ON SATURDAY THE New York Times published an investigation into a "secret U.S. exit poll" in Kenya's 2007 presidential election. In December of that year, Mwai Kibaki defeated Raila Odinga to win a second term as President of Kenya. According to the official results, Kibaki overcame Odinga's pre-election lead in the polls to win by some 200,000 votes, about 2 percent of the total ballots cast. Just three days later, Kibaki was sworn in as violent protests by Odinga supporters spread across the country.
Despite widespread suspicions of fraud, the State Department promptly congratulated Kibaki on his victory, only to retract the statement after the European Union voiced doubts about the integrity of the election. A few days later reports emerged of an unreleased exit poll conducted by the International Republican Institute (IRI) that showed Odinga had actually won the race by 8 points.
The New York Times reports that IRI officials in Washington decided "the poll numbers would be kept secret," and that "at least four people" involved in IRI's Kenya operations disputed the official explanation -- a lack of confidence in the numbers -- for withholding the results. However, only one name is provided: Kenneth Flottman, then IRI's director for East Africa. Earlier this month, the Nation magazine likewise accused IRI and the State Department of suppressing the results of the poll because, as one unnamed source described as a "non-American international official" told the magazine, the results were "unpalatable." Again, the only on-the-record source alleging impropriety was Flottman.
The Times has no actual evidence that IRI withheld the results for reasons other than those stated, but according to the paper, IRI's decision "was consistent with other American actions that seemed focused on preserving stability in Kenya, rather than determining the actual winner." This apparent consistency, along with the allegations from Flottman, served as the basis for the third New York Times investigation into IRI's activities in just the last three years. However, this time IRI refused to cooperate with the paper. Officials there have become convinced that the New York Times is incapable of objectivity in its coverage of IRI. Rather than respond directly to the New York Times, IRI contacted THE WEEKLY STANDARD to offer its version of events.
ALTHOUGH IT IS SUBSIDIZED by federal tax dollars, staffed largely by Republicans, and boasted close ties to the Bush administration, IRI is an ostensibly nonpartisan, non-governmental organization that works to spread democracy and build democratic institutions in more than 100 countries. In practice, IRI shares a worldview with the Republican party that places it squarely at odds with left-wing partisans and media.
Congress also funds a Democratic counterpart to IRI, the National Democratic Institute. In countries where both organizations have a presence, they cooperate by dividing the work -- candidate development, polling, election monitoring, etc. -- among themselves. Elsewhere the two will divide a region by country in order to maximize their reach, with IRI working in Mongolia, for example, while NDI establishes a presence in Nepal. "The best we can do in a foreign country is walk into a room together," says Lorne Craner, the current President of IRI. The goal is to help emerging democracies create a system of government where there is a "competition of ideas, and nobody gets shot and nobody goes to jail when one side loses" he said.
While NDI and IRI both operate in Kenya, IRI had been working with and training opposition figures there for years. The organization had a particularly close relationship with Odinga, who had run in 2007 as a democratic reformer with a strong base of support among Kenya's urban poor. As the results of IRI's exit poll leaked out, journalists began to ask why the group had failed to release the numbers before the election was called -- at the very least, the numbers would have complicated Kibaki's efforts to claim victory. Slate ran a story on January 2, 2008, saying that the poll "could have given the [Electoral Commission of Kenya] pause before it called the election so disastrously," and by January 14, the poll's unofficial result -- an 8 point Odinga win -- had been reported by McClatchy.
According to officials at IRI, there is a much simpler explanation for the decision to withhold the results of their exit poll: the numbers weren't reliable. The call was made, in consultation with IRI officials, by IRI board member Connie Newman, who served as Assistant Secretary of State for Africa under Colin Powell and led IRI's election observation delegation in Kenya. "We determined it wasn't valid," Newman said, explaining that violence and civil disturbances in the immediate aftermath of the election had prevented completion of the poll. Even where violence and unrest weren't a problem, the Kenyan firm IRI was working with "just had problems getting their stuff in," Newman said.
The Times asserts that the "quality [of the data] was not expected to be a concern," but it was. And while Flottman told the paper that, of all those involved, Newman was "most opposed to releasing the numbers," IRI provided THE WEEKLY STANDARD with copies of Flottman's emails which indicate that he, too, had concerns about the quality of the data. According to one email dated January 25, 2008, Flottman wrote his superiors in Washington that "I am inclined to think that recoding the whole thing from scratch here would be the way to go." Further, Newman insists she didn't even know the results of the poll when she decided to hold it for further review because IRI's guiding principle, as Craner said, is "the first bad poll we put out is the last poll we put out."
It took nearly eight months for IRI to release the results of their poll, and they did so only after all the data had been reentered and audited by two independent firms. The new result put Odinga's win at 6 points, meaning that the initial result (an 8 point margin) was outside the poll's 1.35 percent margin of error. As IRI Press Secretary Lisa Gates explained, "if you're outside the margin of error, you know you have some problems." But IRI's critics drew a different conclusion. Writing in the Nation, Karen Rothmyer asserted that while the review had reduced Odinga's margin, 6 points was "still significantly larger than the poll's margin of error." Of course, that's true, but the final numbers would seem to confirm that the initial results were not reliable, and that IRI had acted responsibly in postponing the poll's release. Likewise, the Times notes that the initial results were off by two percentage points but makes no note of the fact that this is well outside the poll's margin of error.
In other words, there was no conspiracy -- at least not involving IRI.
IRI'S CLOSE TIES TO Washington's Republican establishment have made it a frequent target for left-wing media like the Nation and the New York Times. The organization is chaired by John McCain, whose role there was the focus of an investigation by the New York Times in the summer of 2008 after McCain had locked up the Republican nomination. The paper accused McCain of using his perch there "to score points with important Republican figures" and rub shoulders with wealthy donors. Before that, in 2006, the Times published an investigation alleging that IRI had worked to undermine U.S. policy in Haiti -- a charge repeated in this latest piece about the Kenyan exit poll, offered as evidence of a pattern of "meddling." That 2006 story quoted the former U.S. Ambassador to Haiti, Brian Dean Curran, accusing IRI of "attempting to destabilize the government" of President Aristide, whose rule had turned increasingly authoritarian before he fled the country in 2004.
In order to understand why officials at IRI are so convinced that the Times's investigations into their work have been conducted in bad faith, one needs to understand the fascinating back story of the 2006 Times report on IRI's role in Haiti. According to one former Bush administration official, the Haiti piece was at least partly the product of some kind of personal vendetta against the group by Janice O'Connell, a former staffer for Senator Christopher Dodd and the wife of Jeff Gerth, an investigative reporter who worked at the New York Times for more than 25 years, leaving the paper shortly before the Haiti piece went to print. In meetings with IRI staff in 2003 and 2004, O'Connell repeatedly demanded that IRI remove its chief officer in Haiti, Stanley Lucas.
According to Roger Noriega, who served as Assistant Secretary of State for Western Hemisphere Affairs from 2003 to 2005, Lucas had become "a personal obsession" for Ambassador Curran, a holdover from the Clinton administration, and Janice O'Connell "bought into his paranoia." During Noriega's confirmation hearings in the spring of 2003, Senator Dodd pursued only one sustained line of questioning and it focused entirely on Stanley Lucas and Ambassador Curran's claim that he had "caused some difficulties." Dodd asked Noriega if he had worked with Lucas, if he knew Lucas, if anything had "been done about this particular problem." Dodd was so well briefed on this particular personnel issue that after threatening to hand the matter to the inspector general, he said that he'd been "told by staff here that the grant renewal specifically required Mr. Lucas not to be involved in the program."
Curran was "more or less committed to the Clinton policy that we could work with Aristide," says Otto Reich, who preceded Noriega as Assistant Secretary of State for Latin American Affairs. According to Reich, "after spending between $2 and $3 billion propping [Aristide] up and sending the 101st Airborne, they were upset when the Bush administration decided finally that we weren't going to be complicit with a government that was violating human rights, ordering people killed, stealing money -- [Aristide] was no friend of the United States."
Noriega says "Curran had it out for Lucas, thinking Lucas was undermining his credibility, but Curran had very little credibility because of his coddling of Aristide." Curran found a boogeyman in Lucas and IRI, Noriega says, and the result was a "fantastical New York Times account of Lucas as insider and conspirator."
O'Connell was a powerful enemy, able to push her case against IRI, and what she perceived as its unofficial campaign to overthrow Aristide, in the Senate and on the pages of the New York Times. O'Connell was also, according to Reich, "a notorious leaker." Reich said he'd even put his complaint in writing. "I wrote a letter pointing her out as leaker of a briefing I gave -- the contents appeared in the New York Times the next morning. She always leaked to the New York Times," Reich said.
Reich can't conceal his disdain for O'Connell. He concedes he's not objective, calling the former Dodd staffer "despicable," and saying "anytime you have somebody undermining U.S. policy, her name pops up." She was an "attack dog for the extreme left," Reich said, adding that "whenever you find an anti-American group in the last 20 years, you find O'Connell taking up their side, including violent anti-American groups like the FLN, the Sandinistas, Cuba."
O'Connell was also a supporter of Aristide, but IRI had been tasked to work exclusively with opposition parties in Haiti -- a fact never noted in the Times story, which insinuates a political motive to IRI's lack of cooperation with Aristide -- while the National Democratic Institute advised and trained Aristide's governing party. Still, Reich insists that Curran's charge that IRI was undermining U.S. policy was "simply not the case." Reich said of the Times reporters, "they tried to build a case where there wasn't one." Though Colin Powell's views of U.S. policy in Haiti were characterized by the Times as being in line with those of Ambassador Curran, Reich characterized Powell's views of the reporting by the Times, saying Powell believed "the level of professionalism was appallingly low."
O'Connell quietly left her job in the Senate a year ago, just as her patron, Senator Dodd, was in the midst of a presidential campaign and at the peak of his power. Several sources speculated that there may have been one leak too many, but no misconduct was ever officially reported. Ambassador Curran left Haiti under similarly murky conditions which the Times could not pin down, though one executive official with knowledge of the matter told THE WEEKLY STANDARD that given the circumstances of his departure it was "breathtaking that he would claim the moral high-ground."
THE TIMES CLOSES ITS latest piece on IRI describing the rift between the Institute and Odinga in the wake of the 2007 elections. For Odinga, "bitterness lingers," the paper claims. But according to Connie Newman, the relationship has not suffered. Newman met with Odinga during a recent trip to Washington, and says that IRI's new chief in Kenya has a "good relationship" with Odinga's party there. Odinga does contend that his supporters believe the release of IRI's poll "would have made a huge difference," but had IRI released a poll it had little or no confidence in for the sake of propping up their preferred candidate, that most certainly would have been construed as meddling. As IRI President Lorne Craner said, "we're in a very idealistic business, the real trick is not having the passion, but the objectivity." That's a trick the New York Times just can't seem to learn.
Michael Goldfarb is online editor of THE WEEKLY STANDARD.