1.1 Background
On 27 December 2007 some ten million Kenyans went to the polls in what was generally anticipated to be the most hotly contested and close-run presidential, parliamentary and civic elections in the country’s 45 years since emerging from British colonial rule. The register of voters had been swelled since the previous elections by several million new registrations, many of them young first-time voters, and the Electoral Commission of Kenya (ECK) had doubled the number of voting stations to 27 555, arranged in some 20 000 polling centres.
Campaigning at all three levels of the contest had been vigorous, characterised by robust language occasionally lapsing into ethnic hate-speech and deteriorating into violence. Since the constitutional referendum in 2005, political discourse in Kenya had been sustained at a high pitch and tended to focus on the presidential contest. The two main presidential candidates, incumbent President Mwai Kibaki and former ally Mr Raila Odinga, had led opposing sides in the referendum, which was won handsomely by the Odinga side. It was therefore hardly surprising that a prominent feature of the ODM parliamentary and presidential campaigns was the claim that only rigging could prevent their taking power at the elections. This was particularly serious as public comment on the manner and timing of the appointment of the majority of electoral commissioners during 2007 had already cast a shadow of suspicion over the ECK’s impartiality. State power in Kenya, harking back to the country’s colonial past and decades of one-party rule, remained vested in a centralised executive exercising control through a network of provincial administrators/district commissioners, a vocal but relatively powerless legislature and a compliant judiciary exercising few checks and balances. The presidency was, rightly, seen as the ultimate political prize. Elections in Kenya have been characterised by intensified awareness of ethnic divides and deep-seated historical land grievances, especially among rural communities. President Kibaki, heading the Party of National Unity (PNU) ticket and drawing his support mainly from the Kikuyu, Embu and Meru communities of Central and central Eastern provinces, campaigned principally on his socio-economic record.
Mr Odinga at the head of the Orange Democratic Movement (ODM), with the support of largely the Luo, Luhya, Kalenjin and some smaller ethnic communities, vocalised the need for fundamental political and socio-economic reform and devolution of state power. Although the emphasis was more pronounced at the civic and parliamentary levels, and in the rural areas, the ethnic configuration of the PNU and the ODM, and the origins of the two main contenders in the presidential contest, remained a factor. Opinion polls predicted a close contest, Odinga leading but Kibaki later narrowing the gap. The PNU, though registered as a political party under Kenyan law as it then was, was in reality an electoral alliance. The ODM, though also recently assembled, was a fully-fledged political party, more cohesively organised and hence generally posting a single candidate in each of the provincial and civic contests. The PNU, though uniting behind their single presidential candidate, in the other two elections allowed the party’s various components to field candidates under their individual banners, often in competition with one another.
Having regard to the scope and complexity of the undertaking, polling, counting and announcement of results seemed satisfactory – wholly unjustifiably, as would become only too apparent in due course. All also seemed well (once again deceptively so) with the transmission of the requisite documents to returning officers at constituency level and the onward transmission by them of data to the Kenyatta International Convention Centre (KICC) in downtown Nairobi where the ECK had established its national data tally and media centre.
There, however, there were ominous portents from the outset. Commissioners and staff of the ECK proved ill-prepared for the relatively straightforward but highly sensitive exercise of receiving, verifying, tallying, tabulating and announcing the presidential results. The ensuing spectacle left an indelible impression on visitors to the media centre and on millions of television viewers. Six months later, informant after informant around the country could vividly recall their astonishment and anger at the fiasco and could mimic and quote the ECK chairman verbatim.
In the event the PNU and its scattered array of allies were defeated in the parliamentary and civic elections. Also, as results trickled in, first from ODM strongholds and only later from the PNU heartland, President Kibaki trailed most of the time and only started catching up well into the tallying exercise. He was ultimately announced the winner (by 231 728 votes) in the late afternoon of 30 December 2007, and then hurriedly sworn in, notwithstanding vociferous protests that the result had been rigged by the ECK. These protests and an ODM press conference were abruptly silenced by a news blackout and summary security clampdown as armed soldiers bustled candidates, party agents, diplomats and domestic as well as international observers out of the KICC.
Some observers were aghast, others who had been allowed into the tally centre were volubly incensed by what they regarded as evidence of malfeasance on the part of the ECK committed in their very presence. Upward adjustment of already announced results from some populous pro-Kibaki constituencies, seemingly favouring the President, fanned the flames of suspicion. Televised utterances by Chairman Kivuitu only served to make matters worse, as did a hurriedly composed media statement released by four out of twenty-two commissioners, commenting on the turn of events and calling for calm.
Widespread and often ethnically motivated violence erupted and rapidly spread. Over the ensuing six or seven weeks approximately 1,150 people were killed, property damage ran to billions of Shillings and some 300 000 Kenyans were forced to flee their homes and livelihoods.
1.2 Scope of mandate
* Pursuant to the political pact brokered by Mr Annan and his colleagues, the seven members and the secretary of IREC were consensually identified and formally appointed by President Kibaki under the Commissions of Inquiry Act (Cap. 102). IREC’s terms of reference (ToRs) were published in Gazette Notice 1983, Kenya Gazette of 14 March 2008 (annex 1.A) and mandated examination of the 2007 elections from a number of different angles:
* The constitutional and legal framework to identify any weaknesses or inconsistencies.
* The structure and composition of the ECK in order to assess its independence, capacity and functioning.
* The electoral environment and the role of the political parties, civil society, the media and observers.
* The organisation and conduct of the 2007 elections, extending from civic and voter education and registration through polling, logistics, security, vote-counting and tabulation to results-processing and dispute resolution.
* Vote-tallying and -counting to assess the integrity of the results of the entire election with special attention to the presidential contest.
* Assess the functional efficiency of the ECK and its capacity to discharge its mandate.
* Recommend electoral and other reforms to improve future electoral processes.
* Within six months to submit to President Kibaki and the Panel its findings and recommendations which are then to be published within 14 days.
1.3 Overview of report
This report first outlines how IREC set about executing its mandate, then details some of the salient aspects of its activities and findings, broadly discussed by reference to the ToRs, and concludes with a number of specific recommendations aimed at preventing a recurrence of the tragic events that gave rise to IREC’s appointment. The discussion and recommendations broadly follow the ToRs under the following headings:
* Constitutional and legal framework, describing Kenya’s existing legal framework and past efforts to reform it.
* The organisational structure of the electoral management system, reviewing the electoral system, the composition of the ECK commission and secretariat, their appointment, functions and interaction, the functional efficiency of the ECK and its funding.
* Public participation in the 2007 elections, examining in turn the role of political parties, the media, civil society and observers.
* The organisation and conduct of the 2007 general elections, starting with boundary delimitation, voter and civic education, registration of voters and political parties, nomination of candidates, recruitment of election-day personnel, electoral logistics, provisioning, security, campaign regulation and technical assistance received by or offered to the ECK.
* Counting, tabulation, transmission, tallying and verification of results, examining complaints raised in this context by the ODM and the reactions of the PNU and the ECK, criticisms by observer groups, an overall evaluation of ECK procedures, comment on system integrity, a statistical analysis of a sampling of results and developing a finding on the integrity of the results, with special reference to the presidential election.
* Post-election procedures, including swearing-in and assumption of office, dispute resolution and review, audit of results, custody of balloting material and analysis of post-election procedures.
* Lastly and most importantly, detailed recommendations concerning each of the topics listed above.
1.4 Method adopted
1.4.1 Phases of activity
IREC, having met formally for the first time on 19 March 2008 (and being sworn in the following day) set about performing its mandate in relation to each of the ToRs as follows –
o A first phase of analysing the ToRs, delineating and disaggregating the tasks thus identified, reviewing documents made available by the United Nations Development Programme (UNDP), which had implemented a Joint Election Assistance Programme funded by a group of eight foreign donors and by the co
o ordinator in charge of the Programme Management Unit, the International Foundation for Electoral Systems (IFES), the ECK and diverse NGOs and members of the public; devising a work plan and constituting committees to implement such plan; identifying human and material resource requirements; interviewing, recruiting, engaging and briefing staff; finding, furnishing and equipping office accommodation. (IREC wishes to record its gratitude to UNDP and the Government of Kenya, both of which gave invaluable assistance in getting IREC started.)
* A second phase of drafting and promulgating rules and procedures (see annex 1.B) providing for both inquisitorial and adversary procedures for receipt of both oral and written sworn and unsworn evidence while observing the elementary rules of fair play; engaging assisting counsel to lead investigations and present evidence. (Here IREC wishes to record its gratitude to the Attorney General of the United Republic of Tanzania, and through him to his government, for seconding Advocate Yohane Masara to assume this office.) Concomitantly it was decided to adopt an adversary system of determining disputed facts and/or opinions, affording the ODM, PNU and ECK standing. During this phase, a personnel structure was created making provision for legal and other researchers, investigators, rapporteurs and support staff.
* The third phase comprised, first, a series of preliminary meetings with the principal role players, namely the ECK and the three main parties that contested the presidential election, a breakfast meeting with media representatives, and briefing sessions with UNDP and IFES staff who had been intimately involved in the Joint Election Assistance Programme and a USAID-sponsored assistance programme. Then followed a series of publicly advertised meetings in Nairobi with various selected categories of stakeholders, e.g. human rights agencies, professional organisations, civil society and faith-based organisations, the business sector, extending open invitations to them and the general public to submit factual and/or opinion material relevant to the ToRs. IREC also participated in two full-day sessions and subsequent briefings with service providers and donors engaged in reviewing the Assistance Programme. A number of intensive briefing sessions were held with the co-ordinator of the programme management unit responsible for the Assistance Programme in the course of which IREC gained informed, first-hand information about the ECK’s planning and preparation for the 2007 elections, which information was then explored in correspondence and meetings with the ECK.
o The fourth phase was an extensive five-week programme of public meetings at 36 venues throughout the country, starting in Nairobi and covering the widest
o feasible cross-section of Kenyan society, from Mandera to Homa Bay and Turkana to Kwale, IREC mostly splitting into two committees, noting the experiences, concerns, opinions and proposals of over 1,200 speakers of all political persuasions, age groups, walks of life and communities. These meetings were enormously instructive, affording the IREC teams a unique cross-section view of Kenyan society and an invaluable sense of public opinions, perceptions, fears and aspirations. The members attending these meetings were deeply impressed with the determination of ordinary citizens to make a contribution to IREC’s data-gathering. Of particular significance was that, despite this wide spectrum of interlocutors, a number of common themes were encountered virtually everywhere. The most striking was the call for change, change to the Constitution, to the political system, to the electoral system, often an inchoate longing for things to be done differently. Deeply impressed though they were by these communications, the commissioners remained conscious that they were not hearing and testing formal evidence. The speakers were often biased, obviously self-selected and at times had clearly been put up to it by party agents. Their communications were often subjective, ill-informed and/or manifestly unsubstantiated, at times even deliberately untruthful. Many proposals were radical, utopian or far-fetched. Interventions were often characterised by anger at what was seen as irrefutable televised evidence of rigging by the ECK at the KICC, coupled with demands for punitive action against the chairman and his colleagues. It is also important to record repeated reports of rampant bribery, vote-buying, intimidation, abuse of government resources (both human and material), partiality and incompetence of ECK field staff.
* The fifth phase entailed, first, analysing, distilling and evaluating the mass of information garnered in the course of the country visits, and then conducting a programme of six technical workshops with experts in various fields pertaining to the ToRs. An additional day of expert debate was devoted to the presentation (followed by discussion with selected participants and interested parties) of papers presented by political analysts dealing with the integrity of the election results.
o The sixth phase was devoted to a succession of formal hearings in which a number of key witnesses were questioned under oath by IREC’s legal staff, cross-examined by the ODM, PNU and ECK legal representatives and then questioned by IREC commissioners. An extended day was devoted to the evidence of Ms Koki Muli, a foremost Kenyan electoral expert with many years of experience observing Kenyan elections who had moreover witnessed many of the happenings at the KICC during the vital period. The further witnesses were Mr Kivuitu, chairman of the ECK, Commissioner Tumwa, who chaired the ECK committee responsible for the setting-up of the national tallying centre, Mr Chege, the ECK
o secretary who was in charge of the centre, Mr Imbira, the ECK IT manager, Ms Agnes Kisero, the day-shift leader of tally team 5 in the national tally centre, which had dealt with twenty-one Central Province constituencies, including several that had been contentious during the tallying process. The formal hearings then dealt with a number of persons who had been returning officers in problematic constituencies. Interrogation of further witnesses and detailed follow-up of information conveyed at the country meetings was not possible within the time constraints of IREC’s mandate.
* A further half-day was devoted to a statistical analysis of the presidential election results and an eyewitness-based chronology of events at the KICC, presented on behalf of Kenyans for Peace with Truth and Justice (KPTJ), an umbrella of legal, human rights and governance organisations that have been playing an important role in Kenya’s electoral process and politics for the last fifteen years and which was formed after the 2007 elections.
In summary: essentially five methods were used to gather relevant information: (i) a review and evaluation of published material and of internal documents of the ECK, the UNDP project management unit and IFES; (ii) public meetings to receive opinions, comments, factual allegations, complaints, recommendations and whatever else the people of Kenya wished to convey; (iii) more formal meetings and workshops with professional bodies and experts; (iv) consideration and debate of written submissions and oral adumbration by the legal representatives of the ECK, ODM and PNU; and (v) formal hearings where witnesses were examined under oath. Save in the case of one witness, all these proceedings were held in public.
1.4.2 Inquisitorial/adversary method
IREC decided to adopt a flexible approach to its data-gathering activities and crafted its rules and procedure accordingly (see annex 1.B). Although the IREC rules include provision for formal hearings in the nature of court proceedings, intended for determination of disputed questions of fact, the IREC rules also made provision for information to be gathered in less formal ways. This proved most useful in relation to the information of a general nature that was gleaned in the course of the public meetings.
As far as specific allegations of fact were concerned, IREC sought to co-opt the principal role-players, especially the two political adversaries, in a quasi-judicial procedure where the legal representatives would participate in formulating defined issues arising out of their respective sets of contentions and then identifying the relevant witnesses to be called to testify on such issues. Unfortunately this proved unattainable.
Although the two political parties and the ECK, acting through their respective counsel, reciprocally served copies of their formal representations outlining their contentions, and reciprocally responded to such contentions, a precise delineation of issues proved an elusive goal and ultimately IREC relied on the good sense and professional judgment of assisting counsel and his team to identify the questions of fact on which oral testimony was to be presented and the witnesses to be called. The two political parties and the ECK, having been granted formal standing, were invited to cross-examine such witnesses. The resultant discourse proved useful in respect of some of the issues.
1.4.3 Interaction with the ECK
In conformity with its policy to seek the willing co-operation of all potential contributors to its researches, IREC established a sound working relationship with the ECK which was maintained throughout and with one notable exception (relating to minutes of certain meetings) afforded IREC ready access to information, documents and witnesses within the ECK’s control. Two avenues of communication were established, one at a formal level via the legal representatives, one at a more direct and informal technical level. As appears from chapter 6, access to constituency files and analysis of their contents proved extremely useful.
1.5 Summary of conclusions
In the chapters that follow, IREC’s conclusions are presented in their appropriate context. They are, however, summarised here for the convenience of the reader:
The voter register The voter register, which has been updated from time to time since 1997, is materially defective in three respects that in themselves already impair the integrity of the election results: (i) Registered voters represent only 71% of .the voting-age population of Kenya. (ii) The register probably includes the names of some 1.2 million deceased persons. (iii) Women and voters between 18 and 30 years of age are significantly under-registered. Furthermore, members of certain marginalised communities encounter difficulties in obtaining their national identity cards, a prerequisite for registration as a voter.
Delimitation The gross disparity in the voting populations of Kenya’s constituencies breaches the fundamental equality principle of democracy, which is clearly articulated and enshrined in section 42(3) of the Constitution of Kenya, namely one person, one vote. This long-standing discrimination in itself impairs the integrity of the electoral process, mainly, but not only, in relation to parliamentary elections.
Fraud Numerous implausibly high turnout figures reported in the strongholds of both main political parties evidence extensive perversion of polling, probably ballot-stuffing, organised impersonation of absent voters, vote buying and/or bribery. This inference is supported by numerous eyewitness accounts given to IREC of various forms of manipulation as well as election observers’ observation reports and ECK submissions.
Indeed, vote-buying and ballot-stuffing appear to be such extensive and universally condoned practices in Kenyan elections that the question can rightly be asked whether genuinely free and fair elections are at all possible.
Permitted irregularities A likely facilitator and catalyst for ballot-stuffing (of which effect the political parties and the ECK could hardly have been unaware) was the indulgence granted by the ECK shortly before the elections for “black books” (in which the names of voters had been entered at the time of registration) to be used in certain circumstances and for double registrants to be allowed to vote, contrary to previous regulation..
Exclusive strongholds A further contributor and facilitator for manipulation at polling stations is the disturbing feature that in many instances (in the strongholds of both main political parties) effectively only the majority party was represented during polling and counting. The alert self-interest of competitors is all but indispensable for honest elections and it is a matter for serious concern that this safeguard was absent in many instances.
Defective planning The system of tallying, recording, transcribing, transmitting and announcing results was conceptually defective and poorly executed. The ECK had long since been aware of the need to revise the system fundamentally by introducing readily available information and communications technology. Its failure to do so was grossly remiss and contributed to the climate of tension, suspicion and rumour in which the violence erupted.
Electoral system The first-past-the-post electoral system carried over from colonial days, with its potential for distortion, real or perceived, compounded the problems caused by the originally gerrymandered, outdated and grossly skewed constituency delimitation pattern. In particular the circumstance that the ODM, which did not scatter its support in the parliamentary contests, beat the PNU and its motley array of allies constituted proof positive of rigging for many observers unfamiliar with the vagaries of the first-past-the-post system, the more so where constituency sizes differ materially.
Delays The gross disparity in constituency sizes also contributed to defects and delays in tallying, recording, transcribing and transmitting results (having moreover to be done by outdated methods) in the large constituencies, which defects and delays also contributed materially to the explosive political climate which then built up as the country awaited the transmission and announcement of results.
Incompetence The conduct of the 2007 elections was so materially defective that it is impossible – for IREC or anyone else – to establish true or reliable results for the presidential and parliamentary elections. IREC has, however, established by means of statistical analysis of a sample of constituencies that innumerable elementary mistakes in tallying and/or transcribing results as well as patent mistakes of omission, duplication and confusion were made.
Integrity of result Therefore, although there is room for honest disagreement as to whether there was rigging of the presidential results announced by the ECK, the answer is irrelevant, as (i) the process was undetectably perverted at the polling stage, and (ii) the recorded and reported results are so inaccurate as to render any reasonably accurate, reliable and convincing conclusion impossible.
Law enforcement and dispute resolution The system of electoral law enforcement and dispute resolution are conceptually defective and were executed poorly if at all by the ECK and the ordinary law-enforcement agencies of government.
Communication There was no effective communication between the ECK and political parties, observers, the media or the public, particularly regarding the national tally centre, the lack of transparency resulting in misconceptions, suspicion, rumour and anger.
The ECK The manner of appointment of commissioners and the structure, composition and management system of the ECK are materially defective, resulting in such a serious loss of independence, capacity and functional efficiency as to warrant replacing or at least radically transforming it.
Constitutional and legal framework Although Kenya has a legal framework for the conduct of elections, material defects in the framework combined with a culture of lawlessness at election time bring into question the capacity of the law to provide a sufficient framework for political competition.
Wider responsibility Though the ECK is primarily responsible for the flaws in the 2007 general elections, Kenyan society has long condoned, if not actively connived at, perversion of the electoral process.
Long-term commitment This culture of electoral lawlessness has developed over many years and cannot be reversed without a concerted, non-partisan commitment to electoral integrity on the part of political leaders, which commitment will need to be sustained and monitored over time.
