I am only aware of one specifically articulated explanation of the “much [that] can happen between the casting of votes and the final tabulation of ballots” as Ambassador Ranneberger described it to Washington on January 2, 2008. Here it is, from my March 2009 submission to the Inspectors General of the State Department and USAID:
Subsequent to the election, I met privately with a highly placed diplomatic official who told me that the theft of the election by the incumbent administration had been carried out through bribery of Kenyan election officials in the field, in particular the Returning Officers at the constituency level. The source said that these officials received large payments which left them financially secure in return for turning off their cell phones and otherwise making themselves unavailable to allow the vote numbers in the presidential race to be inflated. The source stated that the government he worked for was unable to identify this method of rigging in time to do anything about it because it was carried out “at the last minute”, very shortly before the voting. [Months later a story was published in the Standard regarding the vote fraud which stated that the original plan had been for Kibaki’s re-election to be assured by declaring Langata Constituency for Livondo over Odinga, but that as it became clear that the ODM ticket was carrying large margins from Western and Rift Valley Provinces it was decided that this was not tenable and the approach was switched to inflating the votes from elsewhere.]
This discussion took place in January 2008, during the post election violence, with the exit poll issue “pending”. I found it credible and believed it then, as I do now. Nothing in any of the less fact specific analysis produced by diplomatic or social science sources that I have seen over the years is inconsistent or suggests a contradiction with this information. The Kriegler Commission elected explicitly to stay well away from the type of investigation that would have confronted the Commission with the existence of such facts. I promptly reported the conversation to IRI Washington as I consistently reported such conversations during the election campaign and immediate crisis.
FOIA Update: I timed this series based on information from the USAID FOIA office that I would be getting the complete response to my April 2013 request to them for the documents relating to the exit poll by October 17. They were kind enough to call and let me know that it would be delayed to last week and after checking back they sent me a lengthy heavily lawyered letter and some documents. We have broad areas of disagreement at this point and I have asked them to reconsider their approach in some respects. Pending that, I did finally establish by virtue of the letter from USAID that IRI never filed the final report on the 2005-2007 USAID Kenya polling program, covering the 2005 and 2007 exit polls. Likewise, I have an officially public copy of the IRI January 14, 2008 quarterly report where IRI reported to USAID that the poll had been successfully conducted in spite of the challenges presented.