Audit of Aspen's election shows no holes in systemRandom hand count of ballots match up with IRV system
By Carolyn Sackariason
Published May 8th 2009 in Aspen Times
ASPEN — Aspen residents can be assured that their votes in Tuesday’s election were counted correctly, based on a random hand count of the ballots.
Local election officials, city attorneys and a few observers participated in a post-election audit in the basement of City Hall on Thursday. Roughly 10 percent of the 2,544 ballots cast were manually checked against the election administrators’’ computer-scanned images of the ballots.
City Clerk Kathryn Koch and Pitkin County Clerk and Recorder Janice Vos Caudill pulled a total of 260 ballots out of sealed bags — 65 from each of the four precincts. Koch read aloud each voter’s rankings of the 13 candidates, who were assigned numbers in the computer program as they were listed in order on the ballot.
The election administrators, True Ballot Inc., assigned those numbers to what’s referred to as a “string” when they scanned and took images of the ballots on Tuesday night.
On Thursday, Koch’s visual check of the original ballots were matched up against what the computer program had calculated for the rankings. All 260 hand-counted ballots matched what had been scanned Tuesday night.
The audit was conducted to test Aspen’s new instant runoff voting (IRV) system, which was the first one conducted in the country that involved multiple council seats. Other municipalities use IRV but their elections so far have only involved one-seat races; Aspen had two council seats open.
“We are just confirming that the ballot reflects the string,” said Jim True, special counsel for the city of Aspen.
Jim Perry — who was appointed as mayoral candidate Marilyn Marks’ poll watcher and who levied a challenge Tuesday afternoon questioning the procedures in which 801 people cast their votes in the City Clerk’s office during absentee voting — observed Thursday’s audit. He said he was satisfied with the audit results.
Perry’s challenge remains open, but he said he wouldn’t take it further, echoing Marks’ sentiments. If a candidate wants to carry on with the challenge, it will have to be taken up in district court.
Precinct 5, which includes walk-in and mail-in absentee ballots, were left out of the hand count because of Perry’s challenge.
Incumbent Jackie Kasabach, who lost in Tuesday’s election, also was on hand during Thursday’s audit.
“I’m still trying to figure it out, and I’m hoping this will help me understand,” she said shortly before the hand count. She, too, said she was satisfied with the results.
There were several steps taken to tabulate votes in the new IRV system, one of which was scanning the ballots and taking images of them. Another layer of the computer program calculated the top vote-getters by how people ranked their candidates by preference. That tabulation was conducted very quickly on election night before the winners were announced.
The tabulations can be tested by obtaining an open source code that is part of True Ballot’s computer program. The vote tallies and information on where to obtain the computer code is available at www.aspenpitkin.com/depts/38.
Koch said neither local or state election laws require an audit. But because it was a new voting system, the Aspen City Council asked for further assurances. In response, Koch said she would perform an audit of 10 percent of the ballots.
“As a council, we were concerned,” Kasabach said. “We didn’t have to do it, we chose to.”
Local election officials, city attorneys and a few observers participated in a post-election audit in the basement of City Hall on Thursday. Roughly 10 percent of the 2,544 ballots cast were manually checked against the election administrators’’ computer-scanned images of the ballots.
City Clerk Kathryn Koch and Pitkin County Clerk and Recorder Janice Vos Caudill pulled a total of 260 ballots out of sealed bags — 65 from each of the four precincts. Koch read aloud each voter’s rankings of the 13 candidates, who were assigned numbers in the computer program as they were listed in order on the ballot.
The election administrators, True Ballot Inc., assigned those numbers to what’s referred to as a “string” when they scanned and took images of the ballots on Tuesday night.
On Thursday, Koch’s visual check of the original ballots were matched up against what the computer program had calculated for the rankings. All 260 hand-counted ballots matched what had been scanned Tuesday night.
The audit was conducted to test Aspen’s new instant runoff voting (IRV) system, which was the first one conducted in the country that involved multiple council seats. Other municipalities use IRV but their elections so far have only involved one-seat races; Aspen had two council seats open.
“We are just confirming that the ballot reflects the string,” said Jim True, special counsel for the city of Aspen.
Jim Perry — who was appointed as mayoral candidate Marilyn Marks’ poll watcher and who levied a challenge Tuesday afternoon questioning the procedures in which 801 people cast their votes in the City Clerk’s office during absentee voting — observed Thursday’s audit. He said he was satisfied with the audit results.
Perry’s challenge remains open, but he said he wouldn’t take it further, echoing Marks’ sentiments. If a candidate wants to carry on with the challenge, it will have to be taken up in district court.
Precinct 5, which includes walk-in and mail-in absentee ballots, were left out of the hand count because of Perry’s challenge.
Incumbent Jackie Kasabach, who lost in Tuesday’s election, also was on hand during Thursday’s audit.
“I’m still trying to figure it out, and I’m hoping this will help me understand,” she said shortly before the hand count. She, too, said she was satisfied with the results.
There were several steps taken to tabulate votes in the new IRV system, one of which was scanning the ballots and taking images of them. Another layer of the computer program calculated the top vote-getters by how people ranked their candidates by preference. That tabulation was conducted very quickly on election night before the winners were announced.
The tabulations can be tested by obtaining an open source code that is part of True Ballot’s computer program. The vote tallies and information on where to obtain the computer code is available at www.aspenpitkin.com/depts/38.
Koch said neither local or state election laws require an audit. But because it was a new voting system, the Aspen City Council asked for further assurances. In response, Koch said she would perform an audit of 10 percent of the ballots.
“As a council, we were concerned,” Kasabach said. “We didn’t have to do it, we chose to.”
Election Day '09 was a roller-coaster for election reformers. Instant runoff voting had a great night in Minnesota, where St. Paul voters chose to implement IRV for its city elections, and Minneapolis voters used IRV for the first time—with local media touting it as a big success. As the Star-Tribune noted in endorsing IRV for St. Paul, Tuesday’s elections give the Twin Cities a chance to show the whole state of Minnesota the benefits of adopting IRV. There were disappointments in Lowell and Pierce County too, but high-profile multi-candidate races in New Jersey and New York keep policymakers focused on ways to reform elections; the Baltimore Sun and Miami Herald were among many newspapers publishing commentary from FairVote board member and former presidential candidate John Anderson on how IRV can mitigate the problems of plurality elections.