Electronic voting proves to be winner in election
Fewer missed votes than with other methods

By Ken McCall and Lynn Hulsey

Dayton Daily News  14 November 2004

In an election where talk of voting irregularities continues to heat up the Internet, electronic voting machines appeared to score a victory in Ohio.
In the Nov. 2 election, counties that used electronic voting machines had a dramatically lower rate of ballots that did not register a valid presidential vote, according a Dayton Daily News analysis of the unofficial state voting data.

"These data yield a strong endorsement of electronic voting machines, at least with regard to the troublesome undercounts that plagued the 2000 vote in Florida and elsewhere," said Larry J. Sabato, director of the Center for Politics at the University of Virginia.

In the seven Ohio counties that voted electronically only 0.8 percent of the ballots cast, or 6,610 of 870,237, did not register a vote for president.

The so-called undercount was 1.9 percent, almost 2? times higher, in the 68 counties that used punch-card voting. At total of 76,068 punch-card ballots did not register a vote for president.

Results in the 13 counties that used optical scanning technology were only slightly better than punch cards and still had almost twice the undercount rate of the counties that voted electronically. The optical scan counties recorded a 1.5 percent under count, or 9,994 of 670,058 ballots cast.

Statewide, 92,672 ballots did not register a presidential vote. This undercount can happen for several reasons:

? Over-voting, or when a voter mistakenly votes for two candidates ? something that cannot happen on the electronic machines.

? Errors by ballot-reading machines, including cases where the machine doesn't count a punch card vote with a hanging chad or a paper ballot that was poorly marked.

? People not voting in a particular race. The voter's choices in other races are still counted but nothing is recorded in the race for which there was no vote.

There were scattered problems reported around Ohio with the electronic machines Nov. 2, and concerns remain about the potential for computer hacking and the lack of a paper trail ? something Ohio is requiring for the electronic machines that must be used statewide by 2006.

But the electronic machines give voters the chance to review their choices before registering their vote, reducing the opportunity for error, said Steve Harsman, deputy director of the Montgomery County Board of Elections.

The machine's ability to prevent accidental over-voting is a welcome change, Sabato said.

"Less well-educated, less well-off voters are the most likely beneficiaries of this change, given that over-voting happens most often with these groups," he said.

Herb Asher, political scientist at Ohio State University, said the statewide undercount of 1.7 percent likely masks high undercounts of 7 percent to 10 percent in poorer, less-educated, and often urban precincts; and very low undercounts in more affluent areas.

"All that will be reflecting is poor people, the less-educated, are going to have more difficulty with the punch card," he said.

The Daily News analysis also found that Republicans essentially battled Democrats to a draw in the registration and get-out-the-vote ground war.

Republican President Bush carried Ohio by 136,503 votes over Democrat Sen. John Kerry, according to unofficial results. Those results don't include provisional ballots or out-of-country absentee ballots, that have yet to be counted.

In 2000, Bush carried Ohio over Democrat Al Gore by about 3.5 percent. Kerry managed to get closer, but still fell 2.5 percentage points short.

Turnout in the Ohio counties Bush won was 71.77 percent, compared with a 68.05 percent in counties won by Kerry. Urban areas, which favor Democrats, generally see lower turnout than suburban areas, but in both Bush and Kerry counties this year, turnout was higher than four years ago.

While Kerry's counties had slightly more registered voters (4.09 million) than Bush's counties (3.88 million), the number of votes cast was strikingly close, coming within 3,136 of each other.

"It's important to stress here that both parties did a superb job in getting their people registered and out to vote," Sabato said. "But Democrats have re-learned a vital lesson: mobilization in politics begets counter-mobilization. So the well-publicized Democratic get-out-the-vote programs also generated an energetic Republican program."

In fact, the chairmen of the Ohio Democratic Party and the Ohio Republican Party each praised the other side for how well it mobilized voters.

Robert Bennett, state GOP chairman, said the secret weapon for both was effective use of volunteers at the grass roots level.

He said in the 72 hours before the election he got on the phone with party leaders in the 57 smallest Ohio counties. Bennett told them that Kerry would win Cuyahoga County by 180,000 votes ? he ended up winning it by more than 217,000 ? and said he was counting on the smaller counties to bring out enough votes for Bush to make up the difference. It worked. Those counties accounted for 274,000 votes for Bush, he said.

Nancy Martorano, assistant professor of political science at the University of Dayton, said that, while Kerry lost Ohio, it is important to note that the ratio ? 51 percent to 48.5 percent was very tight.

"I think Kerry had his people fired up. In the end I don't think there was any distinct failure by the Kerry campaign in Ohio," she said.

Even though Kerry won critical counties such as Cuyahoga, Franklin and Montgomery, it simply wasn't enough.

"In fact the Democrats did very well in their areas and the Republicans did very well in their areas," Asher said. "I guess we had always underestimated how many Republicans were out there who had never voted."

"Certainly Issue 1 helped them with turnout," said Asher, referring to the constitutional amendment to ban gay marriage.

Both sides spent millions of dollars and untold volunteer and staff hours attempting to register new voters and get people to the polls. And voter turnout increased by nearly exactly the same margin ? just under 6.3 percent ? over the 2000 presidential election in the 72 counties that went for President Bush and the 16 counties that went for the Democratic candidate.

The Bush counties tallied a 71.77 percent turnout in 2004, up from 65.5 percent in 2000. The Kerry counties tallied 68.05 percent in 2004, up from 61.8 percent in 2000.

The turnout calculations for 2004 will change some once boards of elections finish counting the 155,428 provisional ballots, which were cast by people who, for one reason or another, weren't on the voting rolls at their precinct. Also still being counted are out-of-country absentee ballots. Final results will be available Dec. 3.

If all of those provisional ballots are found to be valid ? something officials do not expect ? turnout could increase by 1.8 percentage points in the Bush counties and 2.1 percentage points in the Kerry counties.

The rate of provisional ballots was slightly higher in Kerry's counties. A total of 85,228 provisional ballots, or 3 percent of all ballots, were cast in those 16 counties, while a little more than 70,000, or 2.5 percent, were cast in Bush's 72 counties.

Fast-growing, mostly suburban, counties had the lowest percentage of provisional ballots (2.3 percent) and under-counted presidential ballots (1.3 percent), while urban counties had the highest rate of provisional ballots (3.1 percent) and rural counties had the highest percentage of under-counted ballots (1.8 percent).

Denny White, chairman of the Ohio Democratic Party, believes Kerry was hurt by the long lines at some polling places.

"It's just too bad that thousands of people couldn't vote on Election Day," White said. "Good luck explaining to your Republican boss that you're going to stick around and vote for John Kerry instead of coming to work."

The Daily News analysis found that turnout was almost 5 percentage points higher in the 10 fast-growing, mostly suburban counties that are Republican strongholds than in the six mostly urban counties that include the state's biggest cities. The "exurban" counties ? such as Warren, Butler and Greene ? recorded a total turnout of 72.7 percent, while the urban counties, such as Montgomery, Hamilton and Franklin ? saw a 67.8 percent turnout. The 72 rural counties recorded a voter turnout 71 percent.

Traditionally, urban areas have lower turnout that suburban or exurban areas, Sabato said. That helped Bush.

"It's somewhat easier to get voters with better incomes and educations out to the polls, and sure enough, we see that the GOP was probably more successful at this, marginally," Sabato said. "But elections are won at the margins."