Website Search

Thursday, December 10, 2009

Exit poll Hendersonville instant runoff voting shows mixed results

Exit poll does not prove Instant runoff voting was popular in Hendersonville NC. If anything, the exit poll provided inclusive or mixed results. 322 Hendersonville voters were surveyed. Looking at the actual data, there were more than twice as many people who said that they preferred ordinary voting as said that they'd be upset not to have IRV in future elections -- so it's not as if Hendersonville is FIRED UP for IRV. As for whether IRV actually worked, obviously the exit poll doesn't go there.

New Data Support Use Of Instant Run-Off Voting (from Press Release)December 3, 2009 ( -- New data collected as part of a North Carolina State University study during the 2009 municipal election in Hendersonville, N.C., show that voters prefer instant run-off voting (IRV) to traditional voting - a finding that may build support for IRV.

Professor Cobb's survey on IRV is available on his web page

"I recently conducted an exit poll of Hendersonville, NC, a city that used Instant Run-off Voting (IRV): Here is the survey and here are the preliminary results. Here are the 2007 findings."

So this is the actual survey, it takes up two pages of a scantron style paper
Here are the preliminary results
Here are the results of the 2007 survey of Cary and Hendersonville.
Cary has since dropped IRV.

The exit poll methodology is not yet available as of Dec 10, 2009. 322 voters responded to the poll.

Here are the instructions given to exit pollsters

Election results are posted at the State Board of Elections website here:

Hendersonville, located in the county of Henderson. (Do not confuse Hendersonville with the town of Henderson. Henderson is in Vance County and is not the town using IRV. )
Here is the Henderson County Board of Elections website

State law requires that at least one early voting location be open, and in Hendersonville, early voting would be done on the touchscreen voting machines. Early voting is considered "in person absentee voting". There is always one early voting site but more if funding and demand allow. Early voting has become more popular since our state offers same day registration during early voting only.

Voters also had the option of voting by mail, which is fairly popular in Henderson County, since this area has alot of retirees. Absentee by mail ballots are cast on optical scan ballots.

Hendersonville uses E&S iVotronic touchscreen voting machines, with a "paper trail" that prints on a cash register type roll. This is called an RTAL printer, it prints all voter selections and de-selections, and does not print a summary. The screens could be disconcerting to voters who aren't forwarned about IRV.

Here is what that ballot looked like:

Voter Education: Henderson County Board of Elections Director Beverly Cunningham described it as presentations and "educating all votersthrough the Hendersonville TimesNews, WLOS and our website."
Presentations were done at the continuing care retirement community Carolina Village (and also a polling place), city bill-pay dept, League of Women Voters presentation & the Apple Festival

Here is the 5 page single spaced set of instructions for using an excel spreadsheet to tally the IRV votes.
Experts warn that this spreadsheet tallying method is error prone, lacks an audit trail, and is not good enough for elections.
Philip B Stark Comment Exel Workaround
Tom Dahlberg Comment Exel Workaround

The spreadsheet tallying method was never used in 2007 or 2009, since a winner was found in the first round. The additional rankings were never reported or counted to the public.

At this time, we await Professor Cobb's IRV methodology as well as a reply from the NC State Board of Elections in answer to our request for all vote data.

Some questions: Did Hendersonville voters flood the polls to take advantage of the new system? Were their preferences reflected with exquisite precision? Did the system even work correctly? You won't find the answers in the exit poll results...

We have real concerns with the fact that IRV votes were considered back up votes, only to be counted or reported upon later, if it were determined that they were "needed". Votes are votes and should be counted or at least reported to the public. We also have serious concerns about the tallying method, which is not transparent to voters and is error prone.

To receive updates by email visit this link