In Nairobi I had remained optimistic through the end of January 2008 that IRI would ultimately decide to release the results of the exit poll.  I had been told by my immediate superior in Washington repeatedly that the then-president of IRI would be the final decision maker and “wanted” to release the results.  I told people in Nairobi including senior politicians when the poll came up that the issue remained under consideration.

After I was told there was to be a final meeting in Washington to make the decision on Friday, February 1, I scheduled lunch for the next Tuesday with Shashank Bengali, East Africa correspondent for the McClatchy papers (now with the Los Angeles Times) who had broken a story on the results previously, with the expectation that I would then finally be able to disclose them. I prepared a memo at the request of my boss explaining details of the poll results and their quality, concluding with my recommendation that they should be made public.  I was told to remove the recommendation from the end of the final document. Then I was told following that February 1 meeting that the decision had gone the other way and we would not be releasing, so I had to report that to Bengali at our lunch instead.  I obtained an agreement from my boss that I would be given the opportunity to “fact check” whatever statement would be released from Washington and advance notice to prepare the local IRI staff for the consequences in Nairobi.

In my “fact checking” role I objected vigorously to the draft statement prepared by IRI’s PR shop asserting that the results had been found to be “invalid”.  Recognizing that the IRI Washington hierarchy had the right to decide whether IRI would release the numbers, I was offended by characterizing the poll in a factually unjustified way and misleading the public to make an excuse. This seemed on par with the decision to tell Ambassador Mark Bellamy that he was dropped from the election observation because we could not buy his plane ticket when in fact it was because Ambassador Ranneberger created a showdown over the assertion that Bellamy was perceived to be unfavorable to the incumbent government–except that being less than truthful about the exit poll was a much more consequential matter.

In discussing the matter with my boss and batting away other notions to justify not releasing the results I was told that the bottom line was they were not being released because to do so was “not in the best interests of IRI”.  Those interests were not explained. Likewise it was said that the decision had been “going that way” since mid-January. On the morning of February 8, Nairobi time, I got a call from a staff member in Washington who told me that the offending draft had gone up on the IRI web site overnight, in violation of my agreement, following the events in Washington that day, February 7 (the Senate hearings, which IRI had not attended, where Assistant Secretary of State Frazer and the Assistant USAID Administrator were grilled on why the exit poll had not been released, and a visit to IRI headquarters from ODM representatives to press for release of the poll.  Also that day ODM met with Deputy Secretary of State Negroponte to press unsuccessfully for release of the poll.)

Our longtime pollster, the late Dr. Peter Oriare, who was predictably devastated by the statement from Washington, defended the poll in a story in The Star which quoted both my boss in Washington and me, and was published in print (no website yet) the morning of Friday, February 15.  I had declined to tell The Star that the Washington statement “reflected my personal opinion”.  I left later that day for my scheduled and overdue trip to Somaliland and after I left the Washington office dispatched Nairobi staff from our Sudan and Somaliland programs to personally deliver copies of the February 7 “invalid” statement to key political negotiators on both sides, including Raila Odinga, Musalia Mudavadi and Martha Karua–to which I objected from afar.

These dispatched IRI employees had not been involved in the exit poll and would have had no way to know the statement from Washington was inaccurate. It was particularly noteworthy that the delegation, as I learned later, had missed the new Justice Minister Martha Karua in her office and chased her down at the Serena hotel where the Kofi Annan-led mediation was taking place.

Hon. Karua was leading the PNU mediation team and was identified at the time as one of those key Kibaki hardliners who opposed compromise.  She had her own stake as newly elevated to a key cabinet post by Kibaki on January 8 and as the leader of the NARC-Kenya party.  She has said subsequently that she in fact believed at the time that Kibaki had legitimately won the election–I cannot evaluate retroactively her subjective beliefs, but I was and remain troubled by the fact that she was told in person on February 15 that the exit poll–to be released in July and August showing a large Odinga win–was “invalid” and would not be released because of that lack of validity.  The mediation at the Serena quickly broke down, although Kibaki and Odinga later reached their deal of February 28.

Whatever the machinations of the Ambassador and/or anyone else in the State Department, IRI’s Washington office at that time managed to make the situation more embarrassing and more difficult for the Kenya program aside from failing to have to courage to meaningfully stand against election fraud until too late to really matter.