Eu Eom Kenya 2022 en

European Union
Election Observation Mission
Kenya 2022
Final Report
General Elections
9 August 2022
The Election Observation Mission is independent from the institutions of the European Union. The
views and opinions expressed in this report are those of the authors and do not necessarily reflect
the official policy and position of the European Union.
EU Election Observation Mission in Kenya Final Report
General Elections, 9 August 2022 Page 2 of 72
Table of Contents
I. Executive Summary ........................................................................................................................ 6

II. Introduction ................................................................................................................................. 7
III. Political Background................................................................................................................... 8
IV. Implementation of Previous EOM Recommendations ............................................................... 9
V. Legal Framework ...................................................................................................................... 10
A. International Principles and Commitments ........................................................................... 10
B. Constitutional Rights ............................................................................................................. 10
C. Electoral Legislation ............................................................................................................. 11
D. Electoral System and Constituency Delimitation.................................................................. 11
VI. Electoral Administration ........................................................................................................... 11
A. Structure and Composition of the Election Administration .................................................. 11
B. Administration of Elections .................................................................................................. 13
C. Voter Education ........................................................................................................................ 14
VII. Electoral Technology ................................................................................................................ 15
VIII. Voter Registration ................................................................................................................. 17
A. The Right to Vote .................................................................................................................. 17
B. Voter Registration Procedures .............................................................................................. 17
IX. Registration of Candidates and Political Parties ....................................................................... 19
A. Registration of Candidates .................................................................................................... 19
B. Registration of Political Parties ............................................................................................. 20
X. Campaign Environment ............................................................................................................ 21
A. Election Campaign ................................................................................................................ 21
B. Campaign Finance ................................................................................................................. 23
XI. Media ........................................................................................................................................ 24
A. Media Environment ............................................................................................................... 24
B. Legal Framework for the Media............................................................................................ 25
C. Media and Elections .............................................................................................................. 25
D. Media Monitoring Findings .................................................................................................. 27
XII. Social Media and Digital Communications .............................................................................. 28
A. Social Media Environment .................................................................................................... 28
B. Legal Framework .................................................................................................................. 28
C. Social Media Monitoring (see also Annex II) ....................................................................... 29
XIII. Inclusion of Women, Persons with Disabilities .................................................................... 31
XIV. Citizen and International Election Observation .................................................................... 33
XV. Electoral Justice ........................................................................................................................ 34
A. Political Party Nomination & Candidate Registration Disputes ........................................... 34

B. Constitutional Petitions on Electoral Rights ......................................................................... 35
C. Election Offences .................................................................................................................. 35
XVI. Polling, Counting and Tabulation of Results ........................................................................ 36
A. Overview of Voting............................................................................................................... 36
B. Closing and Counting ............................................................................................................ 37
C. Tabulation of Results ............................................................................................................ 38
XVII. Results and Post-Election Environment ................................................................................ 40
A. Declaration of Results ........................................................................................................... 40
B. Complaints Relating to the Election Results ......................................................................... 41
C. Post-Election Developments ................................................................................................. 45
XVIII. Recommendations ................................................................................................................. 48
XIX. ANNEXES ............................................................................................................................ 59
A. ANNEX I – EU EOM MEDIA MONITORING .................................................................. 59
B. ANNEX II – SOCIAL MEDIA MONITORING RESULTS................................................ 66
List of Acronyms
AU African Union
BBI Building Bridges Initiatives
BVR Biometric Voter Registration
CEDAW UN Convention on the Elimination of All Forms of Discrimination Against Women
COMESA Common Market for Eastern and Southern Africa
CRO Constituency Returning Officers
CRPD UN Convention on the Rights of Persons with Disabilities
CRS Civil Registry Services
CS/CEO Commission Secretary / Chief Executive Officer
CRO County Returning Officer
CSO Civil Society Organisation
CVR Continuous Voter Registration
DIS Department of Immigration Services
DRC Dispute Resolution Committee
DRO Deputy Returning Officer
EAC East African Community
EACC Ethics and Anti-Corruption Commission
ECOWAS Economic Community of West African States
EISA Electoral Institute for Sustainable Democracy in Africa
ELOG Elections Observation Group
EOM Election Observation Mission
EVV/EVI Electronic Voter Verification and Identification
ECVR Enhanced Continuous Voter Registration
EU European Union
HC High Court
ICCPR UN International Covenant on Civil and Political Rights and Optional Protocols
IEBC Independent Electoral and Boundaries Commission
IGAD Intergovernmental Authority on Development
IGP Inspector General of Police
ILO International Labour Organisation
IPOA Independent Policy Oversight Authority
IRI International Republican Institute
KEG Kenya Editors Guild
KIEMS Kenya Integrated Election Management System
KMSWG Kenya Media Sector Working Group
KNHCR Kenya National Commission on Human Rights
KUJ Kenya Union of Journalists
MCA Members of the County Assembly
MCK Media Council of Kenya
NA National Assembly
NCCK National Council of Churches of Kenya
NCIC National Cohesion and Integrity Commission
NDI National Democratic Institute
NRB National Registration Bureau
NRO National Returning Officers
NTC National Tallying Centre
ODM Orange Democratic Movement
ODPP Office of the Director of Public Prosecution

ORPP Office of the Registrar of Political Parties
OSCE Organization for Security and Co-operation in Europe
PJAK Political Journalists Association of Kenya
PPDT Political Parties Disputes Tribunal
PVT Parallel Vote Tabulation
PwD Persons with Disabilities
RO Returning Officer
ROV Register of Voters
RTS Results Transmission System
SC Supreme Court
TJRC Truth Justice and Reconciliation Commission
UDA United Democratic Alliance
UN United Nations
UNCAC UN Convention against Corruption
UNDP UN Development Programme
WFD Westminster Foundation for Democracy
I. Executive Summary
The 9 August 2022 general elections in Kenya took place in a peaceful environment overall, with
fundamental democratic rights respected. Six separate votes were held at different levels concurrently
and this made for a complicated process that took nearly a week to finalise, with numerous challenges
emerging. Nevertheless, the fear of violence that was present in previous elections and was a concern
prior to this one did not emerge, and a well-developed network of conflict prevention organisations
and activities served to mitigate this apprehension. This involved committed state and civil society
actors whose actions often positively impacted the process.
Another positive element was the emergence of socio-economic issues as a key focus of the electoral
campaign. Although this element is just developing and ethno-regional cleavages continue to underlie
voter choice, there is hope that the emergence of issue-based elections is something that might continue
to develop in the future.
One of the main features of these elections was that the complicated process, combined with the late
start in preparations by the Independent Electoral and Boundaries Commission (IEBC). This meant
that many elements were not considered as early as they should have been and were often dealt with
in an ad hoc fashion, rather than systematically.
The IEBC’s administration of the elections was challenged by a lack of resources, both financial and
human, a dearth of proper communication that undermined transparency, as well as political divisions
that became fully apparent as the process developed. It also called into question the institutional
independence of the body, which needs to be further buttressed. Robust institutions that are truly
impartial are essential for public trust in future electoral processes.
The IEBC’s difficulties were further compounded by ongoing complaints and appeals regarding
various elements of the process, with decisions often coming late in the process and challenging the
IEBC’s implementation. One example of this was the decision of the Court of Appeals on the eve of
elections to allow for the use of manual voter registers, whereas a previous court decision had
determined that they should not be used; thus, instruction to lower-level commissions that had already
been transmitted down had to be rescinded at the last moment.
The elections were marred by the lack of a proper campaign finance framework, which meant that
incomes and expenditures went largely unregulated and corrupted the political space. Numerous
allegations of corrupt resources being used by political contestants and of voters being paid for their
consideration were rife. This significantly undermines democratic principles and practices. Without a
proper redress of the role of money in politics in the future, democratic progress will be stymied.
Another key factor in these elections was the results transmission process, which was complicated and
time-consuming, despite the technology used. One positive element was that the public results portal
for the presidential race was able to be established, although at the very last minute. This led to greater
transparency in that results forms were uploaded at the precinct level and could be checked online by
interested parties. This also, however, formed a major part of the eventual election petition that was
brought by the main presidential challenger before the Supreme Court.
Post-election petitions led to a contentious legal battle that pitted the main rivals against each other
and against the IEBC, leading to questions of public trust in the results. Despite the lack of evidence
presented to the Supreme Court (SC) by the challenger and its subsequent decision to uphold the
presidential election results, the legal proceedings raised questions regarding the proper
implementation of the process that should be addressed in the future in order to build better trust in
electoral processes.
The elections highlighted many positive elements in Kenya’s continued democratic development and
desire to improve the electoral processes. Nevertheless, there are a number of areas that require urgent
reform and which form part of the recommendations made by the EU EOM in this report. The priority
recommendations are listed here for ease of reference:
Priority Recommendations:
• IEBC to be provided sufficient funds immediately after the end of the current electoral cycle
and throughout the next one, to implement continuous voter education activities.
• IEBC to implement improvements in election technology based on regular audits and
stakeholder consultations and to improve consistency and transparency.
• IEBC to operationalise the Election Campaign Financing Act to regulate the amount of money
received and spent by candidates and political parties during an election or referendum.
• IEBC to introduce procedures for special voting of election officials, security personnel on
duty, homebound and hospitalised voters as well as nomadic pastoralists, to vote, if necessary,
in another location than the one they registered. This should be balanced against security
measures to avoid multiple voting.
• Parliament to detail the legal definition of hate speech in line with international human rights
obligations (so both intention to incitement and imminent violence are demonstrated).
• Parliament to remove section 22 and 23 of the Computer Misuse and Cybercrimes Act and
explore less intrusive measures for addressing disinformation.
• IEBC to introduce clear and detailed procedures for managing staff and the flow of information
at tallying centres to increase the consistency and efficiency of the process, especially the
public display of results being aggregated there, and the uploading of all results forms to the
public portal to ensure verifiability.
II. Introduction
The European Union (EU) deployed an Election Observation Mission (EOM) to observe the 9 August
general elections in Kenya following an invitation from the Kenyan authorities. The EU EOM was
present from 26 June to 6 September 2022. The mission’s mandate was to observe all aspects of the
electoral process and assess the extent to which the elections complied with international and regional
commitments for elections, as well as with national legislation. The EU EOM comprised a core team
of 12 experts based in Nairobi and 48 long-term observers deployed to all areas of the country.
For election day on 9 August, the EU EOM deployed an additional 48 short-term observers and 44
locally recruited short-term observers to monitor voting, counting and tabulation of results. In total,
the EU EOM deployed 182 observers from 27 EU member states, and from partner countries Canada,
Norway, and Switzerland as well as a seven-member delegation from the European Parliament. The
EU EOM is independent in its findings and conclusions under the leadership of the Chief Observer,
Ivan Štefanec, Member of the European Parliament. The mission followed an established methodology
and adhered to the “Declaration of Principles for International Election Observation”, endorsed under
United Nations auspices in October 2005, now espoused by over 50 organisations. The EU has
previously deployed EOMs in 2002, 2007, 2013, and 2017.
III. Political Background

On 9 August 2022, Kenya held its seventh general elections since the return to a multiparty system in
1991 and it third general elections under the 2010 Constitution, voting in six different elections. The
presidential race uses the majoritarian system by which the winner is required to gain 50 per cent plus
one vote and at least 25 per cent of valid votes ca in at least 24 of the 47 counties. If no winner emerges,
a second round of elections is organised among the two candidates who received the most votes.
The five other elections are first-past-the-post: elections for 47 county governors, 290 National
Assembly (NA) members, 47 senators, and 1,450 county assembly members (MCA). Additionally,
members representing women, persons with disabilities (PwD), youth, workers, and minorities are
nominated by parties proportional to their vote share in parliament and in the county assemblies.
Kenya has a history of disputed presidential elections with the 2002 polls being considered the only
election since 1992 concluded without major disputes. The 2022 elections were conducted against the
backdrop of the three previous, highly disputed presidential elections, of which one led to large scale
post-election violence (2007 / 08), and one had been nullified by the Supreme Court in a landmark
2017 ruling. Kenya elections are also considered an elite and male-dominated affair due to the high
costs associated with candidate nomination which highly impacts the ability of women, youth, and
minority groups to be fairly represented.
For the 2022 general elections, the political landscape was reorganised. President Uhuru Kenyatta and
his Deputy William Ruto had fallen out after Kenyatta’s ‘handshake’ in March 2018 with his opponent
from the 2013 and 2017 presidential elections, Former Prime Minister Raila Odinga, and their common
Building Bridges Initiatives (BBI). The BBI aimed to expand senior governance structure to
incorporate more political leaders to overcome problematic impacts of the winner-takes-it-all electoral
system and at an upward revision of the financial resource allocation formula for counties and
constituencies. However, court rulings by the High Court (HC) and the SC decided BBI to be
unconstitutional for failing the legally binding public participation criteria and it could, consequently,
not be implemented. This controversy over BBI further cemented the political fault lines between the
Kenyatta / Odinga and the Ruto camps.
In presidential elections, Kenyans have traditionally voted broadly along ethno-regional lines whereas
in all other races a complex mix of national, regional, and local factors determined voters’ choice. The
four former presidents came from two communities (the Kikuyu of the Mount Kenya region and the
Kalenjin of the central and north Rift Valley). In the past, ethno-regional tensions have been
pronounced when presidential elections were contested by opposing alliances consisting of mutually
exclusive ethno-regional blocs. In 2022, however, the two newly founded broad alliances were cross-
ethnic in nature. That mediated against strong tensions and escalations.
Kenyatta and Odinga reorganised their parties (Jubilee and ODM / Orange Democratic Movement)
under the new Azimio la Umoja (One Kenya Alliance) coalition, with 23 larger and smaller parties
under its umbrella, and Ruto founded the United Democratic Alliance (UDA) as his party platform,
which became the dominant party under the Kenya Kwanza (Kenya First) alliance with 12 parties.
With Odinga from the Luo community in his fifth run for the presidency and Ruto from the Kalenjin
community in his first bid, there was no major Kikuyu presidential candidate on the ballot for the first
time in the multiparty era. The two main presidential candidates, however, both opted for Kikuyu
running mates, Odinga for the former Minister of Justice, Ms. Martha Karua, and Ruto for Mr. Rigathi
Gachagua, a businessperson. Mr. George Wajackoyah (Roots Party) and Mr. David Waihiga (Agano
Party) also competed.
With nine years of devolution experience, voters had become more interested in elective positions at
the county level, as local development and progress can be initiated there and that they do not depend
entirely on the presidency. Even without the preferred candidate winning the presidency, voters usually
saw some of their choices being elected to some of the elective positions, either at county or national
level. These factors reduced tensions, particularly around the presidential elections.
There was a strong and complex government peace and security architecture in place, anchored in the
2010 Constitution and developed against the backdrop of past post-election violence. It consisted of
several government commissions and institutions, including the National Cohesion and Integrity
Commission (NCIC), the Ethics and Anti-Corruption Commission (EACC), the Kenya National
Commission on Human Rights (KNHCR), the Independent Policy Oversight Authority (IPOA), the
Office of Administrative Justice (Ombudsman Office), and the Office of the Director of Public
Prosecution (ODPP). Funding gaps, and unclear delineations of mandates and responsibilities between
the organisations at times hampered effectiveness.
The state’s institutions peace work was supplemented by multi-level, non-state peace initiatives,
sometimes in cooperation with state institutions such as the NCIC. They include the National Council
of Churches of Kenya (NCCK), the private business peace initiative MKenya Daima, the Inter-
Religious Council, the UWIANO (Cohesion) Platform for Peace and others. In the absence of the
implementation of the Truth Justice and Reconciliation Commission (TJRC) report from 2013, which
would allow longstanding grievances and injustices to be addressed, these state and non-state peace
initiatives have established a strong peace narrative that helps bridge cleavages.
Despite a relatively high regard for the judiciary, the respect for rule of law by the executive and
legislature was not always consistent. This was demonstrated by a lack of will to enact several court
rulings on the two-thirds gender rule of the constitution and on the implementation of the Public
Benefits Organisations Act 2013, which leaves civil society organisations (CSOs) vulnerable to
arbitrary state interventions.
The backdrop to the general elections was marked by a deepening economic downturn, increased social
vulnerability, and rising poverty levels. Drastically soaring prices for fuel, staple food items, and
agricultural products in the wake of the COVID-19 pandemic and the Russian invasion of Ukraine had
contributed to the downturn.
IV. Implementation of Previous EOM Recommendations

The EU EOM 2017 offered 29 recommendations for the consideration of the national parliament, the
IEBC, the Office of the Registrar of Political Parties (ORPP), and other stakeholders. These
encompassed law reform and practical measures, ranging from changes to voter and candidate
registration to the conduct of the campaign, the administration of the election and the adjudication of
electoral disputes.
Issues of conflict with the international commitments of Kenya were raised in several
recommendations in 2017, recommendations which have been reiterated in substance in the
recommendations in this report. These include the removal of unreasonable restrictions on the right to
stand for election; narrowing the definition of hate speech; implementing the law on the regulation of
campaign finance; and improving political participation of women and persons with disabilities.
Areas where recommendations have been implemented include the enactment of the Data Protection
Act in 2019 and the subsequent appointment of the Data Protection Commissioner, thereby protecting
the right to privacy. Amendments to the Political Parties Act in 2022 (although less than a year to the
election) introduced new rules for the conduct of party primaries, which greatly reduced violence
during this electoral cycle. The IEBC also made significant efforts to implement the constitutional
provisions on the advancement of women, persons with disabilities and other minority groups during
candidate registration. Rules on the enforcement of the IEBC Electoral Code of Conduct changed, due
to a court ruling, leading to renewed cooperation between the IEBC and the ODPP in this domain. The
ODPP also pursued the prosecution of electoral offences with great expedition, with the most serious
electoral offences charged within days of commission.
Many of the previous recommendations in the space of electoral administration and political finance,
campaign environment and media remain to be implemented and have been reiterated, where
appropriate in context of these elections in the recommendations table at the end of the report.
V. Legal Framework
A very good constitutional framework exists, providing for democracy, rule of law and the protection
of political rights; some legal principles, however, await elaboration in legislation or through
enforcement of prevailing law.
A. International Principles and Commitments
Kenya is a state party to many of the principal international human rights treaties. These include the
International Covenant on Civil and Political Rights (1972); the Convention on the Elimination of All
Forms of Discrimination against Women (1984); the Convention on the Elimination of Racial
Discrimination (2001); the Convention on the Rights of the Child (1990); the Convention on the Rights
of Persons with Disabilities (CRPD) (2008); and the UN Convention against Corruption (2003). Kenya
has not acceded to the International Labour Organisation (ILO) Indigenous and Tribal People’s
Convention, nor to the Convention on the Reduction of Statelessness.
As a member state of the African Union (AU), Kenya has ratified many regional human rights treaties,
including the African Charter on Human and Peoples’ Rights (1992); the Protocol to the African
Charter on Human and Peoples’ Rights on the Rights of Women in Africa (2010); the AU Convention
on Preventing and Combatting Corruption (2007); and the African Youth Charter (2014). Kenya
ratified the African Charter on Democracy, Elections and Governance in 2021, having signed the
instrument in 2008. Kenya has ratified the Disability Protocol to the African Charter on Human and
Peoples’ Rights (2001), which has yet to enter into force.
The Constitution provides that international instruments that have been ratified form part of domestic
law and can be enforced locally. While few of these instruments have yet been relied upon in litigation,
the symbolic effect of this constitutional provision is to elevate international law to a high status. There
are some gaps, however, between the international commitments and their execution and enforcement
domestically, notably with regard to the CRPD, which has yet to be reflected in legislation.
B. Constitutional Rights
The Constitution of 2010 introduced a total renewal of the electoral system, enshrining extensive
political rights and requiring a new legislative framework for elections. An ever-growing body of
jurisprudence, interpreting constitutional rights in a progressive manner, is promoting the realisation
of civil and political rights, rule of law and democracy. Not all of the principles enshrined in the
Constitution have been fulfilled, however, in the absence of implementing legislation.
Protection is provided for the exercise of political rights, including the right to vote and the right to
stand for election, as well as the freedoms of association, assembly, and expression. The rights to vote
and to stand are subject to restrictions which are, prima facie, contrary to international commitments.
Varying across the six elections, these include qualifications based on intellectual capacity, adjudged
bankruptcy, and completion of tertiary education.
C. Electoral Legislation
There was little progress in electoral law reform. Elections (Amendment) Bills were introduced to
parliament in 2021 and in 2022, followed by a set of five draft Elections Regulations, focusing on the
registration of voters, voter education, party nominations, party lists and the use of technology in the
transmission of results. Their purpose was to bring clarity to important procedural aspects of the
elections, but none was enacted by parliament, despite the urging of necessity by the IEBC. Attempts
by the IEBC to resuscitate the Campaign Financing Act, and regulate campaign financing and
expenditure, were also thwarted by parliament. The enactment of the Data Protection Act, 2019, and
the Political Parties (Amendment) Act, 2022, represented the only progress in law reform, with the
latter having contributed to more peaceful political party nomination processes.
D. Electoral System and Constituency Delimitation
Direct elections take place to elect the president and the governors of the 47 counties. The president
must win more than half of all votes cast, as well as at least 25 per cent of the votes cast in at least 24
counties. Elections to the national parliament (NA and Senate) and to the county assemblies are mixed,
with most members directly elected in first-past-the-post contests in single member constituencies, and
others, representatives of special interest groups, nominated by political parties. Forty-seven women
are directly elected to reserved seats in the NA.
The IEBC must review boundaries of the 290 NA constituencies at intervals of between eight and
twelve years, at least twelve months before a general election. While the distribution of population
should be as nearly equal as possible, divergence by up to 40 per cent is constitutionally permitted,
dependent upon factors such as geography, means of communication, and the historical, economic,
and cultural ties of communities.
VI. Electoral Administration

Overall efficient election preparations despite prolonged litigation, but lack of transparency and a
divided and dysfunctional IEBC.
A. Structure and Composition of the Election Administration
Elections and referenda are conducted by the IEBC, a constitutionally independent permanent body
vested with significant authorities. The IEBC consists of a chairperson and six members, selected
through an open and competitive process, appointed by the president for a six-year non-renewable
term, following a prior approval by the NA.1 The public selection process and parliamentary oversight
of the final appointment guarantee impartiality and independence. The current IEBC comprises three
1
The IEBC Act stipulates that at least six months before the lapse of the term of the chairperson or member, or
within 14 days of the declaration of a vacancy, the president shall appoint a seven-member panel consisting of
four members nominated by the Parliamentary Service Commission, one by the Law Society of Kenya and two
by the Inter-religious Council of Kenya. which is mandated to shortlist and conduct public interviews of qualified
applicants. The panel submits its proposals to the president who makes the final selection, subject to NA approval.
commissioners, including the chairperson, who took oaths of office in January 2017 and four
commissioners who took oaths in September 2021.2
The law sets out the required qualifications of the chairperson and members of the commission.
Ineligibilities include having held office within the past five years or stood for election as a Member
of Parliament or a county assembly, and/or was a member of the governing body of a political party,
or who holds any state office.3 As provided by the law, during the shortlisting, nomination and
appointment process of the IEBC members, no more than two-thirds of the members should be of the
same gender while regional balance should also be ensured.
The IEBC commissioners generally enjoy security of tenure and may be removed from office only
for serious violations of the Constitution or any other law, gross misconduct, physical or mental
incapacity to perform their duties, incompetence, and bankruptcy. Removal of a member may be
initiated through a petition submitted to the NA, upon which the president is authorized to appoint a
tribunal to investigate, report on the facts and make a binding recommendation to the president.
The day-to-day management of operations is carried out by the Secretariat, headed by the Commission
Secretary, who is the chief executive officer (CS/CEO), appointed through a competitive process for
a five-year term, eligible for reappointment once. At the local level, the IEBC secretariat comprises 47
County Election Managers who administer elections at the county level (gubernatorial, senate, and
women representative elections), and serve as County Returning Officers (CROs) and 290
Constituency Election Coordinators, responsible to coordinate the Commission’s work at the
constituency level and serving as Constituency Returning Officers (ROs), with their administrative
assistants serving as Deputy Returning Officers (DRO). Their nominations were gazetted on 28 April
2022. The system of rotation of county and constituency returning officers before elections was
perceived by electoral stakeholders as ensuring impartiality of the election administration at lower
levels, while the permanence of their position leads to sufficient experience levels.
The IEBC Act provides that the IEBC financial independence is secured through a budget allocated
by the parliament and paid into a specialised fund. Regrettably, all EU EOM interlocutors pointed out
that during the last electoral cycle and until less than a year before the 2022 elections, the parliament
allocated only the necessary funds for the conduct of everyday operations while minimum resources
were allocated for programmes and activities necessary to be implemented on a continuously basis for
the IEBC’s successful discharge of its constitutional duties, such as voter education.
2
Following the resignation of one commissioner in October 2017 and three more in April 2018, the IEBC remained
with only three members until 2 September 2021. On 6 June 2018 a petition was filed against the IEBC (Isaiah
Biwott Kangwony v Independent Electoral & Boundaries Commission & another [2018] eKLR) asking the court
to declare, inter alia, that its current composition was illegal and unconstitutional as a result of the resignation of
four commissioners and hence the IEBC lacked the requisite quorum to carry out its business. The court dismissed
the petition and ruled that “the mere fact that there are vacancies in the Commission does not mean that the
Commission becomes unconstitutional […] and that the Commission still meets the minimum threshold of three
members as envisaged under Article 250(1) of the Constitution.”
3
The eligibility of one of the four commissioners was challenged in court on 31 August 2021 (Constitutional
Petition no. E345 OF 2021), on the grounds that she had participated in Jubilee party primaries in 2017 for the
position of Woman Representative, thus her independence was compromised. The court found that she was
ineligible for appointment as a commissioner since she stood for elections in 2017 and at the time of her
appointment a period of 5 years had not passed. However, the court found it prudent to allow her to remain in
office as a matter of greater public interest, as the IEBC was preparing for general elections and her removal from
office would have created a constitutional crisis.
Although the Act stipulates that the IEBC funds may also consist of grants, donations, or other
endowments, an official request by the government to the international community to support electoral
preparedness activities, following a shift from its late 2020 refusal to allow government institutions
such as IEBC to receive donor funding was only submitted in late January 2022. Hence, according to
representatives of the international community, the activities developed to support IEBC were largely
designed based on what was operationally possible to implement only a few months before elections.
E. Administration of Elections
The preparatory stages of the elections were managed effectively.4 The decentralised recruitment of
polling station members was conducted transparently and competitively. The IEBC demonstrated
competence in the delivery of cascade trainings for all levels of permanent and temporary election
officials according to the schedule. IEBC staff and participants reported that it was more intensive and
practice-oriented than in past elections. In some constituencies only one Kenya Integrated Election
Management System (KIEMS) kit was available per several participants during the training. EU EOM
observers noted that the training included simulation exercises and group work, however, limited focus
was placed on filling out the polling station statutory forms and the polling station diary, as well as on
the importance of correct packaging of sensitive materials. This was evidenced also by the fact that
several forms arrived incomplete at the constituency tallying centres and materials were often placed
in the wrong envelopes.
The dismissal by IEBC of 10 polling staff one day before polling5 for participating in meetings with
candidates, thereby potentially compromising their neutrality, demonstrated a proactive effort to
ensure the integrity of the process.
Recommendation: The training of electoral staff to be improved, especially of the presiding officers.
Consideration could be given to complement regular trainings before elections with a knowledge-
based online system for all IEBC staff.
Due to the centralised system of information flow, county and constituency returning officers met by
the EU EOM received consistent, regular, and timely information regarding electoral preparations, and
reported smooth cooperation with the IEBC at the central level. Stakeholders’ perception of
constituency and county officials’ information-sharing was positive overall. On the contrary, the IEBC
tended to be more reactive rather than proactive in its communication with external stakeholders, at
times only releasing crucial information on sensitive matters after problems had occurred. Although
the IEBC made daily use of its social media accounts and established a public information and call
centre as well as a press centre at the NTC, its website was lacking important information and was not
updated on a regular basis.
Recommendation: IEBC to strengthen its communication strategy to provide continuous,

comprehensive, and prompt information to election stakeholders and to undertake regular and
structured stakeholder consultation throughout the electoral cycle.
In the lead up to the polls, some important uncertainties impacted upon election day procedures,
although not as much as was expected. These uncertainties resulted from the IEBC’s desire to be
4
With a total cost of KSH 44.6 billion and an average cost of about KSH 2,000 per voter, the 2022 election is one
of the most expensive elections in the world. Given the high costs, IEBC could explore the cost-effectiveness of
electoral cycle activities, including more extensive audits and budget comparisons / analysis.
5
The six Presiding Officers, two Deputy Presiding Officers and two polling clerks were arrested by the Police and
dismissed from duty. They were from Ndhiwa, Webuye East and Webuye West Constituencies.
responsive to past criticism. Many of these decisions could have been considered sooner but were
instead dealt with as they emerged. Prolonged litigation - due to the absence of time limits to adjudicate
pre-election matters - including a court ruling issued less than 24 hours before the opening of the polls
significantly contributed to confusion.
In particular, in its interpretation of a 2017 appellate court decision6 and with the aim of mitigating
perceived misuse of the printed voter register, the IEBC decided to have printed voter registers in
security envelopes in polling stations and to use them only in case of total failure of the KIEMS kits.
This decision created controversy among electoral stakeholders and was also challenged in court,
resulting in a 4 August HC judgment,7 ordering the IEBC to reintroduce the use of the printed register
on election day. This judgment was then stayed by the appellate court on the eve of elections, thus
obliging the IEBC to change the instructions to its polling staff once again.8
Aiming to prevent electoral fraud, the IEBC also introduced additional safeguards for electronic voter
identification. Nevertheless, the efficiency of the KIEMS kits for voter identification and results
transmission remained a point of concern for most EU EOM interlocutors during the pre-election
period, although the IEBC reiterated its full confidence in the effective functioning of the system and
its security features. Overall, the IEBC could have done more to communicate more clearly with
stakeholders throughout the process to enhance public trust, although this improved in the last weeks.
The IEBC took certain steps, guided by the 2017 Supreme Court ruling,9 to modify polling station
presidential election results transmission, so as to fully base it on the originals of the results forms, and
to require only scanned results form (34A) images to be transmitted instead of sending them together
with manual KIEMS results entries. This decision, together with the paper trail, was viewed by the EU
EOM as a measure strengthening the integrity of the process.
The EU EOM notes that moving away from direct data entry into the KIEMS at polling station level
contributes to the verifiability and transparency of the process. In compliance with regulation 21 of the
Election Technology Regulation, the IEBC published the list of polling stations without 3G/4G
connectivity and took steps to address this problem in the 1,272 polling stations (PS) across the country
identified, via the use of satellite modems for the transmission of results; this was a tangible
improvement from the 2017 elections.10
F. Voter Education
The IEBC conducted voter information campaigns by directly engaging regional voter educators as
well as accrediting some 111 civil society organisations. In particular, the IEBC engaged two voter
educators per ward (so a total of 2900 ward-based voter educators), already during the two phases of
the enhanced voter registration activities, and a constituency voter educator in each of the 290
constituencies. However, the late dispersal of funds for voter education11 delayed the start of activities
6
National Super Alliance (NASA) Kenya v. IEBC & 2 Others, Court of Appeal, Decision No. 258/2017.
7
Kenya Human Rights Commission & 6 Others v. IEBC & 6 others, High Court, Petition No. E306/2022.
8
UDA v. Kenya Human Rights Commission & 12 Others, Court of Appeal ruling No. E288/2022.
9
Odinga and Another v. IEBC and 2 others, Supreme Court Decision no. 1/2017.
10
As IEBC conceded during the 2017 Supreme Court proceedings, on election date of 8 August 2017 it claimed it
was “unable” to transmit results from 11,000 polling stations and such inhibition set in place the use of a
complementary system of transmission of results envisaged under Section 44A of the Elections Act, which is in
essence the physical delivery of Forms 34A to the constituency tallying centres and hence the delay in the
declaration of results from those polling stations.
11
In the last four years after the 2017 elections, the IEBC Voter Education Department received KSH 30 million
per year which is largely insufficient for any voter education activities according to the department’s estimation.
and negatively impacted upon the capacity of IEBC staff on the ground, who had limited logistical
support and were unable to reach remote areas.
Despite efforts by the IEBC, CSOs and religious leaders, all EU EOM interlocutors considered that
voter education was insufficient. Even after important and combined efforts of all stakeholders in
promoting inclusivity and multiple platforms enhancing peaceful dialogue among stakeholders,
incidents of violence erupted between opponent parties and candidates in some areas. Although the
percentage of rejected (invalid) ballots amounted to 0.8 per cent, the high number of voters who asked
for assistance in understanding the complicated voting process with the six different ballot papers, also
influenced the speed of voting, reflecting insufficient voter education.
Priority recommendation: IEBC to be provided sufficient funds immediately after the end of the
current electoral cycle and throughout the next one, to implement continuous voter education
activities.
Most EU EOM interlocutors, however, acknowledged that the IEBC made concerted efforts to promote
the inclusion of marginalised groups, including by creating coordination committees to discuss their
concerns and take concrete actions.12 Through support by the Ministry of Education and in cooperation
with the Kenya University Student’s Association, the IEBC trained two trainers per university to
conduct voter education in campuses of the 78 universities of the country. Voter education materials
were also produced in Braille, demonstrating inclusivity.
VII. Electoral Technology

Although on election night results transmission worked well, the IEBC encountered difficulties to
communicate about KIEMS implementation which undermined transparency and thus public
confidence in the technology.
The Elections Act establishes the use of an integrated electronic electoral system that enables biometric
voter registration, electronic voter identification and electronic transmission of results. The act is
complemented with the Elections Technology Regulations, which has broad coverage, ranging from
the adoption and maintenance of the technological solutions, through its implementation, audit, to
security-related aspects and governance.13
In September 2021, the IEBC published the tender for the supply, delivery, installation, testing,
commissioning, support and maintenance of KIEMS.14 The new KIEMS would integrate the Biometric
Voter Registration (BVR), Electronic Voter Verification and Identification (EVV/EVI) and Results
Transmission (RTS) functionalities, replacing the legacy technological solutions made up of the old
BVR system, that had reached the end of its life, and the 2017 KIEMS software. Following a long and
controversial bidding process, the contract was awarded to Smartmatic International Holding B.V.15
The IEBC did not publish the evaluation either for this or the additional election technology related
public procurement processes, undermining transparency, and leaving room for speculation.
In the year before the 2022 elections the IEBC Voter Education Department received KSH 1.1 billion, but this
was too late for a strong positive impact on the electorate.
12
As of November 2021, the IECB established the Women Co-ordinating Committee, Disability Inclusion
Coordination Committee and Youth Co-ordinating Committee.
13
Amended section 44 of Elections Act 2011 and Elections (Technology) Regulations, 2017.
14
IEBC/OIT/001/21/2021/2022.
15
The award was contested at the Public Procurement Administrative Review Board (PPARB), the High Court, the
Constitutional Court, and the Court of Appeal.
The implementation of the new KIEMS suffered because of the appeals, deficient planning, and lack
of coordination between institutions. Of a total of 55,100 KIEMS kits, 41,000 were purchased in 2017
and still operational. The late contract award and the IEBC’s initial difficulties in getting the previous
KIEMS vendor’s collaboration to provide the necessary information so as to be able to reuse the old
kits provoked delays in preparing them for both the EVV and the EVI processes. The mismatch
between the RTS functionality described in the tender documentation and the legal framework resulted
in delays agreeing on the requirements and, consequently, the implementation of the RTS software.
The aforementioned issues conditioned the preparation of the KIEMS kits, which were launched a few
days later than scheduled. Based on the IEBC’s statements, the arrest of a vendor employee in
possession of stickers used for the kits identification negatively affected the logistics and deployment
plan. The lack of information-sharing, coordination and planning between all involved agencies
undermined public trust. In addition, the IEBC was responsible for providing the KIEMS backend
hosting infrastructure and the Results Public Website, but lack of planning meant that only a few days
before election day there was still no decision on where it would be hosted. Finally, a cloud-based
solution was chosen.
The IEBC contracted an audit of the Register of Voters (ROV) that included the assessment of the new
BVR system. The audit revealed severe information security vulnerabilities, such as a weak access
control to the system, meaning several users could access with the same account and had right to
perform more actions than they should.16 These vulnerabilities impede the accountability of
unauthorised operations, such as voters’ disenfranchisement, besides being a breach of data privacy
regulations. Although the IEBC implemented the corrective measures, the issue evidences the IEBC’s
lack of supervision during the BVR solution implementation and the deficient technological ownership
of the solution.
The IEBC contracted the mandatory audit of the election technology, which was launched a few weeks
before the election day. The scope covered both the acceptance of the KIEMS solution, and the security
and suitability of the IEBC data centres to receive and process the election results. Transferring the
responsibility for conducting the solution acceptance process to an audit firm does not release the IEBC
of its accountability, rather it limited its capacity to gain ownership of the technology. The lack of
information about the audit execution, its findings and the subsequent improvements undermined the
audit’s potential to enhance public confidence in the election technology and contributed to
speculations.
While party agents and stakeholders were given the opportunity to observe the assembling of the
KIEMS kits and the IEBC published information on the security and contingency measures
implemented in the KIEMS kits, no equivalent information was provided on the KIEMS backend
applications used by the Constituency Returning Officers (CRO) and the National Returning Officers
(NRO) nor on the hosting infrastructure, limiting stakeholders’ capacity to assess the election
technology.
Recommendation: IEBC to improve public communication on procurement processes and the

suitability and security of the election technology solutions notably with regard to voter registration
and identification and the results management processes.
16
IEBC/RFP/21/03/B/2021-2022/gk/bds/3 KPMG Final Audit Report 16 June 2022.
Two results transmission simulation exercises were held on 9 June and 19 July, respectively. While
several stakeholders recognised improvements following the second exercise, its small scale and
limited scope did not fully eliminate doubts on the efficiency and reliability of the system under real
conditions. The IEBC focused its communication efforts on reinforcing the message that the
transmission of the presidential results would be reliable, without addressing the questions raised by
several stakeholders about the performance of the system and the efficiency of the solution compared
with that of 2017, when besides the image the results were typed in and electronically transmitted.
Overall, on election day, the KIEMS kits served their purpose, ensuring the identification of the voters
either by fingerprint or searching their biographic information. The mission observed unequal
performance of the kits raising questions on the maintenance of the KIEMS kits and the solution’s
sustainability. The transmission of the images of the presidential results forms was constant, with more
than 80 percent of the images transmitted by midnight and exceeding the 99 per cent publication rate
24 hours after closing the polls, which contributed to the transparency of the counting process.
At the Constituency Tally Centres, the procedures were loosely followed resulting in many CRO not
using the KIEMS application to generate the Form 34B, preventing the traceability of the operations
carried out at that level and resulting in a variety of forms published at the Results Public Website.
The time needed to complete that process varied from one to three days, raising questions about the
efficiency, the integrity, and the transparency of the process. At the National Tallying Centre (NTC),
the KIEMS supported the verification of results by party agents. However, as no access was granted
to observe the results consolidation process following the verification, the EU EOM could not assess
the last step of the tallying.
Priority Recommendation: IEBC to implement improvements in election technology based on

regular audits and stakeholder consultations and to improve consistency and transparency.
VIII. Voter Registration

Despite IEBC’s efforts to boost voter registration, delayed funding impacted on its ability to
encourage young people to register, while a number of voter register shortcomings revealed in the
KPMG report were not sufficiently addressed prior to election day.
A. The Right to Vote
Kenya maintains an active voter registration system based on continuous voter registration (CVR),
which is carried out, as the Constitution and Elections Act stipulate, at all times in the constituencies
except for 60 days before the date of the general elections. Every citizen who has reached the age of
18 as proved by a national identity card or a passport, who has not been declared to be of “unsound
mind” and has not been convicted of an election offence during the preceding five years, has the right
to register as a voter upon application to the IEBC in the manner prescribed by the law. Due to the cut-
off period of 60 days, citizens who turn 18 in the last two months before election day are
disenfranchised.
B. Voter Registration Procedures
Voter registration is conducted using the KIEMS that stores biometric data, alphanumerical data, and
a picture of the applicant. In preparation for the upcoming elections, the IEBC conducted a national
Enhanced Continuous Voter Registration (ECVR) exercise in two phases, the first phase from 4
October to 5 November 2021, during which IEBC registered 1,519,294 new voters, and the second
phase from 17 January to 6 February 2022, during which the IEBC registered 1,031,645 voters.
Therefore, a total of 2,550,939 new voters were registered in both phases out of the projected target of
6 million voters (or four million as two million was shown not to possess ID cards). Upon completion
of the ECVR exercise, eligible citizens could continue to register in their constituencies until the legal
cut-off date.
Although the projected target was not reached, 74.82 per cent of citizens holding national identity
cards are included in the final voter register according to the IEBC. The gap in registration was
attributed mostly to lack of interest or defiance/expression of protest by the youth, as well as to
insufficient IEBC sensitisation especially prior to the first phase of the exercise, caused by delayed
and limited funding. Absence of specific efforts targeting youth, women and persons with disabilities
groups also had an impact. According to figures released by IEBC, 49,12 per cent of registered voters
are women, 39,84 per cent of registered voters are youth, and 18,98 per cent persons with disabilities.
Importantly, EU EOM interlocutors did not raise serious concerns either on the inclusivity of the
process or the quality of the voter register. Nevertheless, politically motivated transfers of voters were
mentioned to EU observers in some areas, while the IEBC itself investigated the illegal transfer of
voters in two counties. Trends of irregular transfers were also identified by the KPMG audit.
Recommendation: IEBC to enhance continuous voter registration efforts to target youth, women,
and persons with disabilities in order to ensure universal suffrage.
Registration for citizens residing outside Kenya to vote for the presidential election was conducted
from 6 to 20 December 2021. A total number of 10,443 diaspora voters were registered in 12 countries,
this figure being more than double the number of diaspora voters registered in 2017.17 Registration in
prisons was also conducted with 7,443 inmates registering to vote for the presidential election.
Although reg. 90 of the General (Elections) Regulations provides for the special voting of election
officials, observers, security forces on duty as well as nomadic pastoralists, hospitalised and
homebound voters, through procedures that the IEBC may adopt, no such provisions were made for
these (or previous) elections, thus de facto disenfranchising many voters.
The law provides that the IEBC may engage an external company to conduct an audit of the Register
of Voters (ROV) at least six months before a general election. For these elections, the audit was
conducted by KPMG, which carried out various tests on the ROV including comparison with the
National Registration Bureau (NRB), Department of Immigration Services (DIS) and Civil Registry
Services (CRS) to ascertain its accuracy. The preliminary audit report was submitted to the commission
on 2 June 2022, revealing a total of 481,711 registrations with duplicate or missing IDs, while 164,269
registrations with invalid reference number compared to the NRB and 4,757 voters who registered
with a passport and national ID. The total number of exceptions was 970,351 entries to the provisional
ROV.
The final audit report was submitted to the IEBC on 16 June. The final register of voters gazetted on
21 June 2022 totalled 22,120,458. Women accounted for 49.12 per cent and persons with disabilities
for 18.94 per cent. Although the general findings of the audit were released by IEBC on 20 June, the
actual report was only made public on 2 August, just a week prior to the elections. This undermined
the transparency of the process, even though the IEBC purported to have addressed shortcomings.
According to the final KPMG report, 70 per cent of those registered with duplicate or blank
IDs/passport were deregistered, as well as 59 per cent of those who registered with both an ID and a
passport number, and only 11 per cent of those registered with invalid document reference number.
The EU EOM analysis shows that in the final ROV there were still several exceptions (anomalies) to
17
USA, Germany, South Africa, Tanzania, Uganda, Rwanda, Burundi and South Sudan, United Kingdom (UK),
United Arab Emirates (UAE), Qatar, and Canada.
the KPMG report that were not taken into consideration by the IEBC. The estimated number of
exceptions within final ROV is 448,576 which constitutes 2 per cent of the published ROV.
To estimate the number of eligible voters, the 2019 census results were used.18 Additionally, the ward
delimitation (used for census purposes) is not the same as constituency delimitation (used for ROV
purposes) in every county. In fact, 182 constituencies (out of 290) are the same as wards throughout
the country. For the remaining 108 constituencies, there was no possibility to compare the ROV with
census data. According to EU EOM statistical analysis, 53 out of the 182 analysed constituencies have
voter registration rates over 100 per cent in comparison to the 2019 census data.19
Priority Recommendation: IEBC to introduce procedures for special voting of election officials,
security personnel on duty, homebound and hospitalised voters as well as nomadic pastoralists, to
vote, if necessary, in another location than the one they registered. This should be balanced against
security measures to avoid multiple voting.
IX. Registration of Candidates and Political Parties

Rules of access to stand for election, for individuals and political parties, were open; numbers of
women candidates were, however, extremely low, at around 11 per cent.
A. Registration of Candidates
The Constitution guarantees the right of every citizen to be a candidate for election to public office.
Candidates may be nominated by political parties or may stand as independents, provided they have
not been a member of a political party for at least three months preceding the election. Independent
presidential candidates require the support of 48,000 registered voters, including at least 2,000 voters
from at least 24 counties. Gubernatorial and senatorial candidates require 2,000 supporters, while NA
candidates require 1,000 and county assembly 500.
A group of independent presidential aspirants challenged the requirement that copies of the identity
cards of their supporters had to be submitted at registration. On 5 July 2022 the HC found the
requirement of the identity cards to be unconstitutional and quashed the relevant parts of the Elections
(General) Regulations, 2012. None of the plaintiffs in the case pursued their presidential aspirations
any further.
Candidates for parliamentary elections must be registered voters and citizens for at least 10 years.
Article 99 (2)(e) of the Constitution excludes “persons of unsound mind” from standing for election.
The UN Committee on the Rights of Persons with Disabilities, in its concluding observations on the
initial report of Kenya in 2015, stated “The Committee is concerned about restrictions to the right of
persons with disabilities to (…) stand for elections. (…) The Committee recommends that the State
18
For the estimation purposes, two main assumptions were made: a) a mortality rate at 6 per cent (according to
official statistics the real figure is 5.3 per cent, but it may be different for different years, thus the higher number
was taken), and b) an assumption that at least half of 15 year-old voters from 2019 would be able to register as
voters for the 2022 elections.
19
For instance, in Mt. Elgon constituency in Bungoma county, the official ROV includes 76,159 voters, however
the estimated number of voters from the 2019 census including 6 per cent mortality rate and that half of the 15
years old were eligible to register for the 2022 elections is 38,405 so 198.3 per cent estimated registration rate.
Likewise, in Lugari constituency-Kakamega county the estimated registration rate is 134.6 per cent, and in
Balambala constituency in Garissa county is 138.6 per cent, while in Kitui East in Kitui county is 234.1 per cent.
party: (a) Repeal constitutional provisions that restrict the right of persons with disabilities to be
elected as members of parliament”.20
A degree from a recognised university is prescribed in the Elections Act as a requirement of candidacy
for several elected offices. This was the subject of much litigation, principally where the genuine nature
of the documentation adduced by candidates was challenged. The requirement of a degree for
membership of county assemblies and of the NA was struck down, but the requirement stands for other
elections. Satisfaction of the educational qualification appears to be in contravention of the
authoritative interpretation of the International Covenant on Civil and Political Rights, which has
characterised educational requirements as unreasonable and discriminatory.21
Recommendation: Consideration could be given to remove restrictions on the right to stand for
elections for person with intellectual disabilities as well as restrictions requiring an educational
degree to stand.
Altogether there were about 16,098 candidates for 1,835 seats in all races, including 1,962 women.
Before the primaries the Ethics and Anti-Corruption Commission (EACC) had sent a list of aspirants
to the IEBC with a request to reject their candidacies. The list included 241 names with integrity issues,
including corruption cases, forgery of education certificates and late resignation as civil servant, that
according to the EACC suffice to disqualify them from contesting the elections. Of that list eventually
193 filed nomination papers of which only 3 were not cleared by the IEBC, as they had been previously
impeached from public offices.22 As the other aspirants had not exhausted their legal appeal options
the IEBC saw no legal grounds for not clearing them. 36 of the candidates from the EACC list ran for
governor (25 for Azimio, 11 for Kenya Kwanza).
B. Registration of Political Parties
The Constitution provides that every citizen is free to form a political party and to participate in the
activities of a political party. Political rights include the freedom to campaign for a party or a cause.
At present 89 political parties are registered with the Office of the Registrar of Political Parties (ORPP);
88 of these parties are standard parties, with individual members. The other, Azimio La Umoja, is a
coalition political party, an entity introduced by amendment to the Political Parties Act in February of
this year.
Not all registered political parties participated in the elections, as the deadline for the submission of
nomination rules to the IEBC was 18 October 2021. Parties were at liberty to use either direct or
indirect methods of candidate nomination, with oversight from the ORPP and the IEBC for
coordinating elements of the process. The IEBC cleared 83 political parties to nominate candidates for
the elections.
20
CRPD/C/KEN/CO/1, 30 September 2015.
21
CCPR General Comment 25, §15: “(…) Any restrictions on the right to stand for election (…) must be justifiable
on objective and reasonable criteria. Persons who are otherwise eligible to stand for election should not be
excluded by unreasonable or discriminatory requirements such as education (…)”.
22
These were: Mike Sonko, who was impeached as governor of Nairobi in 2019 and wanted to contest the
gubernatorial elections on a Wiper ticket in Mombasa; Chitavi Antony Mkhala, who was convicted for abuse of
office on three accounts in December 2011and wanted to contest the governor elections in Mombasa on a UDA
ticket; and Paul Karungo Thangwa who was impeached as Kiambu youth executive over alleged embezzlement
of funds and wanted to vie on UDA ticket the Kiambu Senator race.
X. Campaign Environment
Respect for fundamental rights and largely peaceful campaign with socio-economic issues profiled
but marred by lack of regulation of money in politics and inappropriate campaigning by some public
officials.
A. Election Campaign
EU EOM observers saw that contestants were able to campaign openly throughout the country to get
their messages out to voters and that the fundamental freedom of assembly was respected. Buttressed
by the peace narrative and the close campaign monitoring by a wide array of organisations, including
the NCIC, as well as the cross-ethnic nature of the two broad alliances, inter-ethnic tensions remained
low throughout.
Local competitions, however, between candidates from different alliances and between members of
the same alliance were often tensely fought with some resort to violence in some places like Kisii and
Uasin Gishu. No incidents of excessive use of force by the police were reported. Almost no cases of
‘derogatory’ language were observed by the EU EOM in campaign events.
The elections were characterised by a hybrid power constellation: both candidates enjoyed elements
of incumbency and of opposition. The Kenya Kwanza candidate, Ruto, was incumbent as deputy
president with direct access to state funding and government human resources. The Azimio la Umoja
candidate, Odinga, being endorsed by President Kenyatta enjoyed throughout the campaign the
controversial active support of several cabinet secretaries. At the same time, Ruto lost some executive
privileges through his falling out with the president, and Odinga had no formal representation within
the government.
The campaigns centred around socio-economic issues, with traditional ethno-regional appeals running
as undercurrents. In order to strengthen his appeal (e.g., among the Luhya in the western region) Ruto
enlisted the support of former government ministers, Musalia Mudavadi and Moses Wetangula,
whereas Odinga banked on the support of former Vice President Kalonzo Musyoka for the Kamba
vote from the Ukambani region.
Ruto ran his campaign as a contest between old entrenched political dynasties and himself with his
running mate, Gachagua, as self-made politicians and businesspeople, rising as ‘hustlers’ to national
leadership. Odinga emphasised his and his running mate’s credentials (Karua) in their struggle for a
multi-party system, their fight against corruption, and his previous experience as prime minister.
Odinga’s campaign was somewhat constrained by his alliance with President Kenyatta, being unable
to distance himself from state policies. Both candidates emphasised economic issues and promised
wide ranging economic reforms benefitting the large groups of disadvantaged and the vulnerable of
society. Ruto, with his ‘bottom-up-approach’, became more clearly identified with creating conditions
for upward socio-economic mobility.
Allegations of corruption against Ruto and Gachagua by the Azimio campaign and counter allegations
of ‘state capture’ through Kenyatta, Odinga and their trustees put forward by Kenya Kwanza dominated
the last weeks of the campaign. In end July 2022, Gachagua was ordered by the HC to pay 200 million
Kenyan shillings, that he had failed to account for, having two corruption indictments against him.
The last campaign weeks also saw an increasingly bitter and personal exchange between Ruto and
Kenyatta over their fallout that markedly heightened tensions between the two camps. Kenyatta did
not campaign together with Odinga or Karua, but controversially used his incumbency in public events
like opening of bridges and motorways to drum up support for the Azimio candidate. One day after the
official campaign’s end, President Kenyatta gave a TV/radio interview in Kikuyu, appealing to Mt.
Kenya region voters to unite, making clear that they should vote for Odinga, an apparent breach of the
campaign silence. This assessment was also echoed by EU EOM interlocutors, including analysts,
journalists and CSO members.
Candidates campaigned through a variety of methods. Presidential rallies throughout the country were
usually large scale, attracting large crowds. Often, after one presidential candidate visited a region, the
other followed within a day or two. Both main presidential candidates were often accompanied by their
alliance’s local candidates, seeking endorsement. This, at times, created stiff competition and internal
tensions amongst coalition parties arose in several places. Controversies about campaign venues were
either settled in coordination with the IEBC or by the contestants directly, except for UDA’s final
presidential rally in Nairobi’s Nyayo stadium, which was only granted after a HC injunction.
As a result of the success of the devolution of power since 2013 the gubernatorial and the MCA races
attracted a high interest amongst voters. Almost half of the governor seats were open, as incumbents
concluded their second and final terms. Whereas gubernatorial candidates also conducted some large-
scale rallies, campaigns for other races were smaller (election caravans, door-to-door, town hall
meetings).
Campaign regulation violations and misuse of incumbency were noted by EU EOM observers in
several cases and corroborated by media reports.23 EU EOM observers received reports of a Principal
Secretary as well as Chiefs and Assistant Chiefs organising support for the Azimio presidential
candidate locally, thereby contravening the legally required political neutrality for public officers.24
EU EOM observers also witnessed the misuse of state resources, including vehicles and facilities, by
campaigning officials25 and, in one case, the use of a public function for campaign purpose for Azimio
candidates.26
23
For example Capital FM news reported that the Director General of the Nairobi Metropolitan Service, Lt. General
Mohamed Badi, publicly supported Odinga, see: General Badi endorses Raila's State House bid, pledges to release
impounded boda bodas » Capital News (capitalfm.co.ke). Chief Administrative Secretary (CAS), Rachel Shebesh,
organised support for Odinga ahead of his rally in Kisumu, see: UPDATE: CAS Rachel Shebesh Leads Azimio
Campaigns in Kisumu Nightclubs Ahead of Raila's Rally Today - Opera News. Principal Secretary Interior
Karanja Kibicho was reported by media houses for campaigning for Odinga, for example see: Kibicho:
Intelligence data shows Raila will beat Ruto in round one (pd.co.ke)
24
Campaigning by chiefs and/or subchiefs was reported to EU EOM observers in the counties of Nyandarua,
Kirinyaga, Meru, Nyeri, Kisumu, Kisii, Nyamira, and Nairobi. Principal Secretary of the Interior Karanja Kibicho
was reported to EU EOM observers as actively campaigning and organising suppport for Azimio in Nyandarua
and Kirinyaga. In Kirinyaga, Nyeri and Muranga counties chiefs and subchiefs were threatened with transfer or
other sanctions if they do not comply with supporting Azimio, according to EU EOM observer interlocutors.
25
Nyandarua Governor Francis Kimemia (Jubilee/Azimio) admitted to EU EOM observers that his whole staff
including the press office are fully campaigning for him.EU EOM observers saw ODM gubernatorial candidate
in Kakamega, Barasa, using an official county government car for campaigning. EU EOM observers in Nakuru,
Kirinyaga, Nairobi, Kiambu, Trans Nzoia and Machakos received reports from multiple interlocutors about the
misuse of government vehicles and other assets (fuel, speakers, generators) for campaigning.
26
On July 20 an official event for the distribution of land title deeds in Kisumu was used by the county
commissioner, the deputy county commissioner, the governor and the speaker of the county assembly to campaign
for Azimio presidential candidate Odinga and to call for six-piece ODM vote on August 9 (i.e. all six ballots
should be cast for ODM). A similar incident was reported about Treasury Cabinet Secretary Ukur Yattani, who
used the issuance of 6,000 title deeds in Isiolo to drum up support for Odinga, see: CS Yatani issues 6,000 titles
to Isiolo residents (the-star.co.ke)
Cabinet Secretaries and members of the county executive committees are exempted from the campaign
ban,27 which contravenes the equality of opportunity principle of the UN Convention against
Corruption (article 19 of the CAC). EU EOM observers witnessed Cabinet Secretaries using this
controversial legal exemption, including CS for the Interior in Kisii and Nyamira and CS for Defence
in Western Kenya. Video footage also show their active support for Odinga.28 Media houses also
reported about other Cabinet Secretaries campaigning for the Azimio coalition.29 Civil Society
organisations took particular issue with the active support for Odinga by CS and his Principal
Secretary.30
Recommendation: Strengthen the ban on public resources being used for campaigning, including
by removing the exemption for cabinet secretaries and members of county executive committees.
B. Campaign Finance
The Election Campaign Financing Act has been in place since 2013, but no regulations to give effect
to the legislation have ever come into force. The intention of the legislation had been to create a regime
of regulation and accountability for funds raised and spent on election campaigns. The IEBC would
set spending limits and enforce compliance. To this end, the IEBC submitted campaign finance
regulations to the NA in 2016, but they were rejected. These regulations were resubmitted by the IEBC
in 2021, with the intention that they would apply to these elections. Despite publication in the Gazette
by the IEBC, the regulations were subsequently nullified in parliament.
In May 2022, the HC decided that the requirement that the regulations should be approved by the NA
was unconstitutional. This means that, for future elections, the IEBC will have the power to introduce
regulations without NA approval. This did not, however, ameliorate the situation for the 2022
elections, as there was insufficient time to draft new regulations and undertake the necessary public
consultation.
The absence of regulation of campaign finance is in apparent breach of the legal commitments of
Kenya, principally in the Convention against Corruption, which requires, in Article 7(3), that “each
State Party shall also consider taking appropriate legislative and administrative measures, consistent
with the objectives of this Convention and in accordance with the fundamental principles of its
domestic law, to enhance transparency in the funding of candidatures for elected public office…”
With this marked and continued absence of campaign finance regulations, there was widespread
payment of cash from a multiplicity of candidates to voters during the campaign. This is counter to
standards for democratic elections and has a corrupting influence on the voters. Candidates spent large
27
The Leadership and Integrity Act (section 23 (1)) specifically exempts cabinet secretaries and county executive
committee members from political neutrality requirements.
28
Am not ashamed to support Raila Odinga - CS Fred Matiangi | (thekenyanman.co.ke); (394) CS Wamalwa
launches Azimio lobby group - YouTube; further media reports add details to their support for Odinga: Gusiiland:
Will Matiang'i Retain Interior Ministry Post In Azimio Government? - Uzalendo News; List of Raila Odinga’s
Powerful Presidential Campaign Teams Spread Across the Country - Tuko.co.ke
29
This includes CS for Health, Mutahi Kagwe: CS Kagwe roots for Raila in Mt Kenya region | Nation; ICT CS Joe
Mucheru:, ICT CS Mucheru vows to ‘guard’ Raila’s votes - Nairobi News (nation.africa); CS Environment CS
Keriako Tobiko: Environment CS Tobiko hits out at DP Ruto (pd.co.ke)
30
10 Local Human Rights Groups Ask Matiangi, Kibicho To Keep Off Politics Ahead Of August 9 Polls - Shahidi
News : Shahidi News
sums of money in direct handouts to organisers, supporters, and event attendees, as observed by EU
EOM observers in 23 cases (11 involving Azimio parties and 12 Kenya Kwanza parties).31
As reported by a variety of interlocutors at national and county levels including campaign advisors on
both sides, the current economic hardship fuelled increased expectations among the electorate for
handouts. Also, the airing of commercials in traditional and/or social media and the use of super-large
posters by candidates, including sometimes those for MCA, indicated the unaccounted spiralling of
campaign spending.
The unregulated use and influence of money in politics has a negative impact on transparency and
accountability and creates incentives for corrupt political behaviour that distorts the playing field,
disadvantaging candidates with less funds, particularly, women, youth, and minority representatives
who have less opportunity to be fairly represented.
Priority Recommendation: IEBC to operationalise the Election Campaign Financing Act to

regulate the amount of money received and spent by candidates and political parties during an
election or referendum.
XI. Media
The media reported extensively on elections, with the focus on the presidential race, giving voters
an adequate level of information on which to base their choices.
A. Media Environment
The media landscape in Kenya is vibrant, with a spectrum of nearly 200 radio stations, 92 television
channels and 100 print publications licensed in the country.32 Radio plays a predominant role
countrywide, particularly in rural areas, where private vernacular radio stations broadcast in 19
community languages. Despite the variety of media outlets, a handful of major media houses with ties
to political and business interests controls most influential media groups. Dominant media
conglomerates include the Nation Media Group, the Standard Media Group, and Royal Media
Services.
During the pandemic, media houses introduced pay cuts or suspended certain work benefits
temporarily, leaving dozens of journalists jobless. In addition, generalised cutbacks in publicity
revenue forced media companies to downsize their newsrooms and lay off senior journalists. This
context damaged the quality of reporting and made the media sector more vulnerable and dependent
on state advertising.
The Media Council of Kenya (MCK) successfully engaged in a major effort to bridge with these
capacity gaps and increase the standards of media reporting. Throughout the pre-election and the
election periods, the MCK carried out trainings across the country which benefited over 3,500
journalists – more than a half of the estimated news community. These trainings provided media
practitioners with additional skills in ethical reporting, hate speech, debunking misinformation, and
conflict sensitive journalism, which contributed to an improvement in election reporting.
31
EU EOM observers witnessed distribution of money and/ or food items (rice, sugar, maize meal) in Malawa and
Lurambi (Kakamega county), Kisumu, Endebess (Trans Nzoia), Baringo Central (Baringo), Homa Bay, Kitutu
Chache (Kisii), Busia (Siaya), Machakos and Taita Taveta.
32
Media Sector Legislative Review 2021, Media Council of Kenya.
Although freedom of the press was generally respected, journalists continued to face attacks and press
associations admitted pressure by media owners over coverage of political topics. Credible journalists
acknowledged to the EU EOM that they are forced into self-censorship when covering corruption or
sensitive political issues.
During the campaign, the Political Journalists Association of Kenya (PJAK) expressed concerns over
an increase in the attacks on media practitioners and the profiling of journalists. This included
threatens, insults and journalists ejected from political meetings.33 Certain media owners publicly
supporting Azimio might have contributed to the alleged bias of the journalists on their staff.
Recommendation: Uphold the right to freedom of expression, including for media practitioners and
ensure that attacks against journalists are properly investigated and sanctioned.
B. Legal Framework for the Media
The legal framework provides an adequate basis for freedom of expression. However, the 2010
Constitution sets limits in the case of advocacy of hatred, which is defined as the “vilification of
others”, a terminology judged too broad by media advocates.34 Concerns were additionally raised with
the re-emergence of section 13 of Kenya’s 2008 National Cohesion and Integration Act, whose
definition of hate speech is also considered vague and would need amendment to promote freedom of
expression as part of the country’s international obligations.
Priority Recommendation: Detail the legal definition of hate speech in line with international
human rights obligations (so both intention to incitement and imminent violence are demonstrated).
In a positive development, the HC protected freedom of expression by declaring unconstitutional the

Penal Code provision on criminal defamation.35 Nevertheless, defamation and libel remain
misdemeanours that may carry jail term of up to two years.
The Access to Information Act 2016 entitles a framework for public entities and private bodies to
proactively disclose information and to provide information on request in line with the constitutional
principles of the right to access information. However, regulations are yet to be passed by the
government for its implementation.
The Communications Authority of Kenya is the state regulatory body tasked to regulate frequency
spectrum usage and licensing, while the state financed MCK prescribes standards for journalists,
ensures the protection of rights, and assess media compliance with the national legal framework.
C. Media and Elections
Election campaign started on 29 May and ended on 6 August, 48 hours before election day. The
Election Law regulates the access of political parties and candidates to state-owned media during the
campaign period, specifying that they shall be allocated reasonable airtime on public radio and
television.36 Failure to put in place regulations under the Election Campaigning Act (see section XI.B)
undermined limits on contributions through paid-up media coverage.
33
Video to PJAK’s public statement in Naivasha, 30 July 2022. During the election period, the Media Council of
Kenya reported various forms of harassment on at least 43 journalists.
34
Commentary on the Regulation of Hate Speech in Kenya, Article 19. Freedom of speech advocates defend that
not all hateful speech reaches the level of prohibition, and a clear threshold should be developed.
35
7 February 2017 HC ruling.
36
Elections Act, articles 41 and 108.
In a spirit of transparency, the IEBC, the Kenya Editors Guild (KEG) and the Kenya Union of
Journalists (KUJ) signed a Memorandum of Understanding in which they committed to an open-door
policy of sharing election information.
To ensure a credible and peaceful coverage, the MCK and media stakeholders abided to the ‘Guidelines
for Election Coverage 2022’ aimed at helping media practitioners to provide accurate coverage of
elections. Further, the IEBC, the KEG, and the KUJ adopted, on behalf of the Kenya Media Sector
Working Group (KMSWG),37 ‘The Media Guidelines on Elections Management Reporting 2022’, in
which a collaborative framework for election management reporting was established. These guidelines
included the delivery of media monitoring monthly reports by the MCK. However, in the last stage of
the campaign, the state regulatory body interrupted the publication of findings without further
clarifications. The journalist community claimed that these reports were never shared with them
beforehand.
Additionally, the MCK teamed up with the Media Owners Association (MOA) and the KEG to
organise broadcast debates at three levels: Nairobi governor, deputy presidential and presidential
candidates. The withdrawal of two candidates from the presidential debate made it less representative.
Further, given the 2+2 format initially envisaged, the two remaining presidential candidates, Ruto and
Waihiga, responded to questions separately, rather than in debate.
Notwithstanding, the presidential debate gave citizens the opportunity to evaluate some candidates’
performance and fact-checking was done by organisers and media on candidates’ statements. During
Ruto’s presentation, power cuts were noted in various parts of the country, as corroborated by EU
EOM observers. The national power company officially apologised on its website, but many EU EOM
interlocutors thought it was politically motivated.
For the first time and in a positive effort to enhance transparency, the IEBC provided a web portal for
political parties, observers, the media, and the public to access elections results in real-time.
Nevertheless, as election day approached, media interlocutors shared with the EU EOM concerns about
IEBC’s poor communication flow, particularly regarding the results transmission on election night.
National and international media conducted parallel tabulations based on the downloaded 34a and 34b
forms. However, the different approaches used to process the forms and the subsequent discrepancies
between their figures created confusion among voters.
Three days after the vote media outlets significantly slowed down their unofficial tallies, when around
80 per cent of the forms were processed. Main media groups claimed staff was exhausted and quality
controls were needed. The MCK denied speculations about instructions to stop this exercise, arguing
that a review and alignment of numbers was needed.
Divergent information on media tallying results continued to be published on the following days and
some international media acknowledged human errors in their exercise, which contributed to
allegations on Twitter regarding the hacking of media portals. Dis/misinformation on media
preliminary results populated the social media environment and some media published false news
without verification.38
37
The KMSWG comprises the major media players in Kenya. It was formed in 2017 to be a coalition of all like-
minded organizations to spearhead synergy in deliberating and getting solutions to emerging issues in the media
industry.
38
Namely a fake joint press release by Royal Media Services and the Standard media group on their own tallying
results stating Odinga’s leading in the presidential race.
Though the media unofficial tally emerged as an additional transparency tool during the tally process,
media houses delivered an incomplete exercise without providing timely clarifications.
D. Media Monitoring Findings
The media reported extensively on elections, with the focus on the presidential race, giving voters an
adequate level of information on which to base their choices. Media outlets allocated special
programmes and sections to elections, and electronic media organised debates at the gubernatorial,
women’s representatives, NA and senatorial levels.
The state-owned broadcaster KBC provided free airtime for political parties and candidates during the
campaign in line to international standards for equitable treatment and access to the public media,
although no criteria for slots’ allocation was established (see annex 1, graphic 1). With regards to the
presidential race, the Media Monitoring Unit (MMU) findings revealed that private TV channels
provided more airtime to Kenya Kwanza candidate Ruto, mainly due to the debate. When combining
all six levels of elections, private TV channels allocated a more balanced coverage to both main
political coalitions. Citizen TV reported more extensively on elections, followed by NTV and KTN
(see annex 1, graphic 2).39
This trend was followed in newspapers with a slightly more space allocation to the Azimio coalition.
Dailies reported on elections with a general neutral tone mostly through news and opinion articles. The
Kenya Kwanza coalition got more critical coverage (see annex 1, graphic 3).
Radio remains the most widespread source of information for most voters. Vernacular radio stations
showed preference to one or other major camp in terms of coverage. This was most notable on Ramogi
FM, the Luo community radio, which promoted Azimio’s presidential candidate in terms of time, and
Kass FM, the Kalenjin community radio, with a significantly allocation of time and positive tone for
the Kenya Kwanza’s candidate (see annex 1, graphics 4 and 5).
The gubernatorial race was the second most covered election. TV and print outlets allocated most of
the coverage to both UDA and Jubilee’s candidates, while on radio stations ODM’s prevailed (see
annex 1, graphics 6 and 7).
The Azimio coalition conducted a more comprehensive paid advertisement campaign across the media,
particularly concentrated in the presidential ticket. The number of TV advertisements from both
presidential front runners tripled during the last week of the campaign, while the more frequent
television paid airtime came from UDA’s candidates (see annex 1, graphics 8, 9 and 10).
IEBC’s communication efforts on voter education entailed a collaboration with national broadcasting
media outlets for live simulations on voting procedures. Broadcasted television and radio paid adverts
were less, due to a lack of financial means. The NCIC displayed a broad campaign on peaceful
elections through advertisements across the media.
39
The EU EOM media monitoring sample included the state-owned broadcaster KBC Channel 1 TV and KBC radio;
three private TV channels (NTV, KTN and Citizen TV); six private vernacular radios in six different community
languages (Citizen Radio – Kiswahili, Kameme FM – Kikuyu, Kass FM – Kalenjin, Mulembe FM – Luhya, Musyi
FM – Kamba and Ramogi FM – Luo) and the three main national daily newspapers Daily Nation, The Standard
and The Star, plus their online versions. Electronic media were monitored daily from 6 July to 6 August during
morning and evening prime time slots: from 6AM to 9AM and from 7PM to 10PM. The Facebook pages of the
media included in this sample were also monitored.
Of the total of candidates, 87.8 per cent were male and 12.2 per cent were female. The media provided
a gender-sensitive reporting and an adequate coverage of women candidates. TV channels, and radio
stations coverage was above the percentage (18 per cent), while on newspapers was slightly behind
(10.9 per cent) – see annex 1, graphics 11, 12 and 13.
Roots presidential candidate Wajackoyah was the protagonist of the most over-performing election
related posts in television channels and print Facebook pages in the first weeks of the campaign.40 In
the last stage, posts about final rallies performed significantly on newspapers and TV channels
Facebook pages, while on radio stations Facebook pages was Wajackoyah’s running mate’s statement
on Root’s candidate supporting Odinga.
XII. Social Media and Digital Communications

Extensive and sophisticated disinformation campaigns distorted the online political discourse,
contributing to the erosion of public trust in the electoral process and affecting voters’ ability to
make decisions free from manipulative interference.
A. Social Media Environment
There were some 12 million social media users in Kenya in 2021, active on multiple platforms.41 While
recent data show that Facebook, YouTube, Instagram and Twitter lead in popularity, TikTok became
the most downloaded app during the campaign, surpassing WhatsApp.
Parties and candidates extensively used different online platforms to campaign, setting up websites,
employing bloggers and influencers, as well as digital campaign managers to manage their accounts.
The EU EOM observed that the digital teams of the two main political camps used Twitter to shape
the narratives of the presidential campaign and the general political discourse online, while Facebook
was mainly used to address the local electorate. As a newcomer in the campaign environment, TikTok
offered an unexplored space where propagandistic and more divisive content could easily spread.
WhatsApp was broadly used to organise campaign activities, mobilise voters and share disinformation
through networks of groups. The main communication language on Facebook and Twitter was English;
vernacular languages appeared more popular on TikTok.
Hashtags were extensively used by parties to push narratives, create trending topics, criticize
opponents and spread disinformation campaigns. The EU social media team observed manufactured
amplification and coordination of messages online by fake accounts and malicious, bot-driven activity
in support of the presidential candidates. “Puppets for hire” contributed to amplify falsehood to a larger
audience throughout the election process.42
B. Legal Framework
Kenya’s constitution stipulates that freedom of expression does not extend to propaganda for war,
incitement to violence, hate speech, or advocacy of hatred. Section 77 of the Penal Code contains
further provisions on incitement to violence and disobedience of the law and the 2008 National
Cohesion and Integration Act established the NCIC to tame hate speech and promote national unity
and mitigation of ethnically motivated violence. In 2020, the Communications Authority of Kenya and
40
From 5 to 21 July. In Facebook analytics, a post overperforms when it gets more on-post interactions from the
page's audience than average, interactions meaning likes, comments and shares.
41
We Are Social, report Digital 2022: Kenya.
42
Sock puppet: an online account that uses a false identity designed specifically to deceive. Sock puppets are used
on social platforms to inflate another account’s follower numbers and to spread or amplify contents to a mass
audience.
the NCIC signed a Memorandum of Understanding to curb hate speech and misuse of online platforms.
As part of the measures to reduce inflammatory language, the NCIC launched Kenya’s National Action
Plan against Hate Speech earlier this year and scaled up capacity to monitor and deter it online during
the campaign, setting up a dedicated social media monitoring unit in cooperation with CSOs.
Some figures on the number and nature of the violations identified by the NCIC and its project partners
during the election process were made public before the election day. The scope of the definition of
‘hate speech’, as defined by law and as implemented by NCIC, is overly broad to be in line with
international standards in this regard (See Media Section). As well, NCIC’s role seems to have been
limited to purely a monitoring role, which is of questionable utility. The EU EOM noted that the NCIC
did not provide the public with regular social media monitoring reports. While it did not have a specific
legal obligation in this regard, releasing broader information may have helped publicly identify
incendiary online postings.
While fundamental freedoms are recognised and protected, the Computer Misuse and Cybercrimes
Act criminalises the spread of falsehoods and misinformation, falling short of regional and
international standards for freedom of expression.43 The vague prohibition of “false”, “misleading”
and “fictitious” data is highly subjective, and this law has been used to harass journalists, bloggers and
activists in the past.44
Priority Recommendation: Remove section 22 and 23 of the Computer Misuse and Cybercrimes Act
and explore less intrusive measures for addressing disinformation.
Beyond a general hate speech prohibition, the election law does not reflect the increased use and
specificity of social media. While a Data Protection Act was adopted in 2019 in line with previous EU
EOM recommendations, online advertising lacks adequate regulation and there was no code of conduct
covering parties’ online activities.
Recommendation: Include provisions on electoral advertising on online platforms, and explicitly

extend the existing campaigning rules to the digital sphere.
C. Social Media Monitoring (see also Annex II)
Concerns over the possible misuse of online platforms to amplify the reach of hate speech and
disinformation operations dominated the public discussion since long before the beginning of the
election campaign and in the period leading up to the elections CSOs made a real effort to monitor
social media to counter disinformation and raise public awareness. Meta, Twitter and TikTok
announced tailor-made measures to help ensure a safer digital space and provide voter information
during the electoral process and engaged in bilateral and multi-stakeholders’ conversations with the
43
The Joint Declaration on Freedom of expression and the Internet, 2011, paragraphs 1(a) states: “Freedom of
expression applies to the Internet, as it does to all means of communication. Restrictions on freedom of expression
on the Internet are only acceptable if they comply with established international standards […]”. Also see: African
Commission on Human and People's Rights Declaration of Principles on Freedom of Expression in Africa,
Principle 2 on “Non-interference with freedom of opinion” and Principle 5 on “Protection of the rights to freedom
of expression and access to information online”.
44
Joint declaration on freedom of expression and “fake news”, disinformation and propaganda by UN, OSCE, OAS
and ACHPR states that: “General prohibitions on the dissemination of information based on vague and ambiguous
ideas, including “false news” or “non-objective information”, are incompatible with international standards for
restrictions on freedom of expression […] should be abolished.” Also see UN, OSCE, OAS, ACHPR, Joint
declaration on Freedom of Expression and Elections in the Digital Age: “States should consider supporting
positive measures to address online disinformation, such as the promotion of independent fact-checking
mechanisms and public education campaigns, while avoiding adopting rules criminalizing disinformation.”
civil society and national institutions to ensure implementation.45 Despite that, all platforms were
largely misused to pollute the digital environment, with multiple tactics used to fuel information
disorder.
The online space was distorted by extensive and sophisticated disinformation campaigns targeting
multiple election stakeholders and largely seeking to undermine public trust in the electoral process
through repetitive attacks to the IEBC and its members. The EU EOM observed that different
techniques were used to distort the information ecosystem, including doctored newspaper front pages,
news alerts meant to discredit election officials, misleading opinion polls, inciteful or false quotes
deceitfully attributed to political aspirants, candidates falsely endorsing their competitors, manipulated
content announcing unverified election results or containing branding of established media outlets, as
well as elements of coordinated inauthentic behaviour, use of false accounts and bots.46 In a report
published after the elections, the MCK flagged seven influential figures for spreading disinformation
online during the campaign, also noting that some social media influencers created accounts in the
name of notable Kenyans to mislead the public.47
The EU EOM identified more than 300 misleading Facebook and Twitter profiles created in the months
leading up to the elections to deceive voters across the political spectrum, with multiple cases of use
of identical or slightly manipulated profile pictures. On Twitter, they were used to increase the reach
of deceptive claims during the campaign and after the elections by retweeting the same or substantially
similar content to artificially influence conversations in a coordinated manner. Many were suspended,
taken down, or changed their name into non-election-related subjects during the campaign, but the
great majority were still active at the time of writing, despite contravening the companies’ community
standard.48 These pages aimed at misleading the electorate, weakening its ability to discern sources of
political information to make a fully informed decision and thus adversely affecting the right to form
opinions on political matters free from manipulative interference.49
A number of international and national players exposed information manipulation. Fact-checking

organisations and civil society groups established partnerships to identify and investigate
misinformation around the elections and advocate to make tech companies more accountable and
promote effective content moderation practices. However, the EU EOM observed that on several
occasions different organisations debunked the same items multiple times. A better coordination
among fact-checkers would reinforce their capacity to timely dismantle false narratives and raise
public awareness. Ensuring that voters have access to fact-checked information is essential for election
integrity.50
45
The EU EOM repeated request for a meeting to Google Kenya and YouTube remained unanswered.
46
The EU EOM monitoring unit isolated 268 accounts created between June and August 2022 that shared
disinformation on election night and processed the Twitter handles with Botometer, a bot detector developed by
Indiana University, selecting those with a probability equal or greater than 80 per cent of being bots, and then
checked and labelled them manually. See: “Annex – Social Media Monitoring Results”.
47
Report on media performance during the 2022 General Election, Media Council of Kenya, 23 August.
48
Twitter community standards on manipulation and spam policy and Facebook Transparency Centre.
49
See United Nations (UN) Human Rights Committee General Comment No. 25 paragraph 19: “Voters should be
able to form opinions independently, free of violence or threat of violence, compulsion, inducement or
manipulative interference of any kind.”
50
Joint declaration on freedom of expression and “fake news”, disinformation and propaganda by UN, OSCE, OAS
and ACHPR states that: “All stakeholders – including intermediaries, media outlets, civil society and academia –
should be supported in developing participatory and transparent initiatives for creating a better understanding of
the impact of disinformation and propaganda on democracy, freedom of expression, journalism and civic space,
as well as appropriate responses to these phenomena.”
Recommendation: Develop a coordinated effort by the media, civil society and tech companies to
build a network to debunk disinformation in order to reinforce the existing fact-checking initiatives
in view of future elections.
A Council for Responsible Social Media brought together representatives from civil society and
eminent individuals before the elections. It demanded better accountability from big tech companies
in Kenya. It also encouraged the authorities to develop and publicly sign a self-regulatory Code of
Practice on Disinformation, to address threats to the integrity of the election process in line with
international good practice.51
Recommendation: Develop a self-regulatory Code of Practice on Disinformation by a multi-

stakeholder coalition against disinformation to promote greater transparency and accountability of
big tech companies.
Inciteful quotes falsely attributed to political aspirants were observed by the EU EOM social media
monitoring unit, with content often migrating among different online platforms and WhatsApp.
TikTok’s delayed response to violations of its policies on hateful behaviour and harmful
misinformation enabled the spread of incendiary rhetoric about the election through threats of ethnic
violence targeting certain communities. Several pieces of manipulated content containing
disinformation, inciteful language and digital forgeries widely circulated on the platform and content
evoking Kenya’s violent electoral past for political gain was observed.52 Women candidates were also
victims of harassment on social networks.
While the establishment of partnerships with local stakeholders enhanced the tech companies’ capacity
to identify harmful content and take action, their ability to conduct effective content moderation in
Swahili and other vernacular languages remains unclear.53 The EU EOM observed 101 instances of
inciteful messages and 24 gender based attacks published on multiple platforms during the campaign,
54 per cent of which were in Kiswahili, other local languages or mixed.54 More clarity around the
resources and techniques devoted to content moderation and a bigger effort to develop a network of
experts who speak the local languages and understand the context would improve their capacity to
safeguard the online space from detrimental behaviours.
XIII. Inclusion of Women, Persons with Disabilities

While important steps have been made to promote the inclusion of women and persons with
disabilities, more needs to be done to operationalise the provisions to make them meaningful.
Affirmative action was included in the Constitution, as part of the transformative agenda of the
document, comprising measures for “minorities and marginalised groups”, to ensure that they
“participate and are represented in governance and other spheres of life”. No more than two-thirds of
51
In October 2018, leading tech platforms voluntarily signed the Code of Practice on Disinformation, submitting
themselves to transparent self-regulation as laid out by the European Commission. A recent revision of this
initiative brought 34 signatories (including online platforms) to join the strengthened Code of Practice on
Disinformation, which sets more ambitious commitments and measures to counter online disinformation.
52
The content analysis of 62 TikTok accounts sharing political content using a mixed combination of 40 political
hashtags allowed the EU social media monitoring unit to identify 22 videos collectively viewed over 7 million
times as of 7th August that violated the platform community guidelines.
53
A study on Meta’s content moderation of political advertisements involving hate speech released a few days before
the elections exposed the company’s ineffective content moderation approach, despite previous announcements
of an increasing effort in this direction.
54
See: “Annex – Social Media Monitoring Results”.
the members of elected bodies should be of the same gender but no legislation has yet been enacted to
implement this two-thirds gender rule. A succession of judicial orders against parliament to implement
the principle has been ignored over the last decade, while efforts by the IEBC to mandate compliance
by political parties in this electoral cycle were judicially restrained at the instigation of the UDA.
Numbers of women elected in direct elections were very low,55 although reserved seats for women at
least maintained the current situation.56
The Constitution includes the principle that at least five per cent of the members of elected bodies
should be persons with disabilities (PwD), but this has not been given effect in law or in practice. A
PwD was successful in the HC this year in securing an order for special treatment in registration for
the presidential election, but the order was stayed by the Court of Appeal and appears unlikely to be
upheld when eventually heard by the Court of Appeal. An election petition challenging the results of
the presidential election was filed by the plaintiff in this case but was dismissed.
A small number of seats in parliament and in county assemblies are reserved57 for representatives of
PwD, nominated by political parties proportionate to their share of the vote. Certification of disability
status, by the National Council for Persons with Disabilities, was required by the IEBC for the first
time for these elections as a condition precedent to nomination. The Persons with Disabilities Act,
2003, which pre-dates ratification of the CRPD, addresses political rights only to the extent of
facilitation of access to voting, and requires amendment to protect the right to stand for election.
The Constitution also requires promotion of the parliamentary representation of ethnic minorities, but
no dedicated law has been enacted to achieve this. This strand of affirmative action has, instead, been
addressed through piecemeal amendment of electoral legislation, leaving the area under-regulated. The
only legislative definition of the beneficiaries is the vague formulation “a group that is not the dominant
one in a given society”. The Constitutional Implementation Oversight Committee introduced the
Special Interest Groups Laws (Amendment) Bill, 2019 to the NA in 2019, to impose obligations on
political parties and on the IEBC to adopt affirmative action for the target groups. It was not enacted.
The UN Human Rights Committee, in its concluding observations on the fourth periodic report of
Kenya58 stated that: “The Committee is concerned about: (a) The absence of dedicated legislation to
provide specific protections for indigenous peoples in the State party (…). The State party should: (a)
Develop and enact dedicated legislation to expand specific protection for indigenous peoples.” Kenya
noted the recommendation during the UPR that it should ratify the ILO Indigenous and Tribal Peoples
Convention, 1989 (No. 169).59
Recommendation: Enact constitutional affirmative action provisions. In particular: application of

the 2/3 gender principle; application of the principle that 5 per cent of members of elected bodies
are PwD; provisions on inclusion of marginalised ethnic minorities.
Members representing women, persons with disabilities (PwD), youth, workers, and minorities are
nominated by parties proportional to their vote share in parliament and in the county assemblies.
55
Women elected in direct elections: Senate – 6.4 per cent; County Assemblies – 7.9 per cent; NA – 13.3 per cent;
governor – 14.9 per cent.
56
The NA has 47 seats reserved for women; the Senate has 16; half of the representatives of special interest groups
and youth and PwD must be women; members are added to the MCA to comply with the 2/3 gender rule.
57
Two PwD are nominated to the Senate, a man and a woman; while PwD are among those qualified for nomination
to the 12 special seats in the NA & the 6 special seats in county assemblies
58
CCPR/C/KEN/CO/4, 11 May 2021
59
UN Human Rights Council 2020, Report of the Working Group on the UPR, 144.21.
Nevertheless, Kenyan elections are also considered an elite and male-dominated affair due to the high
costs associated with candidate nomination, which highly impacts the ability of women, youth, and
minority group representatives to be fairly represented.
The number of women candidates who contested the elections was very low overall, at around 11 per
cent. Several interlocutors highlighted the intimidating climate against women candidates in many
places. This assessment was corroborated by findings of EU EOM observers in several counties.
Physical harassment of female candidates was reported from Kirinyaga, Kilifi, Nyamira, and Embu,
and women candidates faced stigmatisation through ‘derogatory’ speech, verbal abuse, and threats in
some 14 counties.60
The elections saw a slight increase in the number of women elected, however the figure still falls far
short of the constitutionally required, but still un-legislated 2/3 gender rule. Seven women were elected
governors (4 more than last election), 3 Senators (no change) and 30 MPs (up from 23). In addition,
there are the 47 women representatives (1 per county). In the Senate, there are 16 nominated reserved
seats for women, plus the two special interest group seats that must go to women. Nevertheless, the
13th parliament is the third consecutive parliament that does not meet the 2/3 gender rule enshrined in
the Constitution. For the MCAs, there were 115 women candidates elected, representing a gradual
increase (96 previously). Seven counties did not elect any women candidate (12 in 2017).61
XIV. Citizen and International Election Observation

Extended presence of citizen observers throughout the country enhanced the transparency of the
process.
The main citizen election observation organisation, the Elections Observation Group (ELOG)
deployed 290 long-term observers countrywide, intermittently during 2021 and permanently since
March 2022, who monitored the enhanced voter registration exercise and party primaries, and in
addition 46 monitors of potential hotspots for violence.62 A week before the elections, ELOG launched
together with Ushahidi and 20 other organisations an Election Situation Room, in order to allow
citizens to report election-related incidents so as to enable rapid response and mitigate electoral
malpractices. For election day, ELOG had 5,108 short-term observers on the ground and conducted a
parallel vote tabulation (PVT) for the presidential race. Other organisations, such as the Youth
Empowerment and Development organisation deployed around 3,000 observers. Institutions, such as
the Kenya National Commission on Human Rights, deployed monitors nationwide to follow election-
related human rights issues, while the Independent Medico-Legal Unit monitored the situation since
March with 86 observers.
Several international observer missions were deployed for these elections. These included the African
Union together with the Common Market for Eastern and Southern Africa (COMESA), the
Commonwealth Secretariat, the East African Community (EAC), the Electoral Institute for Sustainable
60
These were Trans Nzoia, Kisii, Nyamira, Kisumu, Home Bay, Siaya, Baringo, Kilifi, Mombasa, Kiambu, Meru,
Embu, Tharaka Nithi and Machakos
61
These counties were Lamu, Garissa, Mandera, Marsabit, Samburu, Kajiado, and Nyamira. County assemblies,
however, become compliant with the two-thirds gender rule as per the Constitution (Art. 177/1b), which stipulates
that as many additional seats are to be created as necessary for compliance.
62
ELOG members include the Catholic Justice and Peace Commission, Center for Governance and Development,
Consortium for Empowerment and Development of Marginalized Communities, Constitution and Reform
Education Consortium, Institute for Education in Democracy, International Federation of Women Lawyers,
National Council of Churches of Kenya, Supreme Council of Kenya Muslims, United Disabled Persons of Kenya,
and Youth Agenda.
Democracy in Africa (EISA), the Intergovernmental Authority on Development (IGAD), and the
National Democratic Institute (NDI) / International Republican Institute (IRI) joint mission. The Carter
Center and the Westminster Foundation for Democracy (WFD) deployed expert missions, focusing on
specific issues. The EU EOM closely coordinated with all of them under the auspices of the
Declaration of Principles framework.
In addition to the requisite freedoms that were fully respected thus enabling the citizen and
international observers to conduct their duties, the IEBC established regular communication with all
observers through WhatsApp groups and information sharing through two official events.
XV. Electoral Justice

Access to remedies for electoral disputes was open and efficient; most hearings were conducted
publicly online.
The legal framework provides for effective access to legal remedies. Electoral dispute resolution is
highly regulated, with well-elaborated procedures in place, which functioned efficiently during this
electoral cycle. Judicial and quasi-judicial bodies were variously responsible for handling different
types of disputes, without any clashes between their respective areas of jurisdiction.
A. Political Party Nomination & Candidate Registration Disputes
The Political Parties Disputes Tribunal (PPDT) manages disputes both within and between political
parties. It is highly accessible, with an efficient system of electronic filing and hearing of cases. In
addition to permanent members, the appointment of 18 ad hoc members to deal with candidate
nomination disputes enabled decentralisation to seven locations. An amendment to the Political Parties
Act in 2022, which diminished the requirement of resort to internal party dispute resolution
mechanisms, facilitated greater access to the PPDT. Despite this change, the number of disputes on
party nominations brought before the PPDT reduced significantly, compared to 2017, due to new party
nomination rules. The number of party nomination disputes determined by the PPDT was 199 in 2022,
compared to 306 in 2017. All cases were processed within a month, as required by law. Fifty-eight
decisions of the PPDT were appealed to the HC.
Disputes relating to candidate registration were managed by the IEBC Dispute Resolution Committee
(DRC). This year, 325 cases were brought to the committee, all of which were determined within the
applicable 10-day time limit. The substance of the cases related to decisions of the IEBC to either
register or reject aspirant candidates for election, based on compliance with the statutory qualifications
determining eligibility to stand for election. More than forty decisions of the DRC were appealed to
the HC.
In a very positive development, hearings of the PPDT were open to all online. The decisions were
made available to the public in a compendium subsequent to the election. Hearings of the IEBC DRC
took place in person. The publication of decisions of both mechanisms, prior to the elections, would
increase the transparency of election dispute resolution.
Decisions of both the PPDT and the DRC could be appealed to the HC, at any location throughout the
country. Such cases were heard in an expedited manner, usually within a few days. There were,
however, no legal deadlines in place prescribing a date after which appeals could not be made to the
HC against the decisions of the PPDT and the DRC. Plaintiffs could delay for several weeks and then
lodge appeals close to election day. The consequence of this was that the IEBC could not be certain of
the final outcome of disputes. Printing of ballot papers was postponed in several local elections, due
to appeals on candidate registration continuing perilously close to the election. The IEBC attributed
misprinted ballot papers, which caused the postponement of eight elections, to litigation in some
instances.
B. Constitutional Petitions on Electoral Rights
Several cases were submitted to the HC seeking judicial review of actions of the IEBC, alleging
violations of constitutional rights. The rules set out in the Constitution allow any individual or
organisation to bring such a case, due either to a personal interest or in the public interest. This
jurisdiction of the HC is exercised through a constitutional bench, which was staffed by two judges for
most of the electoral cycle, although by just one for six days following the reassignment of judicial
personnel.
The subject matter of the election-related cases heard by the HC included the voting times of the
diaspora, the inclusion of photographs on ballot papers, the applicability of the two-thirds gender rule
to political party lists, the display of the voter registers outside polling stations, and the use of the
manual voter register in polling stations. These cases all took the form of standard constitutional
petitions. This meant that the expedition with which the above-mentioned election disputes were
handled did not apply.
The HC, instead, dealt with the matters in their standard course of business, but, nonetheless, relatively
expeditiously. However, the adjudication of the case on the manual voter register caused great
difficulties for the IEBC, as it was uncertain as to the procedures which would have to be followed
until the HC decided the case. A decision was delivered on 4 August, but this was followed, on 8
August, by an order from the Court of Appeal which restrained the enforcement of the HC order.
While the open rules of access to the courts to vindicate constitutional rights are extremely important,
there is clearly a need for some restraint on the filing of petitions dealing with areas which have
represented settled law for a long period in advance of an election. The subject matter of the manual
register case could have brought the matter before the courts long in advance of the election. The
introduction of a restraint, by way of an administrative practice direction for the HC, prohibiting
challenge to settled areas of electoral law within two months preceding an election, would assure
greater legal certainty for the conduct of elections. The jurisdiction of the HC must, however, remain
available to urgent applications when infringement of electoral rights is apprehended.
Recommendation: Consider imposition of a deadline on appeals to the HC against PPDT decisions

on candidate nomination and IEBC DRC decisions on candidate registration a d consider
introduction of a deadline to restrain cases against the IEBC to the HC Constitutional Bench in the
two-month period prior to election day. Exceptions should be permitted for urgent matters, where
the law is not settled before elections.
C. Election Offences
An extensive range of electoral offences is established by law, set out in the Election Offences Act,
2016, while many election offences also amount to breaches of the Penal Code. These include offences
related to the register of voters, to multiple registration and to voting. Impersonation, bribery, undue
influence, use of violence and misuse of public resources are also offences, as are attempts to procure,
aid or abet the commission of these acts. The Office of the Director of Public Prosecutions (ODPP)
published a compendium of election offences, with a useful explanation of the elements of all offences
included. A Code of Conduct for Political Parties, within the Political Parties Act, and the Electoral
Code of Conduct, within the IEBC Act, additionally prescribe expected standards of behaviour during
the campaign period.
The power to order investigations and to prosecute electoral offences lies with the ODPP. The IEBC
had, until April this year, enforced the Electoral Code of Conduct, but the HC determined this power
to be quasi-judicial in nature and ultra vires the IEBC, leaving the sole authority with the ODPP. The
IEBC provides information on offences to the ODPP, further to a Memorandum of Understanding
between the two institutions on 18 July this year. Proceedings must be commenced by the ODPP within
12 months of the election. Reported incidences of elections offences were few during the election
campaign, although that changed on election day, with one election-related killing that day and another
two days later, as well as some subsequent non-lethal attacks on IEBC officials. Allegations of election
offences were made in presidential election petitions but, at the time the EU EOM left the country, it
was premature to assess the outcome.
XVI. Polling, Counting and Tabulation of Results

Although polling and counting were well-administered, the magnitude of the tabulation exercise led
to a lengthy process, albeit professionally conducted.
A. Overview of Voting
For these elections, 46,229 polling stations were created across the country. This represents an increase
of 13,08 per cent compared to 2017, in accordance with the 12,79 per cent increase in the number of
the registered voters and the legal requirement of no more than 700 voters per polling station. A Greek
company was awarded the tender for the printing of 132,722,748 ballot papers for all elective positions.
There were eight security features on the ballot papers and five on the results sheets.
Voters exercised their democratic right to vote in a generally peaceful manner throughout election day.
The process was calm, albeit prolonged, with dedicated polling stations staff who had to administer
six different elections through complicated voter identification procedures.
Problems with KIEMS kits resulted in the use of printed voter register in about 238 polling stations in
two counties (in Kakamega and Makueni counties), according to IEBC official data and following
authorisation. Errors of ballot papers resulted in postponement of elections in four constituencies
(Pokot South, Kitui Rural, Rongai and Kacheliba constituencies) and two counties (Mombasa and
Kakamega).
Polling stations (PS) visited by EU EOM observers opened late in 28 out of 33 cases, due mostly to
the unpreparedness of staff and KIEMS kits not functioning; in only 5 of those PS visited the delay
was over 60 minutes. Party agents were present in all PS visited. The overall conduct of opening was
evaluated as “bad” or “very bad” in 4 out of 33 PS visited, a relatively high number. Female Presiding
Officers were observed in 13 out of 33 polling stations visited for the opening, and in 161 polling
stations out of 433 visited for voting.
During voting, the biometric voter identification was problematic in 35,1 per cent of the PS visited,
with KIEMS not being able to immediately verify fingerprints and having to verify alpha-numerically.
This led to the slow processing of voters and long queues. Positively, polling staff followed procedures
in 135 of the 151 PS where this issue was observed (out of the 433 PS visited). Inconsistent application
of the rules on the use of the printed (manual) voter register was noted in a small number of PS visited;
an appellate court ruling issued on the eve of voting had stayed a previous High Court judgment
instructing IEBC to use the printed register together with the electronic one, and instead ordered IEBC
to only use the printed voter register in case of complete KIEMS failure. This contributed to some
confusion among polling staff.
Checking of voters’ fingers for ink was not conducted in 111 out of the 433 of PS visited, which
constitutes an important procedural violation. The possibility, however, of the KIEMS kits to identify
whether a voter has already voted counterbalances this omission. The layout of 90 out of 433 PS visited
was not suitable for people with reduced mobility. Procedures related to assisted voting were not
followed in 60 of the 144 of the PS where assisted voting was observed, while inking the assistants’
finger was not conducted in 62 of these PS. Importantly, the legal provision requiring that when a voter
in need of assistance is not accompanied by a person of his/her choice, s/he will be assisted by the
Presiding Officer in the presence of the agents, significantly violates the secrecy of the vote.
Recommendation: More efforts to be undertaken for the inclusion of persons with disabilities in the
election day process. IEBC should consider making available a tactile ballot paper in each polling
station for voters with visual impairments, to be able to mark the ballot independently.
Secrecy of the vote was not fully protected in 185 of the 433 PS observed, due to PS layout.
Overcrowding was noted in 76 out of 433 PS visited, mostly due to the small classrooms, six ballot
boxes and many party agents, who were present in 432 out of 433 of PS visited. Citizen observers were
present in 191 of the 433 polling stations visited. Overall, the EU observers assessed the conduct of
voting as “good” or “very good” in 400 of the 433 PS observed.
Four EU EOM members of the core team were deployed to Kakamega and Mombasa counties on 28
August to observe the gubernatorial elections rescheduled for 29 August. The two teams observed
voting in 14 PS in total and noted very low voter turnout in all PS visited. No major problems were
observed during voting. Inconsistencies in the use of the printed voter register were noted, however,
in two PS in Mombasa County, where election staff were crossing out the names of voters identified
through KIEMS kits. The PS layout sufficiently protected the secrecy of the vote in only half of the
PS visited, as polling booths were placed open towards the room.
Recommendation: The IEBC to consider revising the layout of the polling stations to better
safeguard the secrecy of the vote.
B. Closing and Counting
More than half of PS observed at closing did not close on time, mostly due to late opening and voters
still waiting in line to vote. Required integrity checks during ballot reconciliation, such as not counting
unused ballots, were not followed in 9 out of 41 PS visited for counting, while spoiled ballots were not
counted in 10 out of 41 PS. Unused ballots were also not packed in tamperproof envelopes in 10 out
of 41 of the cases. These integrity measures are in place for the important reasons of preventing
possible tampering. Counting was conducted in a transparent manner, albeit slowly, and not always
according to the rules. Nevertheless, no official complaints were lodged at PS visited. Results forms
for the presidential race were not immediately displayed in half the PS observed, reducing
transparency.
The EU EOM observed the counting for 66 other elective posts. The transparency of the counting
process was evaluated as not good in 6 out of 66 as opposed to 3 out of the 41 of the counting for the
presidential election, and as bad in 3 out of 66 as opposed to one out of 41 of the counting for the
presidential election.
The overall conduct of closing was assessed as “very good” by both EU EOM teams deployed in
Kakamega and Mombasa counties on 29 August, while counting as “good” by one team and “very
good” by the other. Party agents and citizen observers were present in both PS. Contrary to the EU
EOM observations of the counting on 9 August, this time procedures related to ballot reconciliation
and security of the unused and spoiled ballots were duly followed. Likewise, completion of protocols
was done without problems. However, importantly, a copy of the PS results form was posted outside
of only one of the two PS observed, and copies were not given to party agents despite there being
sufficient copies available. They were instead instructed to take photos.
C. Tabulation of Results
EU EOM teams conducted 242 observations across 145 Constituency Tallying Centres (CTC). In the
direct vicinity of the CTCs, long queues of Presiding Officers (POs) were observed, who at times had
to wait for several hours to deliver their materials and the CTC locations with less yard space were
becoming increasingly chaotic. In 133 observations, all tamperproof envelopes arrived intact at the
CTC. During the intake process, mistakes such as tamperproof envelopes not been sealed upon arrival
of the POs at the CTC, presenting carbonated copies instead of the original, misplaced/unattended
ballot boxes, mathematical and clerical errors, were observed in CTCs visited by the EU teams.
As pressure to speed up the tally process increased during the second and third day of the tallying,
removing 34A forms from the tamper proof envelopes before reaching the CTC intake desk became a
common practice in all the CTCs observed. Although this aimed to increase the speed of the process,
it ultimately decreased the effectiveness of the administrative safeguards in place.
In numerous cases POs had sealed original presidential result form, 34A in the ballot boxes. The EU
teams observed that POs removed these forms from the ballot boxes while queuing outside the CTC
without a proper procedure. In some cases, it was observed and reported that similar situations were
handled according to the procedure – opening ballot boxes and resealing them inside CTC in presence
of agents while registering broken and replaced seal numbers.
Congestion was noted inside all CTCs observed, due to the large number of candidate agents,
supporters and observers from various local level observation groups, and this affected the possibility
to thoroughly observe the tallying process as well as the ability of the CTC staff to control whether
procedures were properly followed. The congestion was exacerbated in some CTCs by insufficiently
spacious premises. In 18 cases, there were unauthorised people inside the CTC, but only in 3 cases
they were interfering in the work of the staff. Whereas the overall assessment of the EU EOM observers
was that breaches of packaging of the sensitive material packaging were not conducted with an aim to
tamper with the results, they may have been interpreted as such by the supporters and agents present
in the vicinity of the CTC.
EU observers were not able to observe the process in 40 cases out of the 242 observations because of
the distance from the tables while in some CTC the layout was such that observers did not even have
visual access to the intake process. Displaying of the data entry and working method was not uniform
in the CTCs observed. Even though projectors and screens were installed in the majority of the CTC
visited, very few ROs opted to simultaneously project the figures being filled in the excel sheets. This
could have managed expectations as to the timeframe of announcement of results and generally would
enhance trust in the overall process, while it may also have prevented some incidents of tension at
CTCs. In addition, although the IEBC special software for entering the presidential results in the
system was made available, in the majority of the CTCs observed the staff was using the Excel
spreadsheet, that does not contribute to the security of the process. In 8 of 242 observations, no party
agents were present.
Overall, the most important challenges faced by the ROs, in addition to the magnitude of the tallying
exercise of six different elections, was the extreme fatigue after several days of tallying without rest,
the frequent reluctance of the security forces to enforce order and control the people present at the
CTC and the growing pressure exercised by candidates and their supporters to announce local level
MNA and MCA election results that at times resulted in tension inside and within immediate vicinity
of the CTCs (both in County and Constituency level tallying centres). However, the EU observers
noted that while having to deal with some party and candidate agents’ agitation and exhaustion of their
own IEBC staff, the constituency ROs handled the tallying process well and acted in a professional
manner.
At county level the tabulation process was much quicker and smoother, reflecting the fewer results
sheets they had to process. In all 32 observations in 23 County Tallying Centres visited by the EU
teams, the overall process was assessed as “good” or “very good”.
The development of an Independent Electoral and Boundaries Commission (IEBC) web portal that
allowed the public to access all presidential election polling station and constituency tallying centre
results forms in real time as they were being uploaded on the system, was a significant step towards
transparency.
Tabulation of the presidential results at the National Tallying Centre (NTC) started on 11 August, with
the arrival of the first original polling station results forms (34A) and of the CTC collation results
forms (34B). Strictly adhering to the procedure stipulated in reg.87(3) of the Elections (General)
Regulations, IEBC staff checked the results of each original 34A form (polling station presidential
results) against the results for each polling station tallied on 34B forms (CTC collations of presidential
election results). Copies of the 34A and 34B forms were distributed to party agents, sitting at tables,
and verifying data simultaneously with IEBC staff. Small arithmetical errors on the reconciliation parts
of the forms were identified in several forms, and either noted and corrected, if on 34B forms or noted
for the IEBC commissioners to take relevant action, if identified on 34A forms.
The slow pace of the process led, in the afternoon of 13 August when only 29.92 per cent of polling
stations results had been verified and 124 ROs were still waiting to handover forms, to a decision to
increase the number of staff as well as to add one ICT operator in every table to speed up the
verification. However, it was apparent that no specific detailed operational plan had been elaborated
in advance, as evidenced by the high number of ROs lining up for many days to deliver the sensitive
materials and the IEBC decision to optimise the use of ICT operators in order to meet the constitutional
deadline for announcement of results.
While at the NTC the results while being aggregated were initially being projected on the screens, on
13 August the projection stopped and only the results verified by constituency and county were being
shown. Since then and until the declaration of results on 15 August, no official information was
provided by the IEBC on the advancement of the process, the number of constituencies still left to be
verified, updated turnout figures, the number of stray and rejected ballots and the votes received per
candidate.
Priority recommendation: IEBC to introduce clear and detailed procedures for managing staff and
the flow of information at tallying centres to increase the consistency and efficiency of the process,
especially the public display of results being aggregated there, and the uploading of all results forms
to the public portal to ensure verifiability.
The EU EOM analysis of all the 290 constituency tallying centres results forms (forms 34B) revealed
errors in a small number of 34B forms. In particular, in 21 forms (or 7,24 per cent) the total number of
valid votes does not equal a sum of all candidates. The total difference between the officially
announced figures and the correct sum of votes for candidates is 4.023 votes and in 12 forms the
difference is greater than 100 votes. These inaccuracies were not of a scale to affect the outcome of
the presidential election. They indicate, nevertheless, procedural shortcomings of the results
aggregation process that could be attributed to human error rather than systemic manipulation.
Overall, party agents of the four presidential election candidates, as well as local observers, in all
observed tallying centres including the NTC, did not raise concerns about the transparency of the
processes, while they praised the cooperation with the IEBC officials. No formal complaints were filed
in any of the Constituency and County Tallying Centres observed. The work of the NTC staff was
characterised by a high level of transparency, being fully observed by party agents and observers who
had unrestricted access to the verification process; the EU EOM did not notice any original polling
station results forms (34A) being different from those uploaded on the IEBC portal and the copies
given to agents. This was also confirmed by the Supreme Court in its 5 September 2022 judgment,
which ruled that “no significant differences captured between the forms 34A uploaded on the public
portal and the physical forms 34A” delivered to the NTC that would have affected the overall outcome
of the presidential election.
XVII. Results and Post-Election Environment

The declaration of the presidential election results exposed deep fractures within the IEBC and
deeply affected public confidence in the process and in the institution itself.
A. Declaration of Results
The results were declared by the IEBC chairperson on 15 August. At the time of announcement, four
commissioners held a press conference at another location during which they announced that they
neither accepted nor rejected the results but instead “disowned” them, due to the opacity in the
compilation of the form 34C and the refusal of the IEBC chairperson to address some of their questions.
Nevertheless, they participated in the announcement of constituency presidential election results at the
NTC on several occasions (a point that was later noted by the SC.
The issue of whether the IEBC had the required quorum to finalise verification of form 34C and
announce the results was questioned in some of the petitions challenging the results. Notably, the court
had previously held that “the issue of the quorum of the commission only arises during the conduct of
the business of the commission” and that “even though tied to the commission’s membership, is not
per se an issue that should lead to a declaration that the commission is improperly constituted as
quorum will only be the subject of a challenge if the policy decisions of the commission are made
without the requisite quorum.”63
At the time of announcement, 28 polling station results forms (34A) were still not uploaded on the
IEBC public portal and results of 27 constituencies (forms 34B) had not been publicly announced; it
was not specified if they had been tallied and included in the form 34C or not. Whereas article 39
(1H)(2) of the Elections Act stipulates that the IEBC chairperson may declare a candidate elected
63
Isaiah Biwott Kangwony v. IEBC & another, High Court, Petition No. 212/2018, eKLR.
“before all the constituencies have transmitted their results if the commission is satisfied the results
that have not been received will not affect the result of the elections”, the small margin between the
two frontrunners required the tabulation of all constituencies before announcement. Notably, as the
IEBC chairperson stated in his sworn affidavit to the Supreme Court, the remaining 27 constituency
results were scheduled to be announced on 15 August, but after the security incident at the NTC with
supporters and agents of the Azimio camp storming the podium just before the declaration of results,
the chairperson decided to proceed with the declaration of presidential results immediately.
Following the declaration of presidential election results by the IEBC, ELOG presented the results of
its parallel vote tabulation (PVT) exercise. It was in line with the IEBC results, with Ruto obtaining
50.7 per cent of the votes and Odinga 48.7 per cent, with a margin of error of ±2.1 per cent. However,
the PVT could only confirm that the results were within the margin of error and could not confirm the
winner given the close result. In a second press conference, the four dissenting commissioners
advanced four arguments justifying their repudiation of the results. One was that the IEBC chairperson
appropriated an excess of power to himself in declaring results without the approval of a plenary of all
seven commissioners. They also alleged that there were mathematical errors in the results as declared,
and that the results declared were incomplete, lacking the numbers of registered voters, of the total
votes cast and of invalid votes. Their final point was that not all constituency results were included in
the final announcement.
B. Complaints Relating to the Election Results
Nine presidential election petitions were filed with the SC on 22 August 2022, challenging aspects
both of the conduct and of the results of the presidential election. An extremely condensed set of
deadlines applied for the hearing of these petitions, with 7 days for submission of petitions, followed
by 14 days for the SC to reach a decision. The petitioners had to serve the respondents within one day,
by 23 August, either in person or in newspapers advertisements. Respondents then had four days to
reply, to which petitioners had another day to file a further response.
The substance of the nine petitions was broad, ranging from allegations of treason on the part of the
IEBC Chairperson and flaws in the decision-making process of the IEBC, to flaws in voter registration,
in the use of technology, in tallying and verifying votes, and of fraud and hacking. The most significant
petition was that submitted by the Azimio candidates, Mr. Odinga and Ms. Karua. They alleged that
the IEBC did not conduct a proper, credible, or valid election and that, as a consequence, there should
be either a new result declared for the election or a nullification.
Similar grounds were adduced in six of the other election petitions. These cases were filed by a total
of 15 additional individuals, with those involved comprising voters, civil society activists, political
activists, and a member of the Senate.
Two petitions differed substantially from the other seven, those filed by Reuben Kigame and Moses
Kuria. Mr. Kigame had successfully challenged the decision of the IEBC not to register him as an
independent presidential candidate before the HC, an order subsequently restrained by the Court of
Appeal. His case centred on the alleged denial of his right to stand for election. The petition filed by
Mr. Kuria asserted that Mr. Odinga, through his agents, had interfered with the conduct of the national
tallying centre.
Twenty-three applications for interlocutory legal orders, precedent to the hearing of the petition, were
filed with the SC. The court responded individually to each of the interlocutory applications. A pre-
trial conference took place on 30 August in which the court consolidated seven of the petitions. The
original petition number five, that of Mr. Odinga and Ms. Karua, became the lead petition. The six
other, similar-fact petitions were consolidated with this petition. The court then framed a set of nine
issues to be addressed in hearing and determining the consolidated petition. The petitions of Mr.
Kigame and Mr. Kuria were dismissed.
Presidential Election Petition No. E005 of 2022 was then consolidated with presidential election
petitions E001, E002, E003, E004, E007 and E008. The hearing of the consolidated petition began on
31 August and lasted for three days. In addition to hearing both petitioners and respondents set out
their cases, the SC adduced questions for both sides to answer. Also, a scrutiny process took place over
48 hours, a report of which, the ICT Scrutiny and Inspection, Tallying and Recount Report, dated 1
September, was placed before the court and interrogated by the parties to the case. The court addressed
the nine issues raised for consideration, eight of substance and the ninth relating to the nature of the
orders which could be delivered in a ruling on a presidential election petition.
The decision of the SC was delivered on 5 September, with the detailed reasoning for the decision
published on 26 September. The seven judges of the court decided, unanimously, to dismiss the
consolidated election petition and to uphold the validity of the result of the election, as announced by
the IEBC on 15 August. Regulation 87(3) of the Elections (General) Regulations, 2012, was declared
to be unconstitutional, to the extent to which it purports to vest the power of verifying and tallying
presidential election results in the chairperson of the IEBC, to the exclusion of other members of the
commission. Each of the parties was ordered to bear their own costs.
The first issue addressed by the SC was whether, as recommended by the Kriegler Commission Report,
and enshrined in Article 86 of the Constitution and Section 44 of the Elections Act, “the technology
deployed by the IEBC for the conduct of the 2022 General Election met the standards of integrity,
verifiability, security, and transparency to guarantee accurate and verifiable results.” The petitioners
contended that the conduct of the election had not reached these standards, while the IEBC attempted
to rebut these contentions.
The SC reaffirmed the burden and standard of proof to be met in the case, drawing on previous
jurisprudence. It was the duty of the petitioners to prove their case, beyond reasonable doubt, through
proffering cogent and credible evidence to prove the grounds alleged to the satisfaction of the court. If
the grounds were established, then it would be for the respondents to rebut these. The analysis of the
evidence led the SC to determine that the IEBC rebutted complaints related to the audit of the voter
register.
The court was satisfied that inconsistencies and inaccuracies identified during the KPMG audit had
been successfully addressed. The broader technology procurement process had been within the law
and there was no abdication by the IEBC of its role in technological aspects of the election. It noted
that the IEBC had visibility and control of the electoral system at all times and there was no evidence
to show that there had been access to the system by unauthorised persons. It was held by the SC that
the technology deployed by the IEBC had not failed the standards of Article 86 of the constitution on
integrity, verifiability, security, and transparency.
The second issue addressed by the SC was whether there was interference with the uploading and
transmission of the Forms 34A from the polling stations to the IEBC public portal. Evidence was
adduced by the petitioners of differences in a sample of 41 polling stations; of 11,000 forms having
been dumped or uploaded in a batch, in order to “stage” results; and the claim was advanced that
greater fraud was perpetrated in 2022 than in 2017. The IEBC responded that the allegations of
compromise or intrusion by third parties had no factual or technical basis and details of the security
measures employed were provided.
The findings of the scrutiny report were relied upon by the SC to determine that there was no evidence
that there had been any interference in the transmission of the Forms 34A. There was also no evidence
that the IEBC had dumped 11,000 Forms 34A together at a particular time, and it noted that there was
no “staging” of results. The court determined that some of the logs submitted in evidence by the
petitioners were either from 2017 or from unknown or unauthenticated sources. No suspicious activity
was found in support of any of the allegations of the petitioners. Allegations of intrusion and
penetration of the results transmission system by foreigners were found by the court not to have been
proven.
The third issue addressed by the SC was whether there was a difference between the Forms 34A
uploaded on the IEBC’s public portal, the Forms 34A received at the NTC, and the Forms 34A issued
to the agents at the polling stations. The petitioners alleged that there was a systematic pattern of
criminal and fraudulent interference with the Forms 34A, the effect of which was the deduction of
votes from Odinga and their addition to Ruto. Forty-one polling stations were identified by the
petitioners as examples of the alleged practices, with an alleged shift in 2,793 votes. Recount of these
polling stations during the scrutiny exercise yielded no variance or discrepancy in the Forms 34A,
except for in four polling stations, where changes were insignificant, except one vote, twice, being
deducted from Mr. Odinga and attributed to Mr. Ruto, for a total difference of two votes.
All the Forms 34A submitted by the petitioners, and purportedly given to them by their agents from
polling stations, differed significantly from the originals, the certified copies and those in the public
portal. In this context, the SC mentioned the consequences of perjury, forgery, and the submission of
misleading or fabricated evidence. The result of the scrutiny exercise was that there was no evidence
of tampering with Forms 34A and the authenticity of the original forms sampled was confirmed. It
was found, overall, that, following the Maina Kiai precedent, the IEBC used the original physical
Forms 34A to tally, verify and declare the presidential election results.
The fourth issue considered by the SC was whether the postponement of gubernatorial elections in
Kakamega and Mombasa counties, and of six lower-level elections, resulted in voter suppression to
the detriment of Odinga and Karua. The allegation of the petitioners was that the postponement was a
deliberate effort to suppress votes in areas which had a history of overwhelming support for Odinga.
They asserted that the postponement was ultra vires the IEBC and unconstitutional as amounting to a
denial of the right to vote.
The SC determined that the IEBC has the constitutional authority to postpone elections and that the
postponements effected were justified, although due diligence in inspection of printing templates in
Greece could have prevented this occurrence. They did not find any evidence of intention to suppress
voter turnout. On the facts, the low turnout in the affected constituencies reflected turnout in
neighbouring constituencies and nationally, without any difference between them. The claim of low
turnout in the affected areas was dismissed as a “red herring”.
The fifth issue addressed by the SC was whether there were unexplainable discrepancies between the
votes cast for presidential candidates and other elective positions. The petitioners alleged that there
was a difference of 33,208 votes between the votes cast for president and those cast for other positions,
and that this differential indicated electoral fraud by way of systematic voter suppression in the
strongholds of the petitioners, and ballot stuffing in the strongholds of the respondents. The SC stated
that no document had been produced by the petitioners to prove systematic ballot stuffing.
The IEBC explained that voters in prison and in the diaspora vote only in the presidential election and
not for the other offices. Their analysis of voting across all 47 counties indicated that the discrepancy
between votes cast in the presidential election and the other election was 791 votes, not 33,208. This
small discrepancy was attributed to stray ballot papers. On the evidence, the SC held that there were
no inexplicable discrepancies between the votes cast for presidential candidates and other position.
The sixth issue addressed by the SC was whether the IEBC carried out the verification, tallying, and
declaration of results in accordance with Article 138(3)(c) and 138(10) of the constitution. The
petitioners argued that the chairperson undertook the tallying, verification, and declaration process to
the exclusion of four of the commissioners. The essence of this issue was whether there was anything
unconstitutional in the discharge of the respective roles of the chairperson and commissioners of the
IEBC during their activities at the NTC.
The SC held that, in general, all the powers and functions of the IEBC are vested in the commissioners
as a collective body and must be exercised by them, acting collectively. Any exclusive authority of the
chairperson must be expressly stated. The SC clarified that the responsibility of tallying and verifying
the results of a presidential election at the NTC vests in the IEBC as a collective entity, while the
responsibility of declaring the results vests exclusively in the chairperson. A chairperson cannot
exclude any member or members of the commission from the execution of tallying and verification.
Consequently, Regulation 87(3) of the Elections (General) Regulations, 2012, was found to be
unconstitutional, to the extent to which it purported to vest the power of verification and tallying of
presidential election results solely in the IEBC chairperson, to the exclusion of the other
commissioners.
On the facts of this particular case, the petitioners argued that the chairperson had excluded four of the
commissioners from the tallying and verification process. On reviewing the proceedings at the NTC,
the SC held that all of the commissioners were involved in activities related to the processing of results.
They announced results from several constituencies upon the conclusion of the tallying and verification
of the results. The court determined that, notwithstanding the divisions apparent between the
chairperson and four of the commissioners, the IEBC carried out the verification, tallying and
declaration of results in accordance with Article 138(3)(c) and (10) of the constitution.
The seventh issue addressed by the SC was whether the then declared president-elect attained 50 per
cent plus one of all the votes cast in accordance with Article 138(4) of the constitution. The exclusion
of rejected ballots from the calculation of results was contested, as was the turnout figure used in the
IEBC result calculations. The SC reiterated precedents which established that rejected votes cannot be
taken into account when calculating whether a presidential candidate has attained the 50 per cent plus
one of the votes cast.
The SC rejected the arguments of the petitioners that the 50 per cent of the votes validly cast, plus one
vote, threshold had not been met. The dispute turned first on establishment of the turnout figures to be
relied upon, with the alternate figures of 65.4 and 64.76 per cent in contention. It was decided by the
court that the turnout figure of 64.76 per cent was correct. It was further decided that the then declared
president-elect had attained 50 per cent plus one of all the valid votes cast, in accordance with Article
138(4) of the constitution.
The eighth issue considered by the SC was whether there were irregularities and illegalities of such a
magnitude as to affect the final result of the presidential election. The litany of deficiencies, of law and
practice, adduced by the petitioners included failures of technology and late opening of polling
stations; interference in the supply and delivery of ballot papers; lack of procedures for special voting;
alleged voter suppression; transposition anomalies; absence of party agents; and offences and ethical
breaches committed by the chairperson of the IEBC, among others. The court concluded that the
irregularities and illegalities cited by the petitioners were not proved to the standard required, or at all.
The court, citing the Harun Mwau case, repeated their previous statement that irregularities must be of
such a profound nature as to affect the actual result or the integrity of an election for the court to nullify
an election. A high, intermediate, standard of proof is applicable, requiring the presentation of cogent
and credible evidence. On an appraisal of the allegations, one by one, the court found that they were
not backed by the evidence needed to satisfy this requirement. In the specific case of the IEBC
chairperson, the court did not find any evidence of violation or breach of any electoral law or regulation
by the chairperson in the management of the presidential election.
The final ninth issue addressed by the court was the matter of what reliefs and orders the court had the
jurisdiction to grant. It was determined, as set out in the constitution and elaborated in the Supreme
Court (Presidential Election Petition) Rules, that the court has the mandate to determine the validity
or otherwise of the election of the president-elect, and, in the case of nullification, to order a fresh
election. Otherwise, jurisdiction does not extend to making any other orders. The court does have the
power to make recommendations or observations and has done so in the determination of previous
presidential election petitions.
Recommendations were offered regarding “the institutional dysfunctionality undermining the optimal
functioning of IEBC”, which included legal, policy and institutional reforms. On corporate
governance, it was recommended that parliament should consider enhancing the statutory and
regulatory framework on the separate policy and administrative remit of the IEBC. The IEBC ought
to introduce guidelines to clarify the distinct policy, strategic and oversight responsibilities of the
chairperson and the commissioners and to manage the separation of the administrative and policy
domains. The roles of the members of the commission and of the staff and other third parties should
be clearly delineated.
It was recommended, on election technology, that, to the extent possible, access to servers used for the
transmission and storage of results forms should be restricted to IEBC staff during the election period,
and such servers should be separate from those used for general administration. It was also
recommended that statutory Form 34A should be simplified and revised to include a record of stray
ballots. There should be thorough training for RO on the determination of valid votes. Special voting
should be brought into effect. As in 2017, the SC repeated the recommendation that the 14-day time
limit for hearing and determination of a presidential election petition should be extended.
The SC also made recommendations which were directed towards counsel involved in the proceedings.
Advocates were reminded of their obligations to preserve the dignity, honour, and ethics of the
profession and to refrain from attempting to influence pending decisions of the court. Language used
in commentary outside the court room by some advocates was characterised as inappropriate and
insulting towards the court, which could amount to professional misconduct.
C. Post-Election Developments
With 50.49 per cent (7,176,141) against Odinga’s 48.85 per cent (6,942,930) Ruto took about 69,000
votes above the 50 per cent plus 1 threshold. Odinga secured a majority in 27 of the 47 counties and
Ruto correspondingly in 20. This is broadly in line with the gubernatorial elections in which Azimio
initially garnered 24 of the 47 positions64 and the results for the NA which initially saw Kenya Kwanza
just slightly ahead of Azimio with 163 to162 members of parliament (MPs).65 Since the declaration of
Ruto as president-elect, the political landscape has started shifting. In the NA, 10 of the 12
Independents have joined Kenya Kwanza as have UDM (7 MPs) and 4 small parties with 1 seat each,
giving Kenya Kwanza a comfortable majority of 184 to 153 for Azimio.
Overall turnout was given as 65 per cent, a drop of 12.5 per cent. Stronger than average decreases were
particularly noted in areas with no presidential candidate, like the Mount Kenya region (a drop from
83 to 67 per cent), Ukambani (from 77 to 61 per cent) and Nairobi (72.7 to 56 per cent). The turnout
in Ruto’s stronghold in the Rift Valley66 decreased substantially less (from 79 per cent to 76 per cent)
than the national average, whereas the decrease in Odinga’s stronghold in the Nyanza region (83 per
cent to 73 per cent) was more marked than in Ruto’s area.67 With 77 per cent of the votes, Ruto could
keep a high level of support in the Mount Kenya region.68 In addition, securing 63 per cent in Bungoma
County (Western Kenya) Kenya Kwanza was able to turn around the result from the past election when
Odinga had received with 87 per cent of the votes. Odinga gained in Nairobi (from 51 per cent to 57
per cent), Mount Kenya (from 7 per cent to 22 per cent) and in Kisii-Nyanza (from 52 per cent to 65
per cent).
Following the declaration of the presidential election result by the IEBC chairperson and the press
conference of the four dissenting IEBC commissioners on 15 August, Azimio presidential candidate
Odinga held a press conference the next day at the Kenyatta International Conference Centre (KICC)
in which he rejected the presidential election result, pronounced a SC petition, and called upon his
supporters to remain calm and await the court proceedings and ruling. This press conference was
closely coordinated with a second press conference by the four dissenting IEBC commissioners and
large parts of their and Odinga’s statement were identical.
Ruto was gazetted as the president-elect on the evening of August 16, paving the way for the transition
process to formally kick-off. A constitutionally prescribed transition team, consisting of
representatives of various state institutions and – after the declaration of the presidential result – was
joined by three representatives of the Kenya Kwanza alliance, had been constituted about a month
before the elections.
After the declaration of the presidential election results few spontaneous protests with some destruction
of cars in Kisumu and Nairobi. The police acted proportionally and with restraint and on 16 August
the protests subsided. In separate incidents some acts of violence against Kikuyu took place in Kalenjin
dominated Uasin Gishu county where two houses were burnt down immediately after the elections and
immediately before the declaration of results. According to EU EOM observers Kikuyu, understood
this as a warning. In this situation, the police also moved in quickly and controlled the situation.
President Kenyatta remained quiet and appeared in public for the first time after the elections when on
18 August meeting a US congress delegation and religious leaders. He assured them of a smooth
64
These figures are before any defections occurred. This then included the two UDM governors (Mandera and
Marsabit in north-eastern), who since the declaration of the presidential election results have defected to Ruto’s
Kenya Kwanza alliance.
65
The NA comprises 290 seats and 47 women representatives. In this, 12 independents were elected.
66
These are the six counties predominantly inhabited by the Kalenjin: Uasin Gishu, Elgeyo Marakwet, Baringo,
Kericho, Nandi and Bomet.
67
This comprises four counties which are predominantly inhabited by Luo: Kisumu, Siayo, Migori and Homa Bay.
68
The wider Mount Kenya region consists of Kiambu, Muranga, Nyeri, Kirinyaga, Nyandarua, Embu, Meru,
Tharaka Nithi, Laikipia and Nakuru.
transition without giving any details. In a press statement, Mr. Odinga accepted the SC ruling although
he disagreed with it. On 13 September, Ruto was sworn in as the fifth president of the Republic of
Kenya in a festive public ceremony at Kasarani stadium in the presence of several heads of state.
Overall, the post-election environment remained calm and as during the election campaign the strong
polarisation between Kenyatta/Odinga on the one hand and Ruto on the other did not translate into
strong polarisation within society in the immediate post-election period.
XVIII. Recommendations
SUGGESTED RELEVANT
FR INTERNATIONAL /
CHANGE IN RESPONSIBLE
NO. page CONTEXT RECOMMENDATION REGIONAL
LEGAL INSTITUTION
# FRAMEWORK PRINCIPLE /
COMMITMENT
LEGAL FRAMEWORK
Women’s Participation;
The Constitution contains Equality between Men and
affirmative action provisions Enact constitutional Women; Freedom from
promoting the participation of affirmative action provisions. Discrimination
women, marginalised ethnic In particular: application of
Art. 4.1 CEDAW: Art.9
groups, and PwD. However, the 2/3 gender principle;
ACHPR Protocol on
1 p.33 successive parliaments have failed application of the principle New legislation. Parliament
Women ACDEG, Art. 3
to introduce the legislation that 5 per cent of members of
required to implement these elected bodies are PwD; Art.29 CRPD
provisions, leaving this provisions on inclusion of
ACHPR Protocol on PwD,
affirmative action largely marginalised ethnic minorities.
Art.21
unfulfilled.
ACHPR, Art.2
Consider imposition of a
The absence of a deadline for deadline on appeals to the HC
appeals to the HC against against PPDT decisions on
decisions of the PPDT and the candidate nomination and
IEBC DRC resulted in cases IEBC DRC decisions on
continuing throughout the candidate registration and
electoral cycle. consider introduction of a Right to Effective Remedy
Legislation. Parliament
2 p.36 Similarly, there are no deadlines deadline to restrain cases
Judiciary ACDEG, Art. 17
on constitutional petitions against the IEBC to the HC
challenging the lawfulness of Constitutional Bench in the
IEBC decisions, which resulted in two-month period prior to
cases on several aspects of election day. Exceptions
election procedures continuing should be permitted for urgent
until close to election day. matters, where the law is not
settled before elections.
Legislation:
Amend Art. 99 of Right and opportunity to
The right to stand for election is
the Constitution to participate in public
subject to restriction on the Consideration could be given to
remove “unsound affairs and hold office
grounds “of unsound mind”. remove restrictions on the right
mind”.
Persons without a degree are to stand for elections for person ICCPR, Art.25
3 p.20 Amend section 22 Parliament
disqualified from standing for with intellectual disabilities as
of the Elections Act CCPR General Comment
certain elections. well as restrictions requiring an
to remove the 25, §15
educational degree to stand.
requirement of a
degree to run for Art.29 CRPD
election.
ELECTORAL ADMINISTRATION
Stakeholders’ perception of
constituency and county officials’
information-sharing, however,
was positive overall. On the
contrary, the IEBC tended to be IEBC to strengthen its
more reactive rather than communication strategy to Transparency and access
proactive in its communication provide continuous, to information
with external stakeholders, at comprehensive, and prompt
No legal change
4 p.13 times only releasing crucial information to election IEBC ICCPR, Art.19.2
required.
information on sensitive matters stakeholders and to undertake
ICCPR, GC 34, Para. 19
after problems had occurred. regular and structured
Although the IEBC made daily stakeholder consultation UNCAC, Art.10
use of its social media accounts, throughout the electoral cycle.
its website was lacking important
information and was not updated
regularly.
No specific focus was placed on

Genuine Elections that
filling out the polling station
The training of electoral staff Guarantee the Free
statutory forms and the polling
to be improved, especially of Expression of the Will of
station diary; this was evidenced
the presiding officers. the Electors
also by the fact that several forms
Consideration could be given No legal change
5 p.13 arrived incomplete at the IEBC ICCPR GC 25, Para. 20
to complement regular required.
constituency tallying centres, as
trainings before elections with African Charter on
well as on the importance of
a knowledge-based online Democracy, Elections and
correct packaging of sensitive
system for all IEBC staff. Governance (ACDEG),
materials.
Art.17
The late dispersal of funds for voter

education delayed the start of
activities and negatively impacted IEBC to be provided sufficient Right of Access to
upon the capacity of IEBC staff on funds immediately after the end Information
the ground, who had limited of the current electoral cycle and
No legal change
6 p.15 logistical support and was unable to throughout the next one, to Parliament ICCPR, Art. 19.2
required
reach remote areas. Despite efforts implement continuous voter
ICCPR, GC 25, Art. 12
by the IEBC, CSOs and religious education activities.
leaders, all EU EOM interlocutors ACHPR, Art. 9.1
considered that voter education was
insufficient.
ELECTION TECHNOLOGY
IEBC to improve public Transparency and Access
communication on to Information
procurement processes and the
Conference of the States
Transparency needed throughout suitability and security of the
No legal change Parties to the United Nations
7 p.16 the election technology project election technology solutions IEBC
required. Convention against
lifecycle. notably with regard to voter
Corruption. Resolution 5/4,
registration and identification
Para. 22
and the results management
processes. ICCPR, GC 34, Para. 19
IEBC to implement Genuine Elections that
improvements in election Reflect the Will of the
Need for regular and agreed upon
technology based on regular No legal change People
8 p.17 improvement of the IEBC
audits and stakeholder required.
technological solutions. ICCPR, Art. 25(b)
consultations and to improve
consistency and transparency. ICCPR, GC 34, Para. 19
VOTER REGISTRATION
The gap in registration was

attributed mostly to youth apathy,
as well as to insufficient IEBC IEBC to enhance continuous
Right to Vote
sensitisation especially prior to the voter registration efforts to
first phase of the exercise, caused target youth, women and No legal change
9 p.18 IEBC UDHR, Art. 21.3
by delayed and limited funding. persons with disabilities in required.
Absence of specific efforts order to ensure universal
ICCPR, Art. 25 and GC 25
targeting youth, women and suffrage.
persons with disabilities groups
also had an impact.
Although reg.90 of the General
IEBC to introduce procedures
(Elections) Regulations provides
for special voting of election
for the special voting of election
officials, security personnel on
officials, observers, security
duty, homebound and
forces on duty as well as nomadic Right to Vote
hospitalised voters as well as
pastoralists, hospitalised and No legal change
10 p.19 nomadic pastoralists, to vote, if IEBC
homebound voters, through required. ICCPR, Art. 25 and GC 25
necessary, in another location
procedures that the IEBC may
than the one they registered.
adopt, no such provisions were
This should be balanced
made for these (or previous)
against security measures to
elections, thus de facto
avoid multiple voting.
disenfranchising many voters.
CAMPAIGN ENVIRONMENT
Unfair incumbency advantage:
Election Offences Act (sec.14) Remove
bans use of public resources for exemptions for
Strengthen the ban on public
electoral and campaign purposes cabinet secretaries
resources being used for
and bars government from and members of Equality of Opportunity
campaigning, including by
advertising its achievements county executive
removing the exemption for
11 p.23 during the election period. Yet, committees from Parliament UNCAC, Art. 19
cabinet secretaries and
the Leadership and Integrity Act the requirements of
members of county executive
(sec.23) allows cabinet political neutrality ACDEG, Art. 12.3
committees.
secretaries and members of (section 23,
county executive committees to Leadership and
campaign, blurring the line Integrity Act, 2012)
between state and party.
CAMPAIGN FINANCE
The Election Campaign Financing
Act, 2013, has never become
operational, as successive
parliaments have rejected subsidiary
Transparency in Funding
legislation proposed by the IEBC to The IEBC to operationalise the
and Free Choice of Voters
enforce it. Most recently, Election Campaign Financing Introduce new
regulations introduced by the IEBC Act to regulate the amount of regulations subsidiary
CPCC, Art. 10
12 p.24 in 2021 were rejected by the NA and money received and spent by to the Election IEBC
were de-gazetted in October 2021. candidates and political parties Campaign Financing
ICCPR GC 25, Para. 19
HC subsequently determined that during an election or Act, 2013.
the IEBC has the power to introduce referendum.
UNCAC, Art. 7.3
regulations without NA approval,
but the decision was too close to the
elections to permit the introduction
of same in time for these elections.
MEDIA
Although freedom of the press is
generally respected, journalists Kenya Police
continue to face attacks and press Uphold the right to freedom Service
associations admit pressure by of expression, including for
Freedom of Expression
media owners on coverage of media practitioners and No legal change Office of the
13 p.25
political topics. Credible ensure that attacks against required. Director of
ICCPR, Art. 19.2
journalists noted that they are journalists are properly Public
forced into self-censorship when investigated and sanctioned. Prosecutions
covering corruption or sensitive (ODPP)
political issues.
Concerns raised with the re- Detail the legal definition of
National Cohesion Freedom of Expression
emergence of sec.13 of the 2008 hate speech in line with National
14 p.26 and Integration Act,
National Cohesion and international human rights Assembly
2008, section 13. ICCPR GC 34, Para. 35
Integration Act; hate speech obligations (so both intention
definition is vague and would to incitement and imminent

need amendment to promote violence are demonstrated).
freedom of expression in line
with international obligations.
SOCIAL MEDIA
Freedom of Expression
African Commission on
Human and People’s
Rights Declarations of
Principles on Freedom of
While fundamental freedoms are
Expression in Africa:
recognized and protected, the
Principle 5
Computer Misuse and
Cybercrimes Act criminalizes the UN, OSCE, OAS, ACHPR,
spread of falsehoods and Joint Declaration on
Remove section 22 and 23 of
misinformation, falling short of Freedom of Expression and
the Computer Misuse and Amend Computer
regional and international Elections in the Digital
15 P.30 Cybercrimes Act and explore Misuse and Parliament
standards for freedom of Age, p.c (i)
less intrusive measures for Cybercrimes Act
expression. The vague prohibition
addressing disinformation. The Joint Declaration on
of “false”, “misleading” and
Freedom of expression and
“fictitious” data is highly
the Internet, 2011, General
subjective, and this law has been
principles, p.1(a); p.1(f).
used to harass journalists,
bloggers and activists. Joint Declaration on
Freedom of Expression and
“Fake News”
Disinformation and
Propaganda UN, OSCE,
OAS, ACHPR, p.3.e.
Beyond a general hate speech

prohibition, the election law does
Include provisions on electoral
not reflect the increased use and Transparency and Access
advertising on online
specificity of social media. […] Amend Elections to Information
16 p.30 platforms, and explicitly extend Parliament
online advertising lacks adequate Act
the existing campaigning rules
regulation and there was no code UNCAC, Art. 7.3
to the digital sphere.
of conduct covering parties’
online activities.
Fact-checking organizations and Right to Access to
civil society groups established Information
partnerships to identify and
Joint Declaration on
investigate misinformation Develop a coordinated effort by
Freedom of Expression and
around the elections. However, the media, civil society and
“Fake News”
the EU EOM observed that on tech companies to build a
Disinformation and
several occasions different network to debunk Public and private
17 p.31 Propaganda by UN, OSCE,
organizations debunked the same disinformation in order to media, civil
No legal change OAS, ACHPR, p.4.e., p. 6
items multiple times. A better reinforce the existing fact- society, tech
required.
coordination among fact- checking initiatives in view of platforms UN, OSCE, OAS, ACHPR,
checkers would reinforce their future elections. Joint declaration on
capacity to timely dismantle false Freedom of Expression and
narratives and raise public Elections in the Digital Age,
awareness. p.2 a (iv)
Freedom of Opinion and

A Council for Responsible Social Expression; Genuine
Media brought together Elections that Reflect the
representatives from civil society Free Expression of the
and eminent individuals a few Will of Voters; Right of
Develop a self-regulatory Code
days before the elections. It Access to Information
of Practice on Disinformation
demanded better accountability Communication
by a multi-stakeholder ICCPR GC 25, Para. 19
from big tech companies in Authority of
coalition against No legal change
18 p.32 Kenya. It also encouraged the Kenya, civil UN Guiding Principles on
disinformation to promote required.
authorities to develop and society, tech Business and Human Rights
greater transparency and
publicly sign a self-regulatory platforms (HRC, 2011), art. 13
accountability of big tech
Code of Practice on
companies. Joint declaration on freedom
Disinformation to address threats
to the integrity of the election of expression and “fake
process in line with international news” disinformation and
good practice. propaganda by UN, OSCE,
OAS, ACHPR, p.6
POLLING, COUNTING AND TABULATION
The layout of 90 out of 433 PS More efforts to be undertaken
visited was not suitable for people for the inclusion of persons
Participation in Political
with reduced mobility. with disabilities in the election
and Public Life
Importantly, the legal provision day process. IEBC should
requiring that when a voter in consider making available a Reg.72 of the Parliament Convention on the Rights of
19 p.38 need of assistance is not tactile ballot paper in each General (Elections) Persons with Disabilities
accompanied by a person of polling station for voters with Regulations. IEBC (CRPD), Art. 29 (a)
his/her choice, will be assisted by visual impairments, to be able
ICCPR, Article 25, HRC
the Presiding Officer in the to mark the ballot
GC 25, Para. 12
presence of the agents, independently.
significantly violates vote secrecy.
IEBC to consider revising the

The PS layout sufficiently Secrecy of the Ballot
layout of the polling stations to No legal change
20 p.38 protected the secrecy of the vote IEBC
better safeguard the secrecy of required. ICCPR, Article 25
in only half of the PS visited.
the vote.
Displaying of the data entry and
working method was not uniform
in the CTCs observed. In addition,
although the IEBC special
software for entering the
presidential results in the system
was made available, in the
majority of the CTCs observed the
staff was using the Excel IEBC to introduce clear and
detailed procedures for
spreadsheet, that does not
contribute to the security of the managing staff and the flow of Promotion of
information at tallying centres Transparency
process. While at the NTC the
results while being aggregated to increase the consistency and
efficiency of the process, No legal change AU Convention on
21 p.40 were initially being projected on IEBC
the screens, on 13 August the especially the public display of required. Preventing and Combating
results being aggregated there, Corruption, Art. 2.5
projection stopped and only the
results verified by constituency and the uploading of all results
forms to the public portal to UNCAC, Art. 7.4
and county were being shown.
Then, until the 15 August ensure verifiability.
declaration of results, no official
information was provided by the
IEBC on the advancement of the
process, the number of
constituencies left to be verified,
updated turnout figures, the
number of stray and rejected
ballots and votes per candidate.
XIX. ANNEXES
A. ANNEX I – EU EOM MEDIA MONITORING
The EU EOM conducted a qualitative and quantitative monitoring exercise of the election campaign
in print outlets and electronic media during morning and evening prime time slots from 6AM to 9AM
and from 7PM to 10PM. The monitoring period runs from 5 July to 6 August 2022, and from 4 July to
6 August 2022 for newspapers. The two-days period of electoral silence (7 and 8 August) was
monitored qualitatively.
The Media Monitoring Unit coded the time and measured the space allocated to election stakeholders
and political actors, including candidates and political parties, as well as the tone of the editorial
content. Paid advertising – from political parties, civil society organisations and state institutions –,
and gender balance in editorial content was also considered.
The sample is comprised by the state-owned broadcaster and a selection of private media with national
coverage, including five radio stations representing the major communities in the country (Kikuyu,
Luo, Luhya, Kamba and Kalenjin). The newspapers online versions and the Facebook pages of the
media sample were also monitored. Facebook pages were analysed with CrowdTangle to find trending
topics and content performing analysis.
The media sample included:

° The state-owned broadcaster KBC Channel 1 TV and KBC radio;
° Three private TV channels (NTV, KTN and Citizen TV);
° Six private vernacular radios (Citizen Radio – in Kiswahili, Kameme FM – in Kikuyu, Ramogi
FM – in Luo, Mulembe FM – in Luhya, Musyi FM – in Kamba and Kass FM – in Kalenjin);
° The tree main national daily newspapers (Daily Nation, The Standard and The Star);
® The total time monitored for radio and television was 2,178 hours;
® A total number of 3,720 newspaper pages and 986 online pages were monitored.
Graphic 1: Time and tone allocated to editorial content of political actors and election stakeholders
on the state-owned broadcaster – six levels of elections
08:24:00
07:12:00
06:00:00
04:48:00
03:36:00
02:24:00
01:12:00
00:00:00
IEBC
NCIC
IEBC
Jubilee
Religious Group
ODM
UDA
UDA
Government
Independent
President
Government
President
Civil Society
Agano Party
Roots
Agano Party
Kenya Kwanza
Kenya Kwanza
Wiper
Azimio
Azimio
KBC TV KBC Radio
Positive Negative Neutral
KBC Channel 1 KBC Radio

Total time 34h 58m 15h 09m
on private TV channels – six levels of elections
14:24:00
12:00:00
09:36:00
07:12:00
04:48:00
02:24:00
00:00:00
IEBC
IEBC
IEBC
NCIC
Jubilee
Jubilee
Jubilee
ODM
Independent
Regional government
UDA
UDA
UDA
President
President
Government
President
Civil Society
Agano Party
Roots
Agano Party
Roots
Agano Party
Kenya Kwanza
Kenya Kwanza
Kenya Kwanza
Wiper
Wiper
Azimio
Azimio
Azimio
Citizen TV KTN NTV
Citizen TV NTV KTN

Total time 49h 00m 38h 12m 31h 10m
in newspapers (space in cm2) – six levels of elections
70000
60000
50000
40000
30000
20000
10000
0
Jubilee
Jubilee
Jubilee
ODM
ODM
ODM
President
President
President
Kenya Kwanza
Kenya Kwanza
Kenya Kwanza
IEBC
IEBC
IEBC
Wiper
Wiper
Wiper
Azimio
Azimio
Azimio
UDA
Independent
UDA
Independent
UDA
Roots
Roots
Roots
The Nation The Standard The Star
* For information, a newspaper page measures around 1,064 cm2.

on vernacular radios (1) – six levels of elections
16:48:00
14:24:00
12:00:00
09:36:00
07:12:00
04:48:00
02:24:00
00:00:00
Jubilee
Jubilee
ODM
ODM
President
President
Civil Society
Agano Party
Tujibebe Wakenya Party

Kenya Kwanza
Kenya Kwanza
Kenya Kwanza
IEBC
IEBC
IEBC
Azimio
Azimio
Azimio
UDA
Government
Independent
UDA
Government
Independent
UDA
Roots
Roots
Citizen Radio Kameme FM Kass FM
Kass FM Citizen Radio Kameme FM

on vernacular radios (2) – six levels of elections
10:48:00
09:36:00
08:24:00
07:12:00
06:00:00
04:48:00
03:36:00
02:24:00
01:12:00
00:00:00
Maendeleo Chap Chap…
Movement for…
ODM
ODM
ODM
President
President
Agano Party
Agano Party
Kenya Kwanza
Kenya Kwanza
Kenya Kwanza
ANC
IEBC
IEBC
IEBC
Wiper
Jubilee
Jubilee
Jubilee
Azimio
Azimio
Azimio
Religious Group
FORD - K
UDA
Government
Independent
UDA
Independent
Independent
UDA
Mulembe FM Musyi FM Ramogi FM
Ramogi FM Musyi FM Mulembe FM

on TV channels – gubernatorial elections
Citizen TV KTN NTV KBC TV

Total time 09h 39m 05h 02m 04h 35m 04h 24m
on newspapers (space in cm2) – gubernatorial elections
Graphic 8: Paid advertisement on TV channels – six levels of elections
01:26:24
01:12:00
00:57:36
00:43:12
00:28:48
00:14:24
00:00:00
Civil Society
Civil Society
Kenya Kwanza
Kenya Kwanza
Kenya Kwanza
Kenya Kwanza
Safina
IEBC
IEBC
NCIC
IEBC
NCIC
IEBC
NCIC
Jubilee
Religious Group
Religious Group
David Mwaure
Religious Group
ODM
Independent
UDA
UDA
Independent
UDA
Raila Odinga
Azimio
Azimio
Citizen TV KTN KBC TV NTV
KTN Citizen TV NTV KBC TV

Total time 03h 34m 03h 30m 03h 04m 18m
Graphic 9: Paid-up advertisement on vernacular radios (1) – Six levels of elections
06:00:00
04:48:00
03:36:00
02:24:00
01:12:00
00:00:00 ODM
Muungano Party
Azimio
Azimio
Azimio
UDA
UDA
Jubilee
Independent
Wiper
Religious Group
President
Mulembe FM Musyi FM Ramogi FM
Ramogi FM Musyi FM Mulembe FM

Graphic 10: Paid-up advertisement on vernacular radios (2) – six levels of elections
01:40:48
01:26:24
01:12:00
00:57:36
00:43:12
00:28:48
00:14:24
00:00:00
International
Jubilee
Jubilee
Religious Group
Religious Group
Religious Group
UDA
Usawa Party
Kenya Kwanza
Kenya Kwanza
IEBC
IEBC
IEBC
Azimio
Azimio
Azimio
Azimio
community
Citizen Radio Kameme FM Kass FM KBC
Radio
Citizen Radio Kameme FM Kass FM KBC Radio

Total time 02h 32m 05m 03m
Graphic 11: Editorial content on TV channels allotted to candidates according to gender – six levels
of elections
43:12:00
38:24:00
33:36:00
28:48:00
24:00:00
19:12:00
14:24:00
09:36:00
04:48:00
00:00:00
Citizen TV KBC KTN NTV
Female Male
Male Female
Total time 91h 43m 17h 29m
Graphic 12: Editorial content on radio stations allotted to candidates according to gender – six levels
of elections
38:24:00
33:36:00
28:48:00
24:00:00
19:12:00
14:24:00
09:36:00
04:48:00
00:00:00
Citizen Radio Kameme FM KASS FM KBC Radio Mulembe FM Musyi FM Ramogi FM
Female Male
Male Female
Total Time 122h 10m 22h 32m
Graphic 13: Editorial content on newspapers allotted to candidates according to gender (space in
cm2) – six levels of elections
160000
140000
120000
100000
80000
60000
40000
20000
0
The Nation The Standard The Star
Male Female
Male Female
2
Total space 338,233 cm 37,037 cm2
B. ANNEX II – SOCIAL MEDIA MONITORING RESULTS

Internet penetration and social media use continue to increase in
the country, however there is a lack of consensus as to the total
number of Internet users, ranging from 12 to 42 million
depending on the source. Nearly 21 per cent of the population
use social media, with almost 12 million social media users.
Facebook is the leading platform, followed by YouTube,
Instagram and Twitter.69 TikTok is rapidly growing in
popularity, and it was the most downloaded app in August 2022.
Instant-messaging apps are also quite popular, with WhatsApp
leading in terms of use.
The EU EOM Social Media Monitoring Unit (SMMU) started its

monitoring activity by mapping the digital landscape in order to
assess the social media presence of gubernatorial and presidential
candidates and clarify the relevance of each platform as a tool to
campaign in the general elections. Different monitoring projects
were designed based on priority subject matters identified
through preliminary consultations with local stakeholders.
The EU EOM analysed the online election-related content with a particular focus on disinformation,
dissemination of inflammatory content and gender harassment. The social media monitoring activity
covered the political campaign, the election silence period, e-day and the post-election developments.
The selection of the sample was made according to multiple analytical purposes, later reflected in
different internal reports.
Subject Analytical approach Monitoring Period No. of Posts

Derogatory Quantitative 29/5 – 8/8 331
Speech
Disinformation Case-study 1/06 – 15/07 26
Gender analysis Qualitative 29/05 – 22/08 15 posts / 20
comments
Subject Analytical approach Monitoring Period No. of clips

Derogatory Quantitative 29/5 – 19/08 558
Speech
Disinformation Case-study 6/04 – 07/08 22
Gender analysis Qualitative 29/05 – 22/08 2
Subject Analytical approach Monitoring Period No. of Tweets

Derogatory Case-study 13 – 30 July 6
Speech
Disinformation Case-study 8 – 22 June 7,328
Bot-driven 8 – 22 June 55
amplification Case-study 9 – 10 August 35
13 – 16 August 30
69
Source: ‘We Are Social, report Digital 2022: Kenya.
The methodology used for producing the insights in the report relies on data sources that include:
• CrowdTangle for Facebook data scraping70;

• Brandwatch for Twitter data scraping;
• SentiOne for punctual analysis of trending topics or key-words;
• Botometer and TruthNest for bot-driven amplification analysis on Twitter;
• Facebook public transparency tool Ad Library;
• Tailor-made reporting forms for Facebook and TikTok content analysis.
Incitement and Derogatory Speech
The social media team established a hate speech lexicon to run searches on Facebook in order to
identify possible instances of derogatory speech. The list of keywords included the hate lex term
announced by the National Cohesion and Integration Commission (NCIC) in April 2022 to establish
popularly used words to be considered derogatory speech and which use could attract legal action.71
During the monitoring period the EU EOM identified 101 cases of inciteful language, 65 of which on
Facebook, 30 on TikTok and 6 on Twitter. The issues that were identified as having potential to trigger
violence include negative stereotypes along ethnic lines, emotive content that could instigate violence,
use of threatening or insulting words to incite the public, hate-filled speeches, especially at political
rallies, as well as claims discrediting the IEBC. Content evoking Kenya’s violent electoral past for
political gain was observed on Facebook,
Twitter and TikTok. Six tweets recalling the
2007/2008 post-election violence containing
content inciting one ethnic tribe against another
with a call to violence were published on 13th
July and taken down by Twitter some days
after. A TikTok video using the hate lex term
“Kihii” to refer to Odinga in a derogatory
manner was taken down by TikTok after it had
summed more than a million views.72
Most of the instances of derogatory speech
identified included elements of ethnicity. A
breakdown of the cases analysed is summarized
in Figures 2-5:
70
727 Facebook Pages and 106 public groups were selected and entered in CrowdTangle for data export. A sample
of 331 Facebook posts was subjected to the analysis of incitement and derogatory speech. Posts generating the
highest number of interactions from Facebook Pages belonging to a sample of women candidates were selected
for manual review of comments as part of the gender analysis component. The pages with lower level of activity
or interaction were object of the social media monitoring in different ways (manual monitoring of comments,
manual checking, case studies).
71
Hate speech lexicon, as defined by the NCIC.
72
The word “Kihii” was included in the hate speech lexicon of the NCIC.
Figure 3. Facebook posts containing elements of derogatory speech by topic. Base: 65 cases.
Figure 4. TikTok videos containing elements of derogatory speech by topic Base: 30 cases.
Figure 5. Twitter videos containing elements of derogatory speech by topic. Base: 6 cases.
In general, the main communication language on Facebook and Twitter was English whereas
vernacular languages appeared more popular on TikTok. However, local languages were clearly more
frequent in inciteful claims. A breakdown of the cases of derogatory speech by language shows that
Swahili and other vernacular languages were used in 32 per cent of Facebook posts, 90 per cent of
TikTok clips and all the tweets analysed.
Figure 6. Main languages of derogatory speech content by platform
C. Gender Analysis
The EU EOM identified 20 elements of gender harassment against 7 women candidates on Facebook
and two against one candidate on Instagram, with insulting words used to demean them. All the attacks
were surfaced in comments to posts published by the victims, who were object of appearance-based
remarks (7), sexualized comments meant to humiliate them through images and memes (10 on
Facebook and 2 on Instagram), criticism around family responsibilities (2) and use of sexist language
(1). Two cases of appearance-based attacks against female public figures were also identified on
TikTok.
Figure 7. Cases of gender-based attacks by platform
While English was the main language in almost half of the Facebook instances analysed, Figure 8
shows that Swahili was also frequently used on this platform and prevailed on others.
Figure 8. Main languages of gender-based attacks by platform
D. Bot-Driven Amplification
The EU EOM SMMU identified 2,838 accounts heavily engaged in disseminating early unverified
results in the eighteen hours following the closing of the polling stations on 9th August. Among these,
65 were created in June, 119 in July and 84 in August 2022.
Figure 9. Profile analysis
The analysis of the Twitter handles of 268 accounts created since June with the bot detector Botometer
classified 62 of them with a probability equal or greater than 80 per cent of being bots. Of these, 12
were created in June, 19 in July, and 39 in August. A manual review applying further filters (such as
profile picture, posting frequency and alpha-numerical nickname) showed that at least 17 accounts
presented strong patterns of inauthenticity. A few were deactivated by Twitter in the following days.
Figure 10. Manual verification
Thirty-five out of 268 accounts showed similar characteristics, such as being fully dedicated to share
campaign content, recent date of creation (8 in June, 10 in July and 17 in August), lack of background
picture, same layout with similar profile picture, alpha-numerical address (62 per cent). Ten of these
accounts were created on election day or the day after.
Figure 11. Troll farm
The analysis of the tweet frequency showed that 30 of these accounts published more than 150
times/day, with the highest performing account publishing virtually every two minutes.
Figure 12. Tweet frequency

The chart below shows the average number of tweets per day published by the 30 accounts publishing
more than 150 times a day.
Figure 13. Average number of tweets per day
The EU EOM social media team observed that this troll farm engaged in a bot-like promotion of
narratives to increase the reach of deceptive claims by retweeting identical or substantially similar
content to artificially influence conversations in a coordinated manner, which is at odds with the
Twitter community standards on manipulation and spam policy.

Eu Eom Kenya 2022 en

Uploaded by

Copyright:

Available Formats

Eu Eom Kenya 2022 en

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Eu Eom Kenya 2022 en

Uploaded by

Copyright:

Available Formats

European Union

Election Observation Mission

I. Executive Summary ........................................................................................................................ 6

A. Political Party Nomination & Candidate Registration Disputes ........................................... 34

ODPP Office of the Director of Public Prosecution

III. Political Background

IV. Implementation of Previous EOM Recommendations

VI. Electoral Administration

Recommendation: IEBC to strengthen its communication strategy to provide continuous,

VII. Electoral Technology

Recommendation: IEBC to improve public communication on procurement processes and the

Priority Recommendation: IEBC to implement improvements in election technology based on

VIII. Voter Registration

IX. Registration of Candidates and Political Parties

Priority Recommendation: IEBC to operationalise the Election Campaign Financing Act to

In a positive development, the HC protected freedom of expression by declaring unconstitutional the

XII. Social Media and Digital Communications

Recommendation: Include provisions on electoral advertising on online platforms, and explicitly

A number of international and national players exposed information manipulation. Fact-checking

Recommendation: Develop a self-regulatory Code of Practice on Disinformation by a multi-

XIII. Inclusion of Women, Persons with Disabilities

Recommendation: Enact constitutional affirmative action provisions. In particular: application of

XIV. Citizen and International Election Observation

XV. Electoral Justice

Recommendation: Consider imposition of a deadline on appeals to the HC against PPDT decisions

XVI. Polling, Counting and Tabulation of Results

XVII. Results and Post-Election Environment

No specific focus was placed on

The late dispersal of funds for voter

The gap in registration was

definition is vague and would to incitement and imminent

Beyond a general hate speech

Freedom of Opinion and

IEBC to consider revising the

The media sample included:

KBC TV KBC Radio

Positive Negative Neutral

KBC Channel 1 KBC Radio

Positive Negative Neutral

Citizen TV NTV KTN

The Nation The Standard The Star

Positive Negative Neutral

* For information, a newspaper page measures around 1,064 cm2.

Tujibebe Wakenya Party

Positive Negative Neutral

Kass FM Citizen Radio Kameme FM

Mulembe FM Musyi FM Ramogi FM

Positive Negative Neutral

Ramogi FM Musyi FM Mulembe FM

Citizen TV KTN NTV KBC TV

Graphic 8: Paid advertisement on TV channels – six levels of elections

KTN Citizen TV NTV KBC TV

Graphic 9: Paid-up advertisement on vernacular radios (1) – Six levels of elections

Mulembe FM Musyi FM Ramogi FM

Ramogi FM Musyi FM Mulembe FM

Citizen Radio Kameme FM Kass FM KBC Radio

B. ANNEX II – SOCIAL MEDIA MONITORING RESULTS

The EU EOM Social Media Monitoring Unit (SMMU) started its

Subject Analytical approach Monitoring Period No. of Posts