Over the years carrying my torch as a witness to what happened on my watch in democracy assistance in Kenya in 2007-08, I have always tried to be mindful of the notion that it has not been my business who Kenyan voters chose, including how they voted in the subsequent 2013, 2017 and 2022 elections in which Raila has continued to be a leading candidate. Rather, my job in 2007-08, and my purpose since, has been to address the facts honestly and support the democratic process so that the choices actually made by Kenyan voters themselves would be honored.

Thus, keeping the 2007 Exit Poll from meeting an untimely demise because it was diplomatically inconvenient was not a matter of “supporting” Raila versus Kibaki as a candidate or politician, but rather doing my job to support the democratic process and “observe” the election with integrity.

There was a little bit more involved in preserving the hope that the Exit Poll would be released and published during the early months of 2008 when I was finishing out my “public service leave” as International Republican Institute Resident Director for East Africa than I have written about over the years.  It is probably time to tell the story.

In summary, after the decision was made in Washington to my surprise and disappointment not to release the Exit Poll showing Raila winning by almost six points, there was still the notion that the original polling forms would be sent to Washington and the original data evaluated and re-entered in digital form to determine whether there were actual doubts or anomalies to justify the announcement that the poll was “invalid”. Initially, this was going to happen when staff from our Nairobi office traveled to Washington in March for IRI’s annual global meeting.  The meeting was intended to be mandatory for me as a Country Director and I was asked on behalf of IRI’s President to prepare a presentation on the process of dealing with the Exit Poll and the release decision.  I explained to my boss, the Africa Director, that this was a terrible idea since I emphatically objected to the decision to say the poll was “invalid” and not release it, but I did not want to get up in front of a bunch of young idealistic IRI employees working around the world and say that, nor surely did “the front office” want me to. I also had a major family conflict for the meeting which had been moved because someone in Washington had forgotten to make hotel reservations.  Since my leave from my job in the States was up June 1 and I had to move back in May anyway, I was comfortable declining and was able to beg off.

The original survey forms, which were in locked storage at the Country Director residence near our office, were going to be delivered to Washington by the other staff members making the trip for the March meeting.  But then those instructions were cancelled and there was no operative plan to re-enter the data or otherwise review the original forms in Washington or elsewhere.  The researchers from the University of California San Diego (UCSD) who were the critical consultants for the USAID-funded poll and had contributed additional funding supplementing that provided by USAID, wanted to do the data work, but IRI Washington did not want to let them without modifying their contract.  IRI would pay  $10,000 as compensation for the additional work, the same sum as the funding Dr. Clark Gibson of UCSD had provided pre-election, but UCSD would have to surrender the right to publish the results after a six month exclusive period for IRI that was provided in the original pre-election contract.  Dr. Gibson, as he told the New York Times declined because he thought “they were trying to shut me up”.

Given the fact that there was no path forward to complete the polling program and answer the questions that had been raised back in Washington without reference to the original data, I had to make a choice as Chief of Party for the polling program between honoring the existing contract with Dr. Gibson of UCSD or breaching it to follow instructions from my IRI superiors.  I elected to honor the program and the contract (and the election process itself as I saw it) and allowed the UCSD graduate student researcher to take possession of the survey forms (I had sent him, along with my wife, to retrieve them from the polling firm and bring them for safekeeping at the residence once things got “hot” when IRI announced from Washington that the poll was “invalid” and would not be released.  (As but one example of what I was concerned about, the possibility of a re-count of the underlying vote in the December 27 election had been eliminated, allegedly, by a fire in the warehouse where the ECK stored the ballot boxes just after the vote.)

Thus UCSD was able to verify the poll and release the results in presentations in Washington at the Center for Strategic and International Studies and Johns Hopkins University in July 2008 after the end of the six month embargo. And USAID reported in their Frontline newletter that the exit poll “disclosed that the wrong candidate had been declared the winner”.

The personal drama was that in April when I was working from the residence (my successor was in place running the office by then but had not yet been approved by USAID as Chief of Party so I was still needed for public meetings and reports and such) when I got a call from my Africa Director in Washington that my successor had not been able to find the survey forms in the office.  I explained that they had never been taken to the office, which seemed obviously less secure, so we had taken them to the residence.  That was a satisfactory answer and nothing further was said.  The fact that the forms were in San Diego at that moment was a “didn’t ask, didn’t tell”.

In May I was to turn over the residence to my successor.  The UCSD researcher was bringing the survey forms back from San Diego with him and put the boxes in checked baggage which was tied up in a big delay at Heathrow in London, so he arrived without them as I was getting ready to vacate the residence.  In the context of the tension between myself and the Ambassador and the non-release of the Exit Poll, I threw myself my own going away party with my family and the staff that reported to me, but I did get invited to a farewell dinner by the Serbian Ambassador and his wife who managed our IRI office for both the East Africa programs and Sudan. Starting out with a homemade Serbian aperitif I felt a bit woozy after a sip and excused myself. I woke up a few minutes later on the floor of the restroom with a bit of blood on the tile from striking the sink on the way down.

A cab was called to take me to Aga Khan hospital where I recovered for a few days while my wife and kids scrambled to finish getting everything out of the residence for turnover in my absence and I hoped that the boxes of survey forms would arrive in time to be back in the residence for my successor.  After a few anxious days the boxes arrived in the nick of time and I was soon out of the hospital and off with the family for a couple of weeks in Uganda before going back to Mississippi and my job as a lawyer in the defense industry. Testing at the hospital indicated that I did not have malaria, just some similar but completely temporary symptoms of who-knows-what.

At some point, IRI ended up hiring a survey firm in Oklahoma to review the Exit Poll and released it themselves in August 2008 just before the UCSD researchers testified about it to the Kreigler Commission which was conceptually charged with investigating the dispute as to the facts of the vote for president.  Raila wrote about how important the Exit Poll was to him in his autobiography, “The Flame of Freedom”.  He got part of the story wrong, but since he has continued to be a candidate for president over the succeeding elections, it has been in his interest not to be overly fastidious about all the details, just as the important thing for current democracy assistance efforts is keep learning and adapting from the lessons that become available.

If Raila ends up being president this time, I hope he does a great job in the spirit that his most loyal friends and supporters, some of whom are also my friends, have always hoped.  I also hope it is because he gets the most votes in a free and fair election that is not marred by violence or more corruption than we have already seen.