Ballot blamed for `undervote'
Unusual percentage of Burke voters didn't pick a presidential candidate
MARK JOHNSON Charlotte Enquirer 15 January 2005
RALEIGH - State lawmakers Friday questioned whether Burke County's voting machines, the same kind that lost more than 4,400 votes in Carteret County, failed to record hundreds of presidential votes.
State elections officials, however, visited Burke in December and said the problems were a poorly designed ballot and insufficient voter education.
"A whole bunch of folks didn't vote for president," said Sen. Ellie Kinnaird, D-Orange, during a meeting of a special committee she co-chairs studying how to reform the state's voting systems.
More than 10 percent of Burke's voters did not cast a ballot for president, a rate three to four times higher than the statewide average, state elections officials said.
Justin Moore, a doctoral candidate in computer science at Duke University who has been advising the elections committee, highlighted that among provisional ballot voters, who used a paper ballot, fewer than 5 percent left the presidential vote blank. The rate was nearly 11 percent among those using the Unilect brand electronic voting machines.
State Board of Elections staffers pored over Burke's results and concluded the machines weren't the problem. One factor they pointed to was the ballot design. The screen on the machines in Burke showed the presidential candidates and the straight party ticket vote on the same screen.
A straight-party vote, which s all candidates from the ed party, applies to all races except for president, a distinction in North Carolina that dates back to the 1960s and continues to confuse voters. Burke voters easily could have thought a straight party vote included the presidency unless they carefully read the wording, said Don Wright, general counsel for the state board who visited Burke.
The Unilect machines in Carteret, by comparison, showed the presidential vote first, on a separate screen, before the straight party vote. Their "undervote" for president was about average, Wright said.
The Unilect machines don't produce a paper ballot, but state officials examined the paper absentee and provisional ballots and found that more than 95 percent of those who didn't vote for president were straight-party voters.
The confusion extended to absentee voters who mailed in ballots in Burke, more than 8 percent of whom did not vote for president, Wright said. Provisional voters, those who show up at the polls and aren't on the voter rolls, get a paper ballot. In Burke, they also received individual guidance from a precinct worker, including a reminder about the presidential race.
"Those voters had the benefit of one-on-one consulting," Wright said.
Moore said confusion over the presidential race doesn't explain it all. A higher-than-average percentage of Burke voters also didn't vote for senator or governor and other statewide races. Burke's provisional ballot voters, who used paper, had a much lower rate of skipping those races than voters who used the machines.
"For all the races," Moore said, "the undervote among provisional voters was about half of what it was on the machines."