getting out early with the benefits of IRV, we headed off criticism of the system that would have accumulated naturally in a vacuum. The landslide win is a testament to this strategy, as history shows voters who don't understand a measure generally vote against it."
The arguments for IRV have been about who likes it - instead of citing proof that it works. Even Fair Vote admits people aren't likely to understand it.
Show the proof that IRV works - instead of attacking those who say IRV doesn't work.
I'm not interested in whether the North Carolina League of Women Voters supports it. I want to see proof that it has helped.
Looking at Australia where IRV has been used for decades, it is helping the top two parties maintain dominence.
If people are looking for an election method that helps third parties or brings some oxygen to the system, then look at other election methods.
Don't turn our election system upside down only to later have to admit that it was a gigantic mistake. Learn from how IRV has worked/not worked in places that already have it.
Imagine how hard it would be to gain acceptance for better election methods after you have pushed one that doesn't work and creates problems?
Google "election methods" and look around the internets.
by
ncvoter (10 articles, 1 quicklinks, 1 diaries, 103 comments)
on Friday, July 25, 2008 at 1:43:32 PM
