Was the New Hampshire vote stolen?

The Web is abuzz with allegations of fraud, and Dennis Kucinich is asking for a recount. The charges don't hold water, but this problem is not going away.

Published January 11, 2008 10:52PM (EST)

Reuters/Jessica Rinaldi

Dixville Notch residents wait to cast their votes in the New Hampshire primary at exactly midnight in the "Ballot Room" of the Balsams Hotel in Dixville Notch on Jan. 8.

Was the election stolen?

In the years I've been at Salon, I've begun more than a few articles with that question. Elections are never placid affairs, but in recent years several factors -- 1) crazily hackable voting machines, 2) generally heightened partisanship, 3) very close races, and 4) a real, honest-to-goodness purloined race (see Bush v. Gore) -- have raised the paranoid in all of us. Wondering if any election outcome is honest has become a standard post-election emotion; not wondering, now that's just crazy.

So it came as no surprise that within a few minutes of Hillary Clinton's unexpected win over Barack Obama in New Hampshire on Tuesday people began to ask, Can we trust this? New Hampshire has a history of clean elections (phone jamming notwithstanding), and the state does not use the notorious, paperless touch-screen voting machines that, for more than half a decade now, everyone who knows anything about voting or computers has been saying we can't trust.

But still, people were suspicious. Most New Hampshire voters cast their votes on optical-scan systems. This means they marked their choices on paper, then fed the paper through a card-reader that translated the ballot into digits; those digits are then counted by a computer at each county's election office. In New Hampshire, the machines that read the ballots and the computers that count the ballots and report the results are made by a company notorious for shoddy practices: Diebold.

Indeed, the Diebold systems used in New Hampshire -- the Accuvote card reader and the county-based counting computer, which runs Diebold software known as GEMS -- has been proven to be vulnerable to hackers. In 2003, Jim March, a tech consultant in Sacramento, Calif., demonstrated to me how he could easily change the votes in any election and then erase evidence that he'd done so. Anyone with access to the county GEMS computer could do this.

You can also change the votes on the Accuvote card reader if you can get hold of its memory card before the race; Harri Hursti, among others, has shown how to do this. (Watch this clip from the HBO documentary "Hacking Democracy" for a good scare.)

Given these vulnerabilities, election-reform activists immediately began examining the results from New Hampshire, and over the past few days they've claimed to find evidence suggesting fraud. In particular, they say they've found significantly different results between the ballots counted by machine and the ballots counted by hand.

I have, as some readers know, been very critical of activists who've confidently asserted that past elections have been stolen -- notably the 2004 race between Bush and John Kerry. (My position: The overwhelming evidence, particularly political scientist Walter Mebane's thorough study, shows that Bush really did win Ohio; if you disagree with me, that's OK.)

I confess I've not been so critical this time. As I'll outline below, the evidence that Clinton won New Hampshire due to fraud is minimal.

But people want certainty, and why shouldn't they get it? When I got Rep. Dennis Kucinich's press release last night announcing that he would file for a recount of New Hampshire's vote, I was pleased.

Under New Hampshire law, the candidate who requests a recount must pay for it; Andy Juniewicz, Kucinich's spokesman, told me that the campaign was willing to do so. Recounts strengthen democracy, and Kucinich should be praised for putting his money where his mouth is.

With that said, let's dive into the details of the New Hampshire vote, FAQ style.

Is there evidence that the vote was stolen?

Eighty percent of New Hampshire's votes are counted by computers, while the other 20 percent -- mostly votes in smaller or more rural areas -- are counted manually. Activists looking into the possibility of theft thus took a straightforward investigative approach: Did the results in hand-count areas match up with the results in machine-count areas? If the results do match, you can pretty much definitively say that the election was not stolen.

But the counts don't match. A number of amateur analysts have pored over the results, each using slightly varying methods, and they've all found some difference in the results from hand-count areas versus those from places where votes were counted by machine.

The most thorough analysis I've seen was performed by an anonymous supporter of Ron Paul. Of the votes that have been counted so far, Hillary Clinton beat Barack Obama in New Hampshire by 39.03 percent to 36.39 percent. The Ron Paulster's analysis shows that in machine-count areas, Clinton beat Obama by a better margin, 40.12 percent versus 35.76 percent. But in hand-counted areas, Obama beat Clinton by 38.76 percent to 34.70 percent.

In other words, if Obama had received the same margin across the state as he got in the areas where votes were counted manually, he would have won.

To some activists, this suggests fraud; the thinking is that Hillary Clinton's margin in machine-count areas is the product of hacking, while Obama's hand-count win represents the true result.

Thankfully, few activists are saying that what they've seen proves fraud -- they're being far more cautious, asking only that someone should look into this phenomenon.

That's because everyone understands that there is a reasonable reason for why Obama would win the hand-count areas while Clinton would win the machine-count areas: Those places simply vote differently.

American elections are hyper-local affairs. Election officials choose voting technologies based on their needs, and their calculation depends on demographics.

Officials in charge of small counties are more likely to choose to manually count their ballots. But if you've got 10,000 or 20,000 voters in your county -- like in Manchester or Concord -- you'll use machines. Money is also a factor; poorer places are less likely to have the resources for machines. Governmental efficiency might also matter -- some elections officials may not have gotten around, yet, to adopting machines -- as might local infrastructure, or any number of other factors.

But, of course, the same demographics would also affect voting results. It's likely, for example, that people in small places or poor places would vote very differently from people in large places or rich places -- and, therefore, variances in the result that look like they were caused by voting-machine fraud might actually only be the product of normal regional differences.

Indeed, there's plenty of evidence showing that Obama did well in hand-count areas because those places were Obama strongholds. Consult the second table on this page, the one that shows hand-count vs. machine-count results in counties with fewer than 750 votes.

There, you see Obama got a blowout in hand-count areas -- 39.59 percent to Clinton's 33.64 percent. Clinton did better than Obama in machine-count places, but her margin is smaller than in other places in the state: just 37.37 percent to Obama's 35.04 percent.

Even more interesting is the third table on this page, which shows how the candidates did in places with more than 1,500 votes. Here, Clinton, not Obama, did better in hand-count areas -- a lot better, 44.17 percent to Obama's 31.61 percent. Meanwhile, in similarly large counties where machines counted the vote, Clinton's margin was smaller -- 40.28 percent to Obama's 35.96 percent.

Let me say that again: In large areas, Clinton did better in places where votes were counted by hand than where votes were counted by machine. Does that mean that it was Clinton who was robbed, in these places, by voting machines?

Almost certainly not. All of these variances are much more plausibly the product of differing demographics. If you check out the county-by-county results, it's possible to find many counties where Obama won handily even despite Diebold (Amherst, for instance, which uses optical-scan voting and where where he got 44.68 percent; if the Diebold machines were hacked for Clinton, they didn't do much good there).

At the same time, Clinton won in many areas that manually count their votes -- she got 44.44 percent in Boscawen, 43.93 percent in Carroll, 43.52 percent in Charleston, and on and on.

Can exit polls tell us whether fraud occurred?

Regular readers might know that I've long been skeptical of efforts to use exit polls -- surveys of voters as they leave voting booths -- as a forensic tool to detect fraud. Academic election experts and pollsters (including the surveyors who carry out the exit polls) say that exit polls, at least as they're practiced in America, are not precise enough to catch fraud in close races.

Many election-reform activists disagree, though, and they've been calling on the news networks to release what they call "raw" exit polling data. The exit information that is currently posted on media sites -- like here at CNN, for example -- have been weighted to match the final voting results. This is a standard practice, and is not at all nefarious; surveyors re-weight the exit poll results to match the election outcome in order to provide a better look at why people voted the way they did.

In the absence of "raw" polling data, activists have pointed to a quote from Chris Matthews, the Clinton-bashing MSNBC anchor, that suggests that Obama was far ahead of Clinton in the exit poll data that news networks were getting throughout the day on Tuesday.

In a video that you can watch here, Matthews says, "Even our own exit polls, taken as people came out of voting, showed him ahead. So what's going on here?"

But information I've been able to find about the intra-Election Day exits suggests that Matthews is letting his Clinton hatred get ahead of the facts.

Daniel Merkle, who heads ABC News' "decision desk" -- which was getting the exact same exit polling data that folks at NBC were getting -- told me that the numbers he was receiving during Election Day did not show a certain Obama win. Merkle said the data indicated "a very close race on the Democratic side," and "that's what it ended up being."

"It was within a couple points," Merkle said. "When we're seeing an exit poll within a couple points, that's a close race." The exit poll numbers, he added, were a "surprise" compared to pre-election polls. "The exit poll was not showing an 8- to 10-point Obama lead. It was showing a close race."

But hey, isn't it bad that we don't know, for certain, that this election was fair?

So far in this piece, I've been careful not to sound too certain. There are many reasons to trust the New Hampshire results. Besides the ones I've outlined here -- reasonable explanations for the machine vs. hand-count variance, an exit poll that showed a close race -- there's also the plain fact that, despite Diebold's hackability, it would have been hard to have rigged the vote all over New Hampshire.

Elections are greatly decentralized. Across the state on voting machines, Hillary Clinton won by an average of 40.12 percent of the vote, far more than she was projected to win. Her surrogates would have needed to get into a lot of card readers and GEMS machines in order to achieve that count through hacking.

They would have needed helpful confederates in many areas of the state, and those confederates would have had to have been disciplined enough to escape the notice of the many campaign officials and get-out-the-vote volunteers that Clinton's rivals had installed all over.

There is a label for such a far-flung scenario: a conspiracy theory.

Wise people on the left are cautioning their fellows against going down this path. "There is ... something perverse about the quick knee-jerk reaction to assume that any election that dramatically doesn't go your way was stolen," Josh Marshall, in a post titled "Enough," wrote yesterday. He added: "There is a sullen childishness at work in this thinking that no robust political movement can ever be built on."

I don't disagree with Marshall; indeed I've just written a whole book about how digital technology is helping a lot of us invest in convenient theories over inconvenient facts. People who confidently assert that the 2004 race was stolen make an appearance.

Still, while we can be pretty sure that Clinton didn't steal the race, we don't know -- if you were to be called into a court of law and asked to prove that she didn't steal it, you wouldn't be able to.

And many activists want courtroom-level proof. And they deserve it -- or something close. We live in an era in which partisans on each side believe that the other side is capable of out-and-out theft. Proving that theft didn't occur should be a routine part of elections -- all elections, all the time.

Last night I had a long discussion with Brad Friedman, who runs the election-reform news Web site Brad Blog. Over and over, he said, "My biggest concern here is that 80 percent of the vote is uncounted by any human being." His request is simple and straightforward: "Why not count the damn votes?"

He's right. Why not count the votes?

And thanks to Kucinich, that's what will likely happen now. It will probably take some time; weeks, if not months. But soon, we'll know what happened.

But as many voting-reform experts have argued, manually counting the votes should be a routine in any race. There are logistical reasons why it would be impractical to hand count every vote in every election. But if we're going to use machines -- optical-scan machines that use paper ballots, that is; touch-screen machines everywhere ought to be burned -- we should, at least, conduct a randomized, accountant-approved audit of ballots.

In other words, after every election, officials should randomly count some number of ballots to double-check the machines' results. It is amazing that this is not a standard procedure across the country; it is a disgrace that election officials aren't rushing to implement such procedures now.

Would this make elections fair? God no. Don't kid yourself for a minute that the way we elect presidents in America is fair. For Pete's sake, in most caucusing states people don't even cast their choice in secret, and in November we will elect a president according to a convoluted process that will give each South Dakotan more say in the matter than each Californian.

Fixing the voting machines won't make this election fair. But it's as good a place as any to start.

By Farhad Manjoo

Farhad Manjoo is a Salon staff writer and the author of True Enough: Learning to Live in a Post-Fact Society.

MORE FROM Farhad Manjoo

Related Topics ------------------------------------------

2008 Elections