The key example of this is Manjoo’s flatly inaccurate claim that the Democratic National Committee report identifies only 129,543 voters, or 2 percent of the electorate, who were disenfranchised by the long lines in Ohio. I can only point to the executive summary of the DNC report, which states:
“Scarcity of voting machines caused long lines that deterred many people from voting. Three percent of voters who went to the polls left their polling places and did not return due to the long lines.”
Manjoo seizes on one line in the 204-page report and then attempts to play a clumsy game of gotcha. But if he had read more carefully he would have understood that the 129,543 votes he refers to were only a subset of those disenfranchised by the long lines. Had Manjoo read a mere paragraph further in the report, he would have seen that it identifies a second group, comprising roughly 48,000 citizens, or 0.83 percent of Ohio’s electorate, whose votes were also suppressed because of the lines and other factors.
The authors of the DNC report aggregate these totals to arrive at the 3 percent figure that I cited. Does Manjoo pretend to have a better grasp on the data than the DNC’s own experts? If so, his beef is with them, not me.
Manjoo also raises the issue of voting preference identified in the DNC’s report, insisting “that those votes would have split evenly between Kerry and Bush.” But, again, the DNC only ventures an estimate for the subset of 129,543 voters, not the full 3 percent turned away by long lines, writing of the additional 0.83 percent, “We do not know the voting preferences of these approximately 47,979 voters.”
I chose not to rely on this incomplete picture from the DNC. It was evidently informed by self-reports of voter preference by citizens surveyed months after the victor had already been determined — a notoriously unreliable measure that routinely overstates the performance of the victor. Instead, I cite the lone report on the Ohio scandal that was produced by the federal government, the Conyers Report. It concludes that the majority disenfranchised by the lines were Democrats. That report — along with every news report from Election Day — identifies that the worst lines were in urban areas and on college campuses, Democratic strongholds both.
Contrary to Manjoo’s insinuation, I never suggest that no Republicans were disenfranchised in 2004. Many thousands were doubtless caught up in the massive voter suppression effort. But they were also clearly in the minority. The most severely affected were urban dwellers, in particular black urban dwellers. Ohio’s biggest cities voted for Kerry by margins approaching 5-to-1. Eighty-four percent of black voters voted for Kerry.
In another specious attack, Manjoo questions whether any of the more than 300,000 voters who were purged in advance of the 2004 election actually showed up at the polls. “It’s impossible to know if those were even real people,” he writes.
Farhad Manjoo, meet Barbara George. She was among the tens of thousands disenfranchised in this manner. “My God. We are sixty-six years old,” she told the Toledo Blade at the time. “We registered when we first turned twenty-one. We have lived in this same house for forty-four years, and [now, because of the purge] I can’t vote. It just seems ridiculous that you have to keep re-registering if you don’t vote.”
Mrs. George was among 28,000 voters who were purged from the voter rolls only a few months before the election in Lucas County. Why before and not after the election? That arbitrary decision was made by the Lucas County Board of Elections, a group so ethically challenged, the whole lot of them, Democrats and Republicans alike, were forced to resign following a state inquiry.
It is also arbitrary that the voter rolls were purged in some cities — Toledo, Cleveland, Cincinnati — but not others, notably Columbus. This disparate treatment raises serious equal-protection concerns under the Constitution. Indeed Ohio stands as a case study in how officials, acting under color of law, can deprive citizens of their constitutional rights.
As to estimating the number of these voters, I took a decidedly conservative approach. The experts we interviewed suggested that 10 percent of those purged were likely purged by mistake. That alone accounts for 30,000 voters disenfranchised. The actual total is in all likelihood higher.
Manjoo’s misguided gotcha campaign continues throughout his piece. With regard to his claim that the exit poll numbers were within the margin of error, Manjoo is referring to the “corrected” numbers that Mitofsky retroactively weighted to more closely resemble the certified vote tallies. Freeman and other statisticians, mathematicians and social scientists disturbed by the poll numbers rely instead on Mitofsky’s raw data. Those numbers, as I report, indicated that Kerry had an advantage outside the margin of error in Ohio, Florida, Nevada and New Mexico. He also had advantages within the margin of error in two other states: Iowa and Colorado, which also went to Bush in the final results. It’s hard to engage in an honest debate with someone who won’t debate the same set of numbers.
With regards to the rural counties, Manjoo identifies two Democratic judges who outperformed Gore in the 2000 election. They are both red herrings. First, it is disingenuous to compare Gore’s results in Ohio in 2000 to Kerry’s in 2004. Gore did not contest Ohio. He crossed it off his list of winnable states nearly a month before the election, and stopped campaigning there. Kerry, by contrast, made Ohio the focus of his electoral effort and campaigned intensively in southwestern Ohio, including the final weekend of the campaign.
As for the first judge, Alice Resnick. She was a wildly popular incumbent, who, with bipartisan support, actually earned more votes than Gore statewide — which tells you something about how little effort Gore put into contesting Ohio. The second judge, Tim Black, is a white man, a former Republican, and a legal luminary from Cincinnati. That he should enjoy a boost in support in the surrounding counties of southwest Ohio is not anomalous.
Contrast these figures to Judge Ellen Connally, an African-American woman who strongly supports both gay marriage and abortion rights. Here is the paradigmatic “activist judge” — from the opposite corner of the state — beating the pants off Kerry in rural southwest Ohio, and nowhere else.
The “Connally Anomaly” is not so easily dismissed. In particular because an analysis performed by the National Election Data Archive for Rolling Stone reveals that Connally’s margin of loss to her opponent Thomas Moyer in these counties falls in line with statewide expectations. The anomalous skew, by contrast, is between Kerry and Bush’s numbers.
Finally, Manjoo faults me for aggregating the voters that the Greater Cleveland Voter Registration Coalition determined were definitely disenfranchised and those who were likely disenfranchised. Once again, however, this analysis is not mine, but the Coalition’s. The report clearly states: “The key point is that the sum of these avoidably lost votes or votes put at risk add up to 72,500 votes or about 1.3 percent (range 0.9-1.6 percent) of votes cast in a (2004) Presidential election decided by a difference of 2.1 percent of Ohio’s votes.” Further, the report repeatedly cautions that its estimates are “conservative” and that “our numbers probably understate the problem.”
As does the 357,000 number that I arrived at. It does not account, for instance, for the untold thousands of voters disenfranchised by Kenneth Blackwell’s absurd 80-pound paper-weight decision, the “ballot crawl” in Cuyahoga County that misallocated Kerry votes to obscure third-party candidates, the suppressive effects of the GOP’s Jim Crow-style bully boy tactics or the party’s campaign to disenfranchise 35,000 urban voters. The list, tragically, goes on and on.
Manjoo has made a cottage industry for himself in attempting to debunk concerns about the validity of the 2004 election. Given that he has staked his professional reputation on the thesis that Bush beat Kerry fair and square, it’s unsurprising that he should be eager to attack my piece. But it is a shame that his faith in the election results has blinded him to the point that he can dismiss the widespread and uncontested evidence of vote suppression as nothing more than a “hit parade” of irrelevant facts and figures. He also remains strangely silent on the transparently crooked recount process, which has kept this debate alive by preventing us from knowing the actual outcome of the vote in Ohio.
Manjoo’s outrage and professional energy would be better directed at those who mounted a concerted campaign to obstruct hundreds of thousands of American voters from going to the poll and having their vote counted in 2004. The nation still needs a thorough and honest exploration of what happened across the country, so we can begin the urgent work of instituting real reforms — ensuring that such abuses do not continue to undermine democracy and cast doubt on the integrity of our entire electoral system.
I appreciate Robert F. Kennedy Jr.’s response to my article, and I’d like to first note that I agree with him on one main point — that we should urgently begin the work of honest election reform. We differ on how to go about that effort, however. Kennedy believes that any such reform effort must begin with an examination of whether Republicans stole the 2004 race. I disagree for many reasons, but mainly because the evidence that John Kerry actually won Ohio is so slight that any such effort is, in my view, doomed to failure — and such a failure would damage the entire reform movement.
Kennedy says that I’ve made a “cottage industry” of attempts to debunk the concerns surrounding the 2004 election. But as my reporting history at Salon shows, I’ve been exploring the various threats to honest elections for several years, and I thoroughly covered the threats to the 2004 race. He’s right that I’ve criticized some who’ve been quick to claim that the race was stolen. But this is not because I think elections in America are perfect — in fact, just the opposite is the case.
We’ll only improve the process if we begin by honestly reviewing the facts — and once again, I’ve got to disagree with the ways in which Kennedy interprets some of the key sources he cites to arrive at his conclusions.
The exit polls
With regard to whether the exit polls showed John Kerry to be substantially ahead in key states, Kennedy is simply incorrect in stating that I’m relying on “corrected” exit poll data while he’s looking at the true numbers. I was relying on the very same numbers he cites — researcher Steven Freeman’s exit poll data — to show that John Kerry’s lead was well within the margin of error in Iowa, Nevada, New Mexico and Ohio (the four states in which the exits showed Kerry ahead that he eventually lost).
You can find the exit poll data that Freeman relies on in his first report on the election, located here: In Iowa, this exit polling data showed Kerry ahead by 50 percent to Bush’s 48 percent; in Nevada, Kerry was ahead 49 to 48; in New Mexico, he led 50 to 47; and in Ohio, he was at 52 to 48. As the pollster Mark Blumenthal has pointed out, the margin of error in these states varied from 5 to 7 percentage points. In none of these states does Kerry’s lead even come close to that level.
The long lines
To estimate the number of people who were disenfranchised due to long voting lines in Ohio, both Kennedy and I rely primarily on the Democratic National Committee’s report on the election, available <a target="new" href="http://a9.g.akamai.net/7/9/8082/v001/www.democrats.org/pdfs/ohvrireport/fullreport.pdf “>here. The DNC arrived at its figures through a poll — it hired a research firm to survey 1,201 Ohioans in order to see how they’d fared on Election Day. The poll found that “two percent of voters who went to the polls on Election Day decided to leave their polling locations due to the long lines.” (Page 20 of the PDF file.) The poll also found that among these 2 percent, “potential voters would have divided evenly between George Bush and John Kerry.” The DNC report estimated that this portion of the electorate represented 129,543 votes.
Kennedy is right that in addition, the poll identified another set of voters who “did not go to the polls at all because they did not receive their absentee ballots, or had heard about long lines, registration challenges, and confusing polling sites.” The report adds that “We do not know the voting preferences of these approximately 47,979 voters.”
I think Kennedy makes two errors in interpreting this data. First, he adds up both these two categories to come up with a total of “more than 174,000 voters” who he says “showed up to vote on Election Day [and] were forced to leave without casting a ballot.” Clearly, though, as the DNC report states, this second category of people “did not go to the polls,” and they were concerned about one of many problems surrounding the election — problems including, but not limited to, long lines. Therefore, it’s just not correct to say that these voters went to the polls and were subsequently forced to leave because of long lines. (As Kennedy indicates, there is some internal inconsistency in the report, since the executive summary also adds up these two categories; it’s only when you drill down beyond the summary into the meat of the 204-page report, as I did, that you find the real numbers.)
Kennedy’s other error is to ignore the very same poll’s second finding that voters who were forced to leave precincts due to long lines would have split evenly between Bush and Kerry. Kennedy says that he doesn’t believe this figure because it’s based on voters’ self-reports months after the election. But, as I’ve indicated, the DNC’s entire study on long voter lines is based on self-reporting months after the election (the poll was conducted from Jan. 30 to Feb. 2 in 2005). I can’t see why he’d place his trust in these self-reports when looking at the number of voters he thinks were affected by long lines, but not when looking to see which voters (Kerry voters or Bush voters) were affected. Moreover, even if he had valid reasons for ignoring this second finding, he misled readers by failing to at least point out that the very same report he cited came to a conclusion very different from his own: “Despite the problems on Election Day, there is no evidence from our survey that John Kerry won the state of Ohio.”
The purported rural vote shift
In his Rolling Stone article, Kennedy suggested that it is extremely unusual for candidates seeking lower office to receive more votes than the presidential candidates of their own party, as occurred in 12 rural counties in Ohio in 2004 (there, Supreme Court candidate Ellen Connally outperformed Kerry). Kennedy quoted Rep. Dennis Kucinich to this effect: “Down-ticket candidates shouldn’t outperform presidential candidates like that,” he says. “That just doesn’t happen. The question is: Where did the votes for Kerry go?”
I think that Kennedy ought to have pointed out to readers that Kucinich wasn’t correct; as I reported, down-ticket candidates do sometimes outperform presidential candidates in some places, and in fact this occurred in 2000, when both Tim Black and Alice Resnick received more votes than Al Gore in dozens of counties.
Kennedy now concedes that such things do happen, but he says that for reasons particular to the 2000 race, the Gore example doesn’t count. We’ll have to disagree on this one. Where Kennedy appears to see no possible explanation other than vote fraud for Connally’s performance in these rural counties, I continue to believe that the fact that the Ohio ballot does not list Supreme Court candidates’ party affiliations accounts for some large measure of what Kennedy calls the “Connally anomaly.” It’s quite possible that many voters — including some Bush voters — chose her without knowing anything about her liberal politics.
Dan Tokaji, an Ohio election-law expert whom Kennedy cites numerous times (favorably) in his article, echoed this sentiment on his blog: “Although I’m an Ohio voter, and was following the 2004 election pretty closely, I must confess that I didn’t know [Connally] was ‘gay-friendly’ or black until reading Kennedy’s article,” he <a target="new" href="http://moritzlaw.osu.edu/blogs/tokaji/2006/06/back-to-ohio-rolling-stone-piece.html “>wrote. “These down-ballot contests simply don’t receive a lot of media attention … I did know that Connally was a Democrat, but other voters might not even have known that. Although judicial candidates in Ohio are typically endorsed by the parties, the offices are nominally nonpartisan and, under Ohio law, “[n]o name or designation of any political party” is supposed to appear by judicial candidates’ names on the ballot. So inferring election fraud in 12 counties based on Connally’s vote total is, in my view, quite a stretch.”
The purged voters
Kennedy argues that voting rolls in key left-leaning cities were cleaned up arbitrarily before the election. I pointed out that this was done in accordance with the law. Kennedy says that even if it was lawful, the procedures were arbitrary and partisan.
But even if you give Kennedy the benefit of the doubt here, I just don’t see how you can count these people into any estimate of “missing voters.” According to Kennedy, of the 300,000 names that were purged from the rolls, 10 percent, or 30,000 people, would actually have voted. He says this is a conservative number. Perhaps. But it’s also unsubstantiated; nobody can know whether or for whom those people, if indeed they were real people, would have voted.
The registration errors
Kennedy is still misinterpreting the Greater Cleveland Voter Coalition study, which he said concluded that “72,000 voters were disenfranchised through avoidable registration errors.”
But the report did not conclude that. As he notes, it stated: “The key point is that the sum of these avoidably lost votes or votes put at risk add up to 72,500 votes or about 1.3 percent (range 0.9-1.6 percent) of votes cast in a (2004) Presidential election decided by a difference of 2.1 percent of Ohio’s votes.” As you’ll notice, the report discusses votes that were “lost” and those that were put “at risk” of being lost — and this second at-risk represented 30,000 of the 72,500 votes. Many of these people, as the report states, did indeed get to vote; they were not, as Kennedy argues, all disenfranchised.