1. The most fascinating race from a political scientist’s perspective is the Republican primary for Governor. It’s still within 1,500 as I write, and since most of the remaining precincts are from Chicago or suburban Cook County, I would expect the gap to close. The leading candidate has just under 20.5% of the vote, and five different candidates have at least 15% in a seven-way race. It is a poster child exhibit for the need for Instant Runoff Voting.

   We’ve also been tracking voting irregularities, and they’ve been out there, though apparently not as bad as two years ago. Some people were denied Green ballots, based on reports, and others had to get into mild arguments with election judges in order to be able to vote, while still others were apparently given provisional and/or federal-only ballots. This isn’t a minor thing, since we have one primary which is currently sitting at 53-51.

2. If Barack Obama had had his way as a state senator in 2002, instant runoff voting would have been used for these primaries. He was prime sponsor of a bill to implement it. See:
   http://archive.fairvote.org/?page=2188

3. Would someone please tell us HOW you would administer and tally a statewide instant runoff voting election?

   IRV is not additive, in other words, you can’t simply tally the votes at the polling places, and since votes have to be centrally tallied due to sorting and reallocating of votes.

   I sincerely want to know how folks think this will be accomplished.

   Will you hire trucking companies to ship the ballots to the state capitol to a central processing center?

   With thousands of polling sites in one state, how will you secure the chain of custody of the ballots being hauled to the central location?

   I’ve never seen a practical explanation on how such an election would be administered in our country.

4. Joyce, why do assume it would work any differently than it does now? They certainly can tally votes in polling places and send them electronically to a central location for the final tally just like they do now. The precincts add up the votes so Candidate A has X 1st place votes, Y 2nd place votes, and Z 3rd place votes, then that info is sent in for the final tally. Its no different that what we currently do.

5. In NY, lever machines were used for the last time, the police department take the sealed results entered by the poll workers and watched by poll watchers, I was a poll watcher in one of the largest polling sites in Manhattan, to a central location to be entered into a database under a lot of oversite.

6. Joyce,

   Somehow they manage to count ballots in Australia, and they use IRV. Is winning a primary with 20.5% of the vote acceptable to you? If not, what is your alternative? Runoffs tend to be expensive and have considerably lower turnout, and fusion doesn’t work in primaries.

7. In #3, Joyce McCloy argues against IRV by asking a question that does deserve a real answer, even though she provides wrong background information. However, in #4 ILLinoize gets the answer wrong.

   Joyce’s error is to claim that IRV requires central tallying, when what it actually requires is centrally coordinated tallying.

   ILLinoize’s error is to say that what precincts (or counting centers) pass on up is counts of 1st, 2nd and 3rd place votes for candidates, when what is actually passed on are (for centrally tallied processes) ballot images (i.e., for each ballot, who is ranked 1st, 2nd, 3rd, etc. on it) or (for centrally coordinated processes with distributed tallying) counts by round.

   One way of centrally coordinating tallying is by tallying all ballots at a single location, and for IRV elections in a city or a county, that’s often the most practical way to do so. After all, the ballots would be going to a single location anyways as part of normal ballot processing (even if they are counted at the precinct, they still are preserved by county election officials until it’s time to destroy them).

   That’s what had been done in most places using IRV or STV (aka choice voting, the multiple winner system similar to IRV) prior to the days of computerized vote counting. The physical ranked ballots would be taken to a single location and would be sorted and counted there. Now, the ballots are scanned (either at the precincts or at a central location), and the scanned ballot images are input to a computer program that does the sorting and tallying. If the ballots are scanned at the precinct level, memory cards with the ballot images are physically transported together with the actual ballots to the tabulation centers, where the cards are read into the tabulating program and the ballots are in secure custody while being available to verify the tabulation later in the process.

   The way this would be extended to a state-wide centrally tallied process would be for regional counting centers to transmit the ballot images from their regions to the state election official. This wouldn’t require solution of the ballot security issues that moving the physical ballots around the state would, and it wouldn’t raise the communications security issues that electronically uploading information from precincts does. In California at least, county election officials already transmit local results in state and federal races to the Secretary of State’s office via secure communication links from election night until the official canvass is complete.

   The other method of centrally coordinating the tallying is to do the actual counting by rounds in regional counting centers (by hand or by computer, possibly by different methods in different counting centers), while coordinating the rounds from a coordinating center where no ballots or ballot images are present. My understanding is that this is how the Irish count their presidential election. The counting centers count the first round and report the results to the coordinating center (just as is done now in non-IRV elections, where counties report regional results to their state election officials). The coordinating center adds up the first round results from the different regions and determines whether there needs to be a second round and, if so, which candidate or candidates are eliminated, and transmits those instructions to the counting centers. The counting centers then count the second round according to instructions and report their results to the coordinating center, which adds up the results and determines whether there needs to be a third round. And so forth, until there is a winner.

8. #2 It never had a committee hearing.

9. #6 http://www.aec.gov.au/Voting/counting/senate_count.htm

   “Counting for the Senate takes longer than the House of Representatives because of the complex nature of the counting system used. It is some weeks before all Senators are declared elected.”

10. If Barack Obama had his way as a state senator in 2002, Rod Blagojevich would have been elected governor of Illinois. See:

    http://blogs.abcnews.com/politicalpunch/2008/12/questions-arise.html

11. IRV votes can be recorded but not tallied at the polling places. Regular votes can be recorded AND tallied at the polling place.

    IRV is not additive. It is the only election method that you cannot just simply add up the votes at the polling places.

    Fusion Voting (which really works by the way) and Approval and Range voting are all additive and simple to tally. You can count the votes where they were cast. We know that Fusion works and it doesn’t even require a change in voting systems – it is used in a few different states already that have levers, DREs and optical scanners. No special software. Oregon just adopted fusion voting.
    Google Fusion Voting plus North Carolina history to see how powerful it was in that state.

    IRV requires a complex algorithm to decide which votes get eliminated and which votes get re-allocated to which candidate. Some votes don’t get counted, they just get eliminated.

    The few countries that use IRV also have very simple ballots with usually 1-2 contests or party lists. This is just not like what we have in the US.

    A big problem with central counting is that it creates a new channel for vote tampering or problems.

12. Instant Runoff Voting actually causes tie/near-tie election-recount-nightmares to be MORE likely, not less. Simple mathematical fact.

    http://scorevoting.net/TieRisk.html

    Score Voting causes them to be LESS likely.

13. ¡§she said they were able to ‘predict that the north tower was going to fall.’ It did just before 10:30 a.m.”Astounding predictive capability… since no steel-concrete structure tower has ever collapsed due to fire before or since

This site uses Akismet to reduce spam. Learn how your comment data is processed.