Kenya’s presidential election petition – it was clear IEBC did not follow the law, even before Supreme Court Registrar showed serious skulldugery with ICT

Discussing Kenyan elections can get tense, even among friends who are not Kenyans and try to be relatively dispassionately analytical. I have copied here one of my emails from an ongoing exchange in late August during the pendency of the Presidential Petition in the Supreme Court. My friend with whom I was corresponding is a Westerner who knows far more about Kenya (and lots of other relevant things) than I do and is someone I greatly respect (he is also a layman as far the legal profession goes). My friend was much more sanguine than I about the IEBC’s implementation and use of the KIEMS Results Transmission System, both in terms of facts and law. This explains how I saw things (and still do):

Uploading an alleged Form 34A offline after the election and reporting of results reflects a failure of the use of the RTS by its terms as consistently represented by IEBC and IFES until well after the election.

It is simply not the same thing at all in my opinion.

Even ELOGs sample in their PVT found 13.5% of Polling Stations did not publicly post Form 34A. If it wasn’t scanned and transmitted in real time, or at least scanned with delayed transmission upon being moved into a coverage area contemporaneously, and it also wasn’t publicly posted, then it cannot credibly treated as if it was reliable without explanation and evidence.

Your figure of 29,000 and the IEBC tweet claiming all but just over 1000 leaves a huge gap in a very short time period. (Further, I understand you to refer to some “backlog in uploading them” which apparently refers to something other than KIEMS transmission, so I am not sure at all that I am really understanding your argument.)

I also disagree with your characterization of “clear rules” of Kenyan election law implementing the Maina Kiai court decision against the IEBC. IFES advised to the contrary in their last pre-election publication on the process that I am aware of, the July 20 FAQ that also explained how KIEMS was to work.

People may have gambled that Chebukati could use the Court of Appeals ruling to announce on day 3 of 7 “final results” from most but not all alleged Form 34Bs without the 34As having been demonstrably transmitted to the Constituencies to generate the Form 34Bs. This tactic might very well win the Supreme Court of Kenya, legitimately or illegitimately, but I don’t find it persuasive myself, nor do I find that provides any justification for the assertive lack of basic transparency.

Kenyan lawyer Nelson Havi’s piece in The Elephant from about the same time gives a good summary of the issues in the Presidential Petition and the Petitiiners basic case: “KENYA ON TRIAL: Truth, Justice and the Supreme Court.”

Kenya Election: How IEBC CEO explained what was legally required for electronic Results Transmission and how KIEMS was to meet requirement

IEBC’s high-tech system to guard against ballot stuffingThe Standard July 22, 2017

The Independent Electoral and Boundaries Commission (IEBC) has assured that the integrity of the August 8 election has been guaranteed through tamper-proof technology.

The Kenya Integrated Elections Management Systems (KIEMS) has unique features that will make double voting, ballot stuffing, and irreconcilable voting patterns impossible. IEBC is already preparing to deploy 45,000 KIEMS kits to be used in all the 40,883 polling stations across the country.

Every polling station has been allocated a Kit with a maximum number of 700 voters depending on the size of the polling centre. By implication, voters in a polling station cannot exceed the allocated number.

The KIEMS technology has two main functions in this election. The first is biometric identification of voters on the election day and results transmission after counting the votes.

The Commission has made it mandatory for all voters to be identified biometrically to close the doors for possibility of resurrection of dead voters.

IEBC Chief executive officer Ezra Chiloba said the Commission has invested heavily in technology and can guarantee successful transmission of election results.

“We have no choice really. The law already demands of us to electronically transmit presidential results from the polling station to the tallying centres,” he said.

After counting of results, the presiding officers in the presence of party agents are expected to type the total number of votes garnered by each candidate into the kit.

The kit aggregates the results automatically and the total number of votes cast for all the candidates is recorded. In cases where the number of voters exceeds the total number of registered voters, the kits shall automatically reject the results. This measure, according to the Commission, effectively makes ballot stuffing impossible.

As an additional measure to guarantee the integrity of elections results, the presiding officer shall scan Form 34A using the KIEMS kit. The Form 34A is signed by both the presiding officer and party agents. Once scanned, the presiding officer shall, together with the text results, send the same to the national tallying centre and constituency tallying centres.

The kit shall equally report turnout trends periodically throughout the day. With this kind of monitoring, the Commission says, ability to identify abnormal voting patterns is guaranteed.

At the end of the voting, said Mr Chiloba, the presiding officers in the presence of party agents are required to reconcile the number of voters recorded by KIEMS as having voted and the number of ballot papers issued.

“We have two procedures that minimise the risk of ballot stuffing. One, the voter turnout as recorded by KIEMS. Two, the ballot papers reconciliation that happens at the end of voting. The number of ballots papers issued and the records of voter turnout as registered by KIEMS should be able to reconcile,” he said.

The Commission contends that once the presiding officer has pressed the “Submit” button, the results cannot be changed by anyone.

Using an encrypted format, the results shall then be transmitted to the tallying centres through a secure network in real-time. The public will be able to view the results online. Similarly, Media will have a dedicated connection to access real-time results as well.

According to ICT sources within the commission, the KIEMS have a unique in-built audit trail. The in-build audit trail enables the commission to collect all the kits and to retrieve records from the SD cards for any analysis at the end of voting. This in-built accountability tools implies that the process of voting can be subjected to objective scrutiny at any point in time after voting.
.  .  .  .

Clearly, much of what Chiloba and the IEBC described here just over two weeks before the election did not actually happen after the votes were cast and counted on election day.

Why?  Well, much of the explanation likely rests on the new information disclosed in the Registrar’s reports on the Forms 34A and Forms 34B and the IEBC ICT review in the Supreme Court litigation.  Other things were going on within the process than described.

Kenya’s Presidential Election in a nutshell:  1) widespread failure or non-use of purchased electonic Results Transmission System (as in 2007 and 2013); 2) lack of transparent or complete “complementary” substitute (as in 2007 and 2013)

The voting and counting, as I have previously noted, is the same this year as in the past.  The voter register remained messy again with likely more than one million dead voters and plenty of ineligibles, and was not fixed and locked down as required.  From outside appearances so far, however, the EVID system seems to have substantially worked this time which may have been a big improvement from 2007 and 2013 in limiting in person voting by ineligibles.

The RTS system which was to transmit from a unique registered and logged-in KIEMS device for each of the polling stations a scanned image of the finalized executed Form 34A simultaneously to the various tally centres, either was substantially misused or failed to work as advertised and/or some combination of the two.  The Jubilee majority in Parliament early this year, coincident with the turnover from the Hassan-chaired IEBC to the Chebukati-chaired IEBC passed over opposition objection the option of allowing a complementary substitute for the electronic system.  As far as I can tell the IEBC did not actually plan and establish such an alternative system, nor certainly did they effectuate it in any comprehensive, demonstrable, traceable way.

Nevertheless, rather than take the seven days alloted by law, Chebukati announced alleged final Presidential results roughly 72 hours after poll closing.

Is this “close enough for horseshoes and Kenyans” or is more required to successfully conduct and conclude a presidential election in Kenya in 2017?

Update: my email to a friend regarding the Court-ordered review of IEBC presidential election data:

I haven’t finished reviewing the Registrars report in detail, but it seems clear to me that the IEBC declined to provide, as directly ordered by the Court, the evidence that would verify or falsify alleged transmittal of scanned Forms 34A by KIEMS sets from Polling Stations to Tally Centres (Constituency, Cty, Nat’l).

Whatever the Court decides to do about the on the ruling petition as a whole, allowing the IEBC to flex its muscle over the Supreme Court openly in this way would probably pretty well tell us where things are headed on rule of law issues over the foreseeable future and whether there will be a serious challenge to Ruto in 2022-32.

See “Audit Report on IEBC Servers: login trails, Forms 34A and B not provided” in The Star.

IEBC having admitted in Supreme Court that Results Transmission System did not work as advertised, March 2017 contract for KIEMS acquisition should be tabled

This is very basic stuff.  Surely one of the minimum steps required for transparency in the administration of Kenya’s election.

And who in good faith can be against that?

I will have more to say about Kenya’s 2017 election eventually, after I finally get the public records I am due from USAID from my 2015 request regarding our assistance in the administration of the 2013 election. 

[To be direct I have heard contradictory things–all hearsay–about what role the U.S. did or did not play in the KIEMS acquisition.  I just do not know.  Likewise the role of other donors.]

Election Litigation in Kenya: What is status of preservation and sharing of forensic evidence from KIEMS on Results Transmission? [update 25 Aug]


Short answer is people involved are extremely quiet on this front.

[UPDATE:  Good news!  U.S. Ambassador Robert Godec spoke yesterday at the annual National Conference of the Law Society of Kenya.  The text of his remarks is here.  Based on his remarks about transparency and the rule of law, I’m sure we will step up and do the right thing here.]

On my end I had my periodic status call today on my 2015 FOIA request to USAID about records relating to our support for the IEBC in 2013, including the failed Results Transmission System. Nothing new since April.

This is part of what I wrote July 3:

According to the EU and Carter Center election observation missions from the 2007 and 2013 elections, perhaps one-quarter to one-third of election officials at individual polling stations did not post the Form 34 showing the presidential vote count as required, so there has been ample room in each of these elections for numbers to change between the count of ballots and sealing of the ballot box at the polling station and the reported “tally” by which the president was named in Nairobi.

Unfortunately, a fair understanding of what happened in 2013 gets worse, in that it turns out that it would surely seem that the IEBC and the donors should have know ahead of time that the electronic reporting system was not going to work–but elected to project what must have been false confidence, followed by “surprise” at its failure. The president of IFES testified to the U.S. Congress in 2013 after the election that the failure was caused by a botched procurement. What was unsaid was that this was not just a procurement failure by the IEBC which IFES would have been expected to know about from its role as “embedded” within the IEBC to provide technical assistance, but that this was apparently also a botched United States government procurement from USAID through IFES, from what I eventually learned recently from my 2015 FOIA request as discussed in my post here from April:

“Kenya Election FOIA news: [heavily redacted] Election Assistance agreement shows US paid for failed 2013 “Results Transmission System”

From the Kenya Election and Political Process Strengthening (KEPPS) Program from USAID for the last Kenyan election:

“Considering the role that results transmission played in the 2007 election violence, IFES will build on its recent work with Kenya’s results transmission system to further enhance it and ensure its sustainability. IFES will ensure this system is fully installed, tested and operational for the 2012 election. Furthermore, IFES will fund essential upgrades and adjustments to this results transmission system.” 

[p.28 of the Kenya Election and Political Process Strengthening 2012 Program – Cooperative Agreement between USAID and CEPPS (coalition of NDI, IFES and IRI)]

This USAID Agreement with the consortium of IFES, NDI and IRI makes up the first 236 pages of what I was told were approximately 1800 pages of documents and attachments provided by the USAID Mission in Kenya to the Washington FOIA office by January 2016 in response to my FOIA request of October 2015.

Unfortunately, I have still not gotten any of the rest of these pages covering contract files and correspondence, as well as USAID transactions with Smith & Ouzman, Ltd., the British firm that was convicted of bribing Kenyan election and education officials to buy their products in the infamous “Chickengate” scandal.

In spite of persistent follow up over these many months, I don’t have any further information as to whether I am likely to get more of these documents released in time for the new election (under the current Kenya Electoral Assistance Program awarded to IFES last year).

The 2017 Kenyan Supreme Court petitions are under a final seven day deadline of today.  [Update: NASA has filed a challenge Friday night in Nairobi, to my understanding joined by The Thirdway Alliance.  Do not know of others.]

“Preliminary Findings” released by Kenyan civil society coalition on election

Update 23 Aug – Here is the latest from the  Kura Yangu Sauti Yangu monitoring:    KYSYElectionDataUpdate-WhyDisputed-22Aug2017

Following the unlawful raid on AfriCOG in Nairobi yesterday, today the Kura Yangu Sauti Yangu election monitoring program which has been engaged since long before any of the International Election Observation Missions were constituted, released its Preliminary Findings.

Please read for yourself (especially if you have commented publicly so far on Kenya’s election).

It would be easier for Mr. Chebukati and Mr. Kerry to make their case to Mr. Odinga’s supporters with much greater transparency

There is a lot that Kenyan voters could be told that they have not been told about how their votes were represented to them by the IEBC over the last several days since they voted and all the ballots were counted Tuesday evening.  As assurances given to the voters in 2007 and again in 2013 in the immediate aftermath of voting those years did not in some substantial respects turn out to be factually sustainable, it is no suprise many Kenyans would want to verify rather than just trust now.

One would expect everyone involved this year to anticipate questions.  There were lots of prominently published warnings of the need for transparency (from the International Crisis Group among others).

Mr. Kerry was Secretary of State in 2013 and presumably has current clearances that would allow him as an individual, now post-government service, to make doublely sure he is fully briefed about the failed Results Transmission System of 2013, as well as other past problems, if he wasn’t before coming to Nairobi last weekend for the Carter Center.  Presumably he could also ask the current US and Kenyan governments to go through the details relating to procurement and use of KIEMS this year.  Then he could answer questions and demonstrate the kind of transparency that would build trust.

Alternatively Mr. Chebukati and the current U.S. government could answer questions irrespective of the Carter Center or other independent Election Obsevation Missions.

[Updated 11 Aug] Kenya election updates 

My update is to just say Kenyan IEBC is not done with their job, but I’m done saying anything at all about this election process so far.  Since I’m not involved it doesn’t help to offer public commentary.

International Crisis Group report on “Kenya: Avoiding Another Electoral Crisis” calls on donors to show “complete transparency”; USAID is apparently not convinced yet

Counting-the original tally

Counting-the original tally

“Kenya: Avoiding Another Electoral Crisis”  March 2017 International Crisis Group paper by Murithi Mutiga

Political tensions are rising in Kenya ahead of elections in August for the presidency and other senior posts. Measures taken now can avert the risk of a repeat of electoral violence that killed hundreds of people in 2007-2008.

.  .  .  .

The equipment for transmitting results from polling places to the tallying centre is as important as the voter kits. Past elections were compromised by lack of transparency in tallying and transmitting. The installation of a transparent, efficient electoral management system would go a long way to assuaging public concerns. Unfortunately, rushed procurement, with little lead-time for testing, may set the IEBC up for failure. That would also deepen suspicions in a situation already marked by significant tension between parties. Government steps to limit the role of external partners, such as the International Foundation for Electoral Systems, that can offer valuable technical assistance, have not helped.

.  .  .  .

International partners should extend technical and financial help to the IEBC to help it better tackle the challenges. This should, however, be done with nuance, flexibility and complete transparency, in light of unfounded claims by the ruling party that external parties are seeking to influence the electoral outcome. International observers should be deployed in time to monitor crucial stages of the electoral process, such as verification of the vote register and procurement of electoral materials.

. . . .

Unfortunately, USAID is still stuck on maintaining minimal, at most, public disclosure, rather than adapt to the recommendations of the Crisis Group and the obvious lessons to be learned from the failure of 2007, especially, and 2013.

While USAID Kenya has confirmed for me that their original December 2015 Request for Agreement (“RFA”) for the $20M “Kenya Electoral Assistance Program 2017” remains a public document at http://www.grants.gov, the subsequent Agreement between USAID and IFES is not being treated as public.  Americans who want to understand our government’s approach to subsidizing the Government of Kenya’s election would be well advised to study the Request for Agreement (rfa-615-16-000001-keap-2017) closely to understand the basic structure, but will need to “ask around” informally to get any actual detail as the election now rapidly approaches.  Likewise, Kenyans who want to have input in the administration of their own election.

Meanwhile, still no documents whatsoever, from my October 2015 request for USAID documents relating to our support for the 2013 Kenya election (!).

See “IEBC must look us in the eye and say, ‘We aren’t ready for August'” by Tee Ngugi in The East African.

Solo 7 — Toi Market

How did Donald Trump get this far?; our actions in Kenya since 2007 are an example of why Americans are frustrated [updated 10/10/16]

[I updated to correct an error — the USAID Inspector General, rather than the U.S. Government Accountability Office, conducted the referenced investigation that found USAID funds went into supporting the “yes” campaign in the 2010 Kenya referendum, rather than providing only neutral process support for Kenyan voters.]

Longtime readers of this blog will well recognize Kenya as a glaring example of the refusal of our government and the surrounding networks of foreign policy elites in the larger Washington Beltway community to seriously self-assess and try to level with the American people in such a way as to build trust and confidence (even  in the face of our serious and determined foes).

The stolen election in Kenya and its aftermath in 2007-08 was clearly a catastrophe for both the Kenyan people–whom we are continually trying to assist to the tune of hundreds of millions of dollars a year–and for security interests of the United States (whatever real or rationalized internal claims might or might not exist to justify our policy of “looking, and pointing, the other way” as we saw the election being stolen).  So far as I can assume, the Kibaki team would surely have done whatever was necessary to obtain the ECK certificate as “winner” of the election irrespective of the actual voting even if “we hadn’t even been here” (see here) but the very least we have to conclude is that our elaborate and expensive electoral assistance effort was in crucial respects a failure.  And we certainly do have to consider the possibility that the other donors could have done better to accomplish what were identified as the common objectives without us and our leading role.

4292493510_a3a02344a7_o

A key reason I have dedicated my “War for History” series to my late friend Joel Barkan–along with my late friend Peter Oriare–was that Joel was one of the rare people in Washington willing to speak out when he saw our country making what he saw as a foreign policy mistake.  He wisely warned IRI that we were risking embarrassment along with the State Department.  Was he thanked when it became obvious that he had been right?–no, he was attacked instead, in the finest Washington tradition of “CYA by pointing your finger at the person who suggested you ought not to show it in the first place”.

Having found myself playing a bit part due to working for a “charity” that got tagged, along with USAID, by our Ambassador to play a role neither my organization nor USAID sought as of the time I moved my family to Kenya to help out, I find myself being the only one seemingly willing to offer any type of public mea culpa for those decisions that I would make differently in hindsight.  And I know that I absolutely did my best even though I was not successful overall.  I cannot help but wonder if that is really the case for everyone, given all the various potential interests to be served.

In spite of how badly things went we have just given ourselves credit–and let the individuals who were in key roles publicly pat themselves on the back–for helping to keep the aftermath of the stolen election from being worse than it was.  I did not have any personal animus against Ambassador Ranneberger and did not want him to be precipitously “recalled” as a result of my complaint about his interference with the election observation, but I would never have imagined that with a big political turnover in the U.S.–based to a great extent on a public loss of confidence in foreign policy decision making–Ranneberger would still end up being one of the most prominent public actors in Kenyan politics–on behalf of my country–for several more years afterward and be our second-longest serving Ambassador to Kenya ever.

Through the persistence of the subsequent Chairman of the House Foreign Affairs Subcommittee on Africa, Rep. Christopher Smith, we eventually learned through a Kevin Kelly story in Kenya’s Daily Nation that the USAID Inspector General had determined that an “eight figure” sum of money had bled over from lawfully neutral process support for constitutional reform into the 2010 “Yes” referendum campaign. Personally believing that on balance Kenyans would be better off to pass rather than defeat the referendum, I was embarrassingly gullible myself in being hesitant to credit Congressman Smith’s concerns in this regard until I saw the reporting on the USAID Inspector General’s findings.  Shocking that the Ambassador who was not neutral in the 2007 vote was not neutral again in 2010!

In the 2013 general election, the administration of the process was in substantial ways even worse than in 2007 as capably pointed out by John Githongo and many others of earned expertise. Our assistance was much more expensive, and while not so controversial, was again not very transparent  at all.  (Still nothing on my public records request to USAID regarding our spending through IFES on Kenya’s IEBC and its corrupt technology procurements.)

And now, here we go again.  The Uhuruto re-election gears up against the ODM-led opposition with the Government of Kenya facing its inevitable referral to the Assembly of State Parties of the International Criminal Court since it–inevitably and predictably–refused to meet its legal obligations to cooperate with the Court.

The individual who served as Assistant Secretary of State during the 2007-08 catastrophe, as a private citizen but identified primarily in her role as a former high ranking diplomat, was a key figure again in the 2013 campaign–this time speaking out (informally I assume) to accuse the United States Government of interfering in the election in the opposite direction, in favor of the opposition and against her preferred candidate, Uhuru Kenyatta.  While she was within her rights, her argument seems counterfactual when you look at how U.S. assistance to the Government of Kenya and NDI/ELOG and IFES for the election was actually used in totality: to sell whatever the IEBC decided, even without a transparent tally and even though we had some real knowledge of the corruption issues that have eventually come out to the point of forcing their buyout after the Opposition was willing to protest on the streets this year.

If you will read the ELOG final report from several months after the election, you will see that it appears that the NDI/ELOG Parallel Vote Count had more problems with falloff of planned data collection than the 2007 IRI exit poll–but since it involved a much smaller universe of locations than an exit poll I’m not sure that this could be adjusted for (if attempted).  So the idea that the 49.7% PVT result “VERIFIED” that Uhuruto received more than 50% looks that much more like advocacy for the IEBC rather than facts for the voters.

I would never vote in a scenario that I can readily imagine for Donald Trump or someone much like Donald Trump as best I understand him.  I agree that his positions–none of which I assume reflect any sincere value judgments–are dangerous to our country now and for my children’s future.  But if you don’t understand why many Americans might have some temptation to go for “the candidate of the middle finger” out of frustration with a sense that “Washington” isn’t actually working on their behalf as they send their taxes, you cannot be getting out enough.