3.1 Overview of the Kenyan electoral management system

The management of electoral processes usually involves the interplay of diverse institutional responsibilities. All these efforts should, ideally, converge towards the common purpose of delivering free and fair elections – elections in which the average citizen, contesting political parties and candidates, and society at large are able to perceive that their various interests are reasonably respected and protected.

Broadly speaking, the Kenyan electoral management system is no different. But, as is often the case, there are differences in the dispersal of responsibilities among various institutions, the nature and structure of the agencies involved, the level of independence the lead agency enjoys vis-à-vis the government and the degree to which the central electoral responsibilities are concentrated in the hands of one principal agency, generically described as the electoral management body (EMB). In Kenya’s case, this is the Electoral Commission of Kenya (ECK).

Though its mandate and powers are dispersed between the Constitution and other statutes, the ECK is the central institutional player in managing the electoral process. It is solely responsible for the principal managerial/administrative functions in the process, such as the delimitation of constituencies, voter education, the registration and identification of voters, the reception and recording of candidate nominations and the organisation and conduct of polling and related operations. The ECK maintains operational independence from governmental structures to the extent that it recruits its own staff down to the grassroots and does not depend on existing local administration structures, as is often the case in countries where elections are organised through a ministry.

In regard to voter registration, the ECK’s work depends on a civil registry function (issuing of national IDs) which falls under a separate agency (the Registrar of Persons), whose mandate is not in itself electoral. The difficulties faced by this office in registering and issuing identity cards to all eligible Kenyans have negatively affected the ECK’s ability to capture the entire mass of Kenyans who have the legal right to register as voters. Proposals to rationalise the efforts of both agencies are already under consideration, with the aim of enfranchising as many as possible of those eligible to vote. This deserves priority attention, even if a complete fusion of the two agencies’ respective databases is not yet possible (see chapter 5 for a fuller discussion of voter registration).
SUMMARY OF INSTITUTIONAL RESPONSIBILITY FOR KEY ELECTORAL
FUNCTIONS
Electoral activities/functions Nature of function Common international practice Responsible agency in Kenya
• Registration of candidates MANAGERIAL Always assigned to EMB ECK
• Organisation of the elections ECK
• Registration of voters MANAGERIAL Frequently – but not necessarily – assigned to EMB ECK
• Identification of voters ECK
• Constituency delimitation ECK
• Decisions on complaints QUASI-JUDICIAL Frequently two-tier arrangements COURTS
• Civic and voter education INFORMATIONAL Frequently – but not necessarily – assigned to EMB ECK, CIVIL SOCIETY
• Regulation of electoral campaigns REGULATORY ECK (weak powers)
• Regulation of political party funding ECK (since 1 July 2008)
• Regulation of media during campaign MEDIA COUNCIL
• Regulation of political party registration ECK (since 1 July 2008)
• Regulation of candidate nomination Rarely assigned to EMB (not considered good practice) PARTIES

The ECK does not have any significant quasi-judicial powers, with the adjudication of complaints regarding election results being handled exclusively by the Judiciary (both first instance and appeals). The ECK’s involvement in election review is limited to ordering recounts and retallies, at the request of candidates and parties, within very specific time limits after the close of the poll. Voter registration complaints of a purely administrative nature are handled by the ECK, while claims and objections are dealt with by the Judiciary.

The ECK’s regulatory functions are also fairly limited, not being backed by robust legal provisions. Much has been said, for instance, about the ECK’s “toothlessness” when faced with errant parties and candidates flouting the Electoral Code of Conduct uninhibitedly in the lead-up to the December 2007 elections. The ECK lacked adequate powers to enforce even its own decisions on such infractions.

The regulation of the media falls under the Media Act (regarding all media houses) and the Kenya Broadcasting Corporation Act (regarding the public broadcaster [KBC] only), and is not an ECK function – save that the ECK must be consulted by the KBC in allocating free airtime for political parties to expound their policies during the campaign period. The ECK does, however, have a regulatory function in relation to the conduct of political parties, especially during campaigns. This arises from the Electoral Code of Conduct, to which all parties subscribe, but enforcement has been difficult.

It might be argued that ECK adopts a self-restrictive reading of some of its powers. For example, it has consistently maintained that once a returning officer has declared election results at the constituency level (particularly in the case of parliamentary and civic elections), it is limited to receiving and eventually gazetting them. This creates a situation in which “errors” can be committed in the tallying of polling station results, safe in the knowledge that they are subject to neither review nor sanction at a higher level.

This leaves the ECK itself in an awkward position (although it seems not to realise this) in which it implicitly certifies results without necessarily verifying them. There is perhaps justification here for an explicit legal provision for the certification by the ECK of results as a distinct step in the electoral process. Indeed, in many countries, especially those following the French tradition, certification is very often the function of a separate body, frequently at the highest judicial level or even parliament itself.

In assigning electoral dispute settlement exclusively to the Judiciary, Kenya’s electoral management system moves justice further away from interested parties. With the ECK not reviewing the work of its temporary staff, the brief (48-hour) period allowed it to review its own decisions, exorbitant filing fees and security for costs levied on election petitions, all this coupled with unlimited time for the hearing of election petitions, the potential that wrongs will never be righted is very high, as has been seen in the past (and in the 2007 elections as well).
3.2 The ECK’s legal framework
Establishment, membership and tenure

The ECK is established under section 41 of the Constitution, which provides for an electoral commission consisting of a chairman and between four and twenty-one members appointed by the President for renewable five-year terms. While the Constitution provides that the chairman and vice-chairman of the ECK should be persons who have held or are qualified to hold the office of judge of the High Court or Court of Appeal, there are no additional qualifications other than Kenyan citizenship for any commissioner, save that no member of Parliament, the public service or the armed forces may be a member of the ECK.

Commissioners enjoy security of tenure and can be removed from office “only for inability to exercise the functions of [their] office (whether arising from infirmity of body or mind or from any other cause) or for misbehaviour, and shall not be so removed except in accordance with [section 41]”. (This is the same security of tenure enjoyed by judges of the High Court [section 62] and the Court of Appeal [section 64], the Attorney-General [section 109] and the Controller and Auditor-General [section 110].) Should the question of removal arise, the President is authorised to appoint a five-member tribunal to “inquire into the matter and report on the facts thereof to the President and recommend to him whether the member ought to be removed”. Pending the tribunal’s report the President may suspend the commissioner in question from office.

The ECK’s institutional independence is provided for in subsection 41(9): “In the exercise of its functions … the Commission shall not be subject to the direction of any other person or authority.” The only limitation on this institutional freedom is in the power provided to Parliament in subsection 41(10) – explicitly “without prejudice to subsection (9)” – to “provide for the orderly and effective conduct of the operations and business of the Commission and for the powers of the Commission to appoint staff and establish committees and regulate their procedure”. Utilising this power, Parliament has authorised the ECK to appoint such staff as it may deem necessary to perform its functions, in addition to the substantive electoral process requirements in the National Assembly and Presidential Elections Act, the Election Offences Act and related election legislation. Added to this are the Code of Conduct for Members and Staff of the Electoral Commission (Second Schedule, Cap. 7 of the Laws of Kenya) and the Proceedings of the Electoral Commission (Third Schedule, Cap. 7) as well as other subsidiary legislation enacted pursuant to powers conferred on either the ECK or other institutions in the foregoing laws.
Functions

Sections 42 and 42A of the Constitution confer upon the ECK the power to –

* divide Kenya into such number of constituencies having such boundaries and names as it may prescribe by order, subject to its reviewing the number, names and boundaries of constituencies at eight-to-ten year intervals and on certain other occasions (this is limited only by Parliament’s power to determine the minimum and maximum number of constituencies)

* register voters and maintain and revise the register of voters

* direct and supervise the presidential, parliamentary and civic elections;

* promote free and fair elections

* promote voter education throughout Kenya

The Constitution gives Parliament the power to prescribe through further legislation such other functions for the Electoral Commission as it may deem fit. Parliament has used that power to mandate the ECK to determine the proportion of nominated members of Parliament and councillors that each political party should nominate and to ensure the observance of the principle of gender equality in such nominations (section 33, Constitution of Kenya; section 26, Local Government Act). Parliament has also conferred upon the ECK the power to divide any municipality, county, township or county division into electoral areas having such boundaries and names as may be prescribed by order (section 10, Local Government Act) and to supervise party nominations on request (section 17(3), National Assembly and Presidential Elections Act). More recently, Parliament has again used that power to establish the office of Registrar of Political Parties within the ECK and to confer upon it wide-ranging powers to regulate the establishment, organisation, management and funding of political parties. In addition, section 17A of the National Assembly and Presidential Elections Act accords the ECK “overall conduct of elections” and the power to “give general directions and exercise supervision and control thereof and take the necessary measures to ensure that the elections are transparent, free and fair”.
Assessment of the ECK’s legal framework

Institutional design and management: A review of the legal framework for the ECK reveals a number of institutional structure issues that require legal attention. Examples, drawn from comparative Kenyan semi-autonomous government agencies and EMBs elsewhere, include: incorporation; chief executive (by whatever name called); oath of office; principal office; meetings and procedure; officers and other staff; guiding principles; powers (including quasi-judicial powers); finances; investment of funds; financial year; annual reports; annual estimates; accounts and audit; and holding of property. Beyond the Constitution and sections 3-3B of the National Assembly and Presidential Elections Act, there is no law governing many of the ECK’s institutional and operational aspects. This is in stark contrast to the situation in some countries in Africa (such as Malawi and Ghana) and elsewhere where this is provided for in a dedicated law to back up constitutional provisions on, for instance, establishment and security of tenure. It is also in contrast to the position of a number of Kenya’s own constitutional bodies, such as the Judiciary and the Parliamentary Service Commission.

Security of tenure: Although the security of tenure provisions exist to good purpose and should be maintained, they have not been interpreted so as to ensure institutional service delivery. This applies not only to ECK commissioners but also to other constitutional office holders. It seems that, with the exception of the case of John Harun Mwau (then Director of the Kenya Anti-Corruption Commission), whose removal after only six months in office was recommended by a tribunal in 1997 on account of incompetence, security of tenure has been interpreted to mean non-removal on any ground that does not point to a serious criminal offence, such as corruption or other economic crime. This attitude must change if constitutional office holders are to respect their oath of office and be fully accountable to the public. Suffice it to say that security of tenure does constitute difficulty in removing officers, but its higher goal is to establish justice in the process of removal, rather than make removal almost impossible.

Institutional independence: The ECK has sufficient institutional independence to discharge its mandate. However, by its own account, this is limited by its lack of financial independence and the general political behaviour of the various actors in Kenyan elections. The former is covered substantively in section 3.6 below, so here we concern ourselves only with the latter. In our analysis, the paper provisions give the ECK the institutional independence to do its work. What is needed is an appointment process that inspires public confidence in the Commission and selection criteria that ensure commissioners have the integrity, sense of judgment and mettle to referee a political contest without being unduly influenced by political pressure. (Section 3.3 deals with this issue.)
Conclusions/findings

Institutional design and management: Certain important institutional matters of design and management are not covered in the Constitution (with good reason, since that is not the place for such detail) and the National Assembly and Presidential Elections Act, and cannot be properly managed merely with the aid of operational manuals and common-sense. They will require more ample legal provisions, following the practice of public institutional design in Kenya as well as best practice elsewhere. This is becoming increasingly essential since, with its additional legal functions, today’s ECK is a far more sophisticated organisation than the one that managed the 1992 general elections, or the ones before that.
3.3 Composition of the ECK and appointment of the commissioners

With a chairman and twenty-one members, the ECK is currently at the full strength constitutionally allowed. While section 41(10) of the Constitution states that “Parliament may provide for the orderly and effective conduct of the operations and business of the Commission and for the powers of the Commission to appoint staff and establish committees and regulate their procedure”, no distinct legislation exists further elucidating in any comprehensive manner the mode of appointment, powers and tenure of the ECK (although the National Assembly and Presidential Elections Act contains provisions describing its mandate in relation to those elections).

Contrary to what has become customary in many countries with independent electoral commissions, the President is not required to consult with any other institution or persons in appointing ECK members. It is also worth noting that there are no legal criteria to guide the President in making such appointments (whether in terms of professional profile, representation of diverse interests or otherwise). This apparent lacuna in the legal framework has been the subject of much controversy over many years.

Commissioners are appointed for (renewable) five-year terms. Because the first set of commissioners was appointed in an election year (1992), and another set appointed in the next election year (1997), this time very shortly before the elections, the end of each five-year cycle has recurrently coincided with an election year, potentially causing significant disruptions in planning and implementation. Ten commissioners were replaced in January 2007 and another five in October 2007, with elections scheduled for December that year (the chairman’s mandate was renewed only in December, after some uncertainty and public debate). Quite apart from the effect of the eventual new appointments (time was, for instance, too short for new commissioners to acquire sufficient familiarity with their functions), uncertainty as to what the President would decide vis-à-vis retaining or replacing the vacating commissioners, as well as the absence of consultation when he did make new appointments, contributed to pre-election tension and undermined the ECK’s credibility as an independent body capable of delivering fair elections.

This was not the first time such a public debate had taken place on this issue. In 1997, dissatisfaction with the President’s exclusive de jure role in appointing ECK commissioners (focusing, in particular, on his partisan status as a candidate in the elections and as the leader of a contesting political party), with mounting pressure from political parties and civil society, resulted in agreement on alternative arrangements for the nomination of members. In terms of this Inter-Parties Parliamentary Group (IPPG) agreement the parliamentary opposition would supply the President with a list of nominees, from which he would appoint ten commissioners. This was the basis on which additional ECK members were appointed prior to the 1997 elections. (It was also the backdrop to the topping-up of the Commission in 1997 to its full complement of twenty-one ordinary members; the Constitution provides only for minimum and maximum numbers, and the maximum number of members had not been appointed before that occasion.) It could be argued that the increased number of members in itself makes reaching consensus on an issue more difficult than it otherwise would have been.

Although the IPPG accord was never entrenched in law, with the terms of the commissioners appointed under the arrangement approaching expiry in 2002, President Moi renewed their appointments. After 2002, when commissioners died or concluded their terms President Kibaki replaced them without reference to the 1997 IPPG accord. This caused little initial acrimony, but as he continued to replace commissioners in this manner – especially as the 2007 general elections drew closer – discontent surrounding the matter was rekindled. The President preferred to follow the letter of the Constitution, using his exclusive prerogative to appoint new members, while the opposition demanded he observe the spirit of the IPPG agreement. The issue became particularly acute in 2007, in which year the President appointed fifteen members.

As has been mentioned, in the absence of any consultative requirement, the Constitution grants the President total latitude in appointing ECK members. The only criterion stipulated – and even then this applies solely to the chairman and the vice-chairman – is that they “shall be persons who have held or are qualified to hold office of judge of the High Court or judge of appeal” (section 41(2A)). In practice, this means that a Commission could comprise a majority of members lacking any professional experience or expertise in election management or other skills that could contribute to the judicious and evenhanded management of an electoral process.

International practice in the composition of EMBs tends to revolve around certain major principles which are sometimes combined, such as professional competence, political balance and consensus. These principles tend to reinforce EMBs in terms of professional stature, credibility and political acceptability (although in certain circumstances, some can create other problems, such as the risk of a deadlock in a consensus-based EMB). As the legislation now stands (and is practised, with the exception of the brief romance with the IPPG arrangements), Kenya’s electoral commission does not seem to conform to any clearly defined appointment principles. Inevitably, its credibility suffers, since key stakeholders cannot feel a sense of (part) ownership of the structure, nor can they place much faith in its inherent professional competence.

Conclusions/findings

In IREC’s public meetings and formal hearings, the ECK and some stakeholders advocated several reforms to be introduced in the way its members are appointed, their security of tenure, duration of their term of office, staggering of the terms and the size of the commission, among other issues. Various models have been proposed and carefully considered by IREC in comparison with EMBs around the world. The most recent constitutional proposal on this question suggested a total of between three and nine Commissioners appointed by the President and approved by Parliament. There is no doubt that this is an area that is ripe for major changes, if future elections are to enjoy the trust of the various stakeholders in Kenyan society.
3.4 The ECK’s operational procedures

There are a number of practical and political features of the process that an EMB needs to bear in mind in planning and conducting elections. These include:

* Elections are high-pressure events. Once an election date is set, election management involves meeting a series of deadlines; the political penalty for missed deadlines is high, for election managers, government and citizens. In Kenya’s case, although the election date is not known well in advance, events following the dissolution of Parliament are all time-bound, exerting a great deal of pressure on the EMB.

* Elections involve high stakes. Their credibility is tied to national stability, and the winning and losing of elections is tied to political party power. As is evident in many parts of the world, elections can precipitate an outbreak of violence or, where this has occurred before, a recurrence of violence.

* Elections are periodic or occasional events. National elections usually take place at widely spaced intervals. While an EMB needs some permanent staff all the time for planning and other continuous activities, an enormous short-term staff is required at the time of elections (and down-scaled between elections). In Kenya, presidential, parliamentary and civic elections are held every five years.

* The electoral process should be predictable; it should be governed by laws and rules that are not only commonly understood but also universally applied.

* Elections must ultimately be a nation-building exercise, rather than a divisive one. Since elections provide a means by which political power can change hands peacefully, they need to be managed in such a way as to deliver not only an arena for peaceful political contest but also to ensure general acceptance of outcomes and stability for other nation-building activities.

Because of these features, any EMB requires – and usually develops – procedures to guide its work, both in election years and in the years between. The ECK has, over time, developed certain procedures to guide its work. Some are written while others are organisational practices that have emerged over the life of the ECK.

An EMB’s success lies in how greatly it inspires public confidence in its role as an impartial manager and referee of elections. An EMB’s procedures can be evaluated according to these criteria:

a) how transparent all its processes are

b) how simple and accessible its procedures and service are to all

c) how much it respects the principles of equity and equality as a measure of ensuring free and fair elections

d) how well it asserts its independence

e) how well prepared it is for all its activities

f) how well it anticipates and deals with any electoral disputes and hitches

g) how well it keeps learning from its experiences for the improvement of its processes

h) how well it can enforce or ensure enforcement of its decisions and processes

With the foregoing in mind, an EMB should also invest in procedures that ensure adequate long-term (even up to twenty years) and operational (at least annually) planning for elections, efficient and effective execution of tasks in the electoral process and a strong system of inter-agency coordination with other government bodies whose support it requires to fully discharge its mandate. In this section, we discuss key aspects of the ECK’s election procedures, both at election time and in the periods between elections.
Between-election procedures

ECK procedures between elections are tailored to the execution of a range of activities, both routine and time-specific. These include election planning, voter education, election petitions, by-elections, voter registration, boundary delimitation, materials design, training, research and electoral law reform. These are easily governed by the procedures developed for the day-to-day management of any organisation.

Commissioners set policy in plenary meetings. Policy is implemented through committees chaired by specific commissioners on which departmental heads serve as secretaries. The ECK technical paper on its organisational structure submitted to the IREC technical workshop in August 2008 lists the following committees:

* General Purposes Committee

* Finance and Logistics Committee

* Electoral Programmes Committee

* Legal and Electoral Reforms Committee

* Public Education and Outreach

* Human Resource and Manpower Development Committee

* Electoral Research Committee

* Audit Committee

* Committee on the Administration of Political Parties

From time to time the ECK also establishes ad hoc committees to deal with various matters. Such committees are disbanded once they have discharged their mandate. Committees, which average six members, meet regularly and make decisions which are then submitted to Plenary for decision.

As to strategic management, the ECK had no strategic plan until 2004 when the current 20042008 Strategic Plan was adopted. The Plan, which places a premium on the attainment of free, fair, credible and professionally managed elections, aims to achieve a number of high-level results. These are reduced electoral violence and malpractices, more equitable representation, informed choices, increased participation (particularly by women and other marginalised groups), professionally managed elections, strengthened political parties, increased Government sensitivity to the electoral process and equitable and responsible media coverage of the electoral process. These are sought to be achieved via a number of strategies: organisational development, mainstreaming gender and diversity, resource mobilization, managing strategic partnerships, public relations and marketing, voter education, and information technology strategies. The achievement of the Plan’s results is to be monitored through a range of predetermined processes.

In regard to other aspects of its work, the ECK has produced a plethora of manuals and other publications over the years to guide its staff and other stakeholders through legal and other requirements. These include various election manuals, a sourcebook on voter registration, a voter education facilitators’ manual, guidelines for election observers, media guidelines, a voter registration leaders’ guide, a returning officer’s guide, a presiding officer’s guide, guidelines for peace committees, guidelines for elections security officers (police), guidelines for polling and counting clerks, various versions of a voter’s handbook, a review of constituencies and electoral areas handbook, a voter education curriculum, an election handbook for political parties and candidates on parliamentary and civic elections, and a polling station operational manual.

For internal use, the ECK has produced guidelines for recruiting permanent staff and for recruiting temporary election officials. Of particular interest is the latter, which are not only detailed but appear to have been drafted to ensure an open and competitive recruitment process. In summary, (a) all posts are to be declared vacant and advertised as and when they fall vacant or when the need arises; (b) detailed job descriptions and job requirements are to be assigned for each category of personnel/election officials to assist in recruitment; and (c) panels for shortlisting and interviewing candidates are to be constituted after receipt of applications to ensure that the exercise is transparent.
Procedures in the 2007 general elections

By contrast, the procedures in an election year require a higher degree of sensitivity to timelines. They work well only if there has been adequate prior preparation. By the ECK’s own account, preparations for the 2007 general elections began in 2004 with boundary delimitation. Since this is a routine between-elections activity based on the intervals prescribed in the Constitution, real preparation must have begun in 2005 with the ECK chairman’s memorandum enquiring into the ECK’s preparedness for the coming elections. In 2007 alone, the ECK concluded its review of constituencies, electoral areas and polling stations, registered voters, procured election materials, recruited and trained election officials, conducted voter education, received nomination papers from candidates and conducted the presidential, parliamentary and civic elections. As part of the preparations, a number of consultative meetings, conferences and workshops were held, some with key stakeholders, including the 13-15 March 2007 conference with political parties.

Owing to the nature and scale of activities in 2007, and in keeping with previous practice in election years, the ECK established a number of ad hoc committees to supplement the standing committees. These were the Code of Conduct Implementation Committee, the Party Liaison Committee and the Steering Committee. The first was responsible for receiving complaints on violations of the Electoral Code of Conduct and enforcing the Code, the second was established in accordance with rule 9 of the Proceedings of the Commission to facilitate consultations with political parties on any matter concerning the elections while the third was in charge of overall preparations for the election.
Conclusions/findings

The ECK has made efforts ever since its establishment to improve its procedures for managing and conducting elections. As things now stand, it is a functioning institution with policies and procedures. However, those policies and procedures– at least in the 2007 general elections – have not managed to deliver a satisfactory election. Structural weaknesses in the relationship between the commissioners and the Secretariat (in terms of the delineation of roles), bureaucratic procedures; and an unwieldy committee structure in some ways hampered smooth preparations for the elections and interfered with staff selection, recruitment, training and deployment.

The manuals for election staff and their training, though improvements on previous versions, still granted some discretionary power to lower-level staff (for instance, whether to give a voter one ballot paper at a time or all three at once, whether to count the presidential or the civic ballot papers first, etc). As a general practice, electoral procedures ought to anticipate the environment in which they will operate and provide as much operational detail as is required to inform decision-making. When matters are left to the good sense or judgment of the field staff, breakdowns in communication (resulting for instance in the reporting of partial results when this had not even been discussed) and non-compliance with laid-down procedure are the likely result. Training seems to have been conducted more or less uniformly, though in some areas implementation did not go according to script. For instance, some election officials went through a comprehensive programme that included simulation (very useful and important for adult learners) while others seem to have been taken through a hurried briefing rather than training. This introduced uneven implementation that in some cases gave rise to uneven performance (this was one reason some regions delayed presidential results significantly, for example). Certainly the tallying staff at the national tallying centre either had not received adequate training or, if they had, did not quite do as they had been instructed. One team leader even used briefing instructions that were later repudiated by the ECK in the formal IREC hearings. The result of this uneven performance and the chaos at the national tallying centre was suspicion and the sullying of the overall integrity of the electoral process.
3.5 The ECK Secretariat

The ECK Secretariat has been built up over many years. Although the Independence Constitution (1963) created an electoral commission, it played a subdued role because the office of Supervisor of Elections, which fell under the Attorney-General, supervised elections instead. Although the Supervisor of Elections had wide powers to conduct elections and register voters, the office was not independent: it neither had operational autonomy from the Executive nor did its officers enjoy security of tenure. The result was most evident in the much discredited mlolongo (queue-voting) elections of 1988. After much agitation for political and legal reform, and with the resumption of multiparty politics, Parliament abolished the office of Supervisor of Elections in 1992 and recognised the ECK’s role as the sole body responsible for the conduct of elections. While commissioners were appointed from 1992, it was not until 1998 that the ECK Secretariat began to take shape.
The period before 1998

According to the ECK, the 1992 elections were conducted by the commissioners assisted only by a skeleton staff seconded from various government departments. The commissioners of the time and their few “borrowed” staff carried out the day-to-day tasks of the ECK through a number of committees. The 1997 elections were conducted by this same staff, although by then some 41 District Elections Co-ordinators (DECs) had been employed on a casual basis.

The lack of permanent staff constrained the ECK to use commissioners for day-to-day work. It could not develop a functional secretariat. In an early publication, Towards Free and Fair Elections (1999), the ECK decried the fact that most of its staff were on secondment from the government, stating that this was “not an ideal situation as it may jeopardise the desired commitment and confidence on the part of the officers”. Thus, with split loyalties, it was as difficult to retool this skeleton cadre to the new institution’s values and mode of work as it was to ensure their operational independence from their mother departments. The ECK therefore continued pressing Treasury for resources to establish an independent secretariat.
The period from 1998 to the present

By July 1998, these efforts had borne fruit and by the end of that year the ECK had 112 permanent staff following an increase of DECs to one for each of the 70 districts on three-year renewable contracts. By 2002, the ranks had swelled to 448.

In terms of structure, the ECK began with an Administrative Secretary as head of the Secretariat. There were also a Deputy Administrative Secretary and heads of departments, each with a basic staff. The first departments were Accounts, Supplies, Personnel, Public Relations, Computer and Legal. Later on, the Administrative Secretary became the Commission Secretary, with accounting officer status once the ECK had its own independent vote. The Deputy Administrative Secretary became the Deputy Commission Secretary, and a second deputy was appointed, while some of the departments were renamed or split to form new departments. The current departments are Finance, Procurement and Supplies, Human Resources, Public Relations, Elections, Cartography, Elections Training, Information Technology, Legal and Registration of Political Parties. Each department is headed by a professional recruited according to the ECK’s hiring procedures. More recently, each DEC has had an Assistant Registration Officer and an Election Officer II added to the original basic structure of a DEC, a secretary and a driver/messenger. The ECK currently has a permanent staff of over 500.

As have the commissioners, the Secretariat staff have their own “management standing committees” that meet regularly to deliberate on various issues and implement decisions in consultation with commissioners. These are the Appointments and Selection Committee, the Tender Committee, the Disciplinary Committee, the Training Committee and the Budget Committee.

As the Secretariat has evolved, so too have roles, although some roles still overlap, resulting in commission-management tension within the organisation. This tension was admitted by the Chairman of the Commission in his evidence to IREC and though it has been downplayed by senior members of staff such as the Commission Secretary (in his evidence to IREC), it is real. It results not from ill-will on either side but from the conflict of roles bound to occur when commissioners who were used to day-to-day management of the ECK’s affairs found themselves with a permanent staff cadre including professionals. Though originally muted, it came into the open in the course of the development of the ECK’s first strategic plan and would need to be addressed, through a re-examination of commission-management roles, if institutional harmony were to be established. In his evidence to IREC, the ECK Chairman stated that consideration could be given to commissioners’ filling both policy-making and executive roles so long as executive lines of responsibility were clear.

On the basis of technical assistance provided by the International Foundation for Electoral Systems (IFES), the Secretariat has drawn up job descriptions for all staff. A human resources manual, which includes gender integration policies, has been developed. Annual operational planning is gaining currency, and the introduction of performance measurement criteria is being discussed. Also under discussion is the improvement of the Secretariat structure, as part of the organisational development process accompanying the implementation of the current strategic plan.
Conclusions/findings

The ECK has professional staff in many of its departments – all departmental heads at least are professionals. Given the Commission’s committee system, however, they will continue to be nothing more than “senior assistants” to the commissioners if they are not given room to apply their professional minds to the issues. This does not bode well for the ECK; in a body where policy-makers come and go according to the appointing authority’s preferences, the need to professionalise the Secretariat and make it the repository of the institutional memory of the ECK cannot be gainsaid. Indeed, one of the ECK’s weaknesses in the run-up to the 2007 elections – that most commissioners had insufficient experience in running an election – could have been offset by a professional secretariat of election workers with substantially devolved decision-making authority on the management of the electoral process.

The current Secretariat structure lends itself to administrative inefficiency and waste. Standing committee business, for example, seems to take up time that would be better spent on operations (some committees do not even keep minutes). The present system makes it possible, as in the approval of the proposal to transmit election results electronically, for a decision to be communicated to the implementer over a month after it has been made! The committee approach to work also reduces departmental multitasking, which puts unnecessary pressure on the organisation every time a major exercise is undertaken. It would be much more efficient for some of these exercises (say the annual voter registration drive) to take place at the same time as other regular business, such as coordination of voter education. Between elections, the effect of this modus operandi is not really felt, save for the wasteful involvement of non-professionals in professional tasks. In an election year, however, when time is of the essence, it can be debilitating. Consultations in committee are necessary, but they should be a prelude to implementation and not take time away from implementation itself.

The role of district staff (DECs and their support staff) is provided for in pre-election (and nonelection year) activities. However, when it comes to election-day activities, district staff are conspicuous by their absence. There is no role for them in a formal sense save as logistics personnel. Even this is a role they seem to lapse into; it is not an allocation on the basis of which they can be held accountable. Yet (subject to the reforms IREC proposes on the ECK’s devolved structure below) they could perform some of the tasks conducted by temporary election officials, although in certain constituencies, depending on geography, they might well be confined to the constituency tallying centre. This would not only reduce expenses, but would also lessen the risk associated with introducing temporary election officials in key areas such as tallying.
3.6 Funding the ECK permanent structure

In order to discharge its constitutional responsibilities, the ECK requires a great deal of money. Commissioners and staff, both permanent and temporary, need to be paid; the ECK’s daily operations and election materials, their delivery and use in the electoral process have to be financed. Funding is therefore required not just for recurrent expenditure but also for development and to support electoral exercises. Commissioners’ salaries and allowances are a charge on the Consolidated Fund and are determined in accordance with the Constitutional Offices Remuneration Act, so they raise no issue regarding independence. It is all the other costs that form the subject of this review of the ECK’s financial independence.

In the course of its investigations, IREC was informed by the ECK and other interlocutors that the ECK had no financial independence. It was recommended that this situation be addressed by making the ECK’s expenses a charge on the Consolidated Fund. As this is a matter that affects the ECK’s independence, capacity and functioning, we examine those arguments in the light of all the available information and the Government Financial Regulations and Procedures.
Current ECK funding modalities

The ECK’s funding mode has been in place since 1998. When the ECK was provided with finances to establish its own secretariat, it was also allocated its own vote (according to its Secretary, this is Vote 33) in accordance with the Government Financial Regulations and Procedures.

The process of funds acquisition begins with the ECK preparing budget proposals for each financial year and submitting them to Treasury. The ECK takes into account all the activities it plans to conduct in the year in question and adds in contingencies and routine costs such as staff emoluments and transport. Treasury may then open discussions with the ECK on the entire budget or on specific items. The agreed budget is then included in the Ministry of Finance financial estimates tabled in Parliament. (In the case of unanticipated expenditure, the Minister includes the additional costs in supplementary estimates.)

Parliament then scrutinises the estimates and, although it has the power to approve an amount lower than that indicated in any vote, it has never in fact made any changes to the ECK’s vote. Once the estimates have been approved, the money can be paid into the ECK account. Treasury does this on the basis of monthly allocations. If lump sums are required (say, for the purchase of motor vehicles), these are included in the monthly requests in addition to recurrent monthly items.

The ECK has, over the years, expressed a number of difficulties with this funding method. One problem has been the need to observe financial procedures “some of which are difficult to apply to elections” because they cause “severe operational limitations”. It is also claimed that Treasury officials sometimes slash budgets on the basis that “there is no adequate funding”. Low allocations to such items as voter education are given as an example of how Treasury consultation has worked against the ECK’s objectives. In the course of the IREC investigation, delays in ensuring that “sufficient funds are made available in good time to the Commission for effective management” of its operations was added to this list.
Assessment of ECK financial independence

The ECK enjoys operational independence but relies on public funds for its operations. That is why it is subjected to the procedure described above and why it agitates for its expenses to be a charge on the Consolidated Fund. What, then, does financial independence mean in this context? Does the ECK have financial independence? If not, how should it be instituted?

Section 99 of the Constitution establishes the Consolidated Fund, into which “all revenues or other moneys raised or received for the purposes of the Government of Kenya shall be paid [and] from which no moneys shall be withdrawn except as may be authorised by this Constitution or by an Act of Parliament (including an Appropriation Act) or by a vote on account passed by the National Assembly …”. The section goes on to provide that moneys for certain authorities may be paid into other accounts established for that purpose so long as they are managed in accordance with the law and that “Parliament may prescribe the manner in which withdrawals may be made from the Consolidated Fund …”.

Payments are made from this account to meet various kinds of expenses:

(1)
public debt

(2)
salaries of constitutional office holders

(3)
international subscriptions

(4)
other government expenditure

With respect to (1), section 103 authorises its payment as a charge. With respect to (2), section 13 of the Constitution provides for “[t]he salary and allowances payable to the President and any pension or gratuity payable to him on retirement” to be a charge upon the Consolidated Fund, while section 104 authorises payment in this manner for judges of the High Court and the Court of Appeal, members of the Public Service Commission, members of the Electoral Commission, the Attorney-General and the Controller and Auditor-General (all referred to in the Government Financial Regulations and Procedures as Consolidated Fund Services). The respective scales are set out in the Constitutional Offices (Remuneration) Act. By dint of chapter 4 of the Government Financial Regulations and Procedures, expenses in relation to (3) are paid with the same authority granted for (1). All other government expenditure is paid out either through an Appropriation Act (passed annually upon presentation of estimates of Government revenues and expenditure for the following financial year by the Minister for Finance) or through a specific Act of Parliament other than an Appropriation Act or a Vote on Account for a specific body, either directly or through a line ministry.

There is no precedent for charging the entire expenses of an institution to the Consolidated Fund. The Kenya National Commission on Human Rights (KNCHR) derives its funding from “such sums as may, from time to time, be appropriated by Parliament for that purpose”. It may receive grants and donations from any other source “provided that such grants and donations shall not be made or received for purposes of influencing the decision or ability of the Commission in any way and shall be disclosed in the annual report of the Commission”. The National Commission for Gender and Development (NCGD) derives its funding from moneys provided by Parliament for the purposes of the Commission; such moneys or assets as may accrue to the Commission in the course of the exercise of its powers or the performance of its functions; and all moneys from any other source provided for or donated or lent to the Commission. Even the Parliamentary Service Commission – one of the examples the ECK relies on in its technical paper on the subject

– does not enjoy the suggested funding modality.

Since the ECK currently has a funding modality similar to those of all other independent institutions, it may seem difficult to support the argument that all its expenses be a charge upon the Consolidated Fund. That argument becomes especially difficult in the light of the fact that the ECK has not been the epitome of financial efficiency. The upshot is that, were the entire expenses of the ECK to become a charge on the Consolidated Fund, this would be a radical departure from the conventional mode of financing semi-autonomous government agencies with operational independence (such as the ECK). However, the necessity for financial independence for whatever electoral management body (EMB) may result from the implementation of the IREC recommendations should be carefully weighed against any fears arising from the ECK’s financial management history. This is probably why the drafters of the Proposed New Constitution of Kenya (2005) provided that:

278. (1) The funds of a constitutional Commission include –

(a)
money voted by Parliament for the purposes of the Commission; and

(b)
any other money received by the Commission in the performance of its functions.

(2) The administrative expenses of a constitutional Commission, including remuneration and benefits payable to, or in respect of, persons serving with the Commission, are a charge on the Consolidated Fund.
Conclusions/findings

The ECK has as much financial independence as all Kenya’s other independent institutions, such as the Judiciary and the Parliamentary Service Commission. Noting the direction that the Proposed New Constitution of Kenya took in making the administrative expenses of all constitutional commissions (such as the ECK) a charge on the Consolidated Fund, it is difficult to understand how this was going to work in practice. IREC, having considered the practice in other jurisdictions, is not prepared to recommend a provision amounting to financial carte blanche.
3.7 The ECK’s advisory role on legal reforms

Besides the roles discussed in section 3.2 above, a role common to EMBs the world over is to advise governments and/or lawmakers on desirable changes to electoral law. Though it is not provided for by any of the Kenyan laws that IREC reviewed, this power has been inferred. IREC was interested in exploring any work the ECK has conducted in this area since in much of its testimony and technical presentations the ECK attached much importance – and rightly so – to enabling legal provisions. A review of letters and memoranda addressed by the ECK since 2002 to various offices within the executive and legislative branches of government tells a sad tale.

Over the years a number of carefully reasoned and cogently substantiated submissions and recommendations by the ECK urging fundamental reform of virtually every aspect of Kenya’s constitutional and legal framework for elections came to naught. A letter dated 3 May 2002 from Chairman Kivuitu to the secretary of the Constitution of Kenya Review Commission, is prefaced with the observation that “it is difficult to separate [electoral law] from the rest of the Constitutional provisions and other laws that indirectly affect the electoral process”. It made a number of recommendations which could sensibly be repeated verbatim today. They relate, to name a few of those more salient in the present context, to the assumption of presidential office, gaps in the enforcement of the Electoral Code of Conduct, the empanelment of an electoral court, service of an election petition and, crucially, the size of the commission, criteria for the selection of commissioners and an inclusive and transparent appointment process. Many other proposals were made, relating, for instance, to reinforcing the ECK’s independence by having it report directly to Parliament, staggering the three elections, establishing a truly professional secretariat with “specialist skills and competencies” and providing for “effective enforcement of electoral rights”. One can but speculate as to the course of events in December 2007 if these recommendations had been taken seriously.

Conclusions/findings

It is indeed ironic that the ECK, the composition and legitimacy of which has been so trenchantly criticised since the 2007 elections, itself made proposals for reform which could have prevented the fiasco. However, these proposals were not pursued, or they were frustrated by a political agenda that did not give them the attention they deserved. As things stand now, there is nothing in law to prevent the relevant policy- and law-makers from taking the ECK’s advice, discussing all or aspects of it with the ECK, and then incorporating the final result of those deliberations into electoral law reform proposals. However, owing to the lukewarm response the ECK’s proposals have received, IREC is constrained to make recommendations on an issue that probably does not require more than a change of attitude. There are other matters, however, on which the ECK requested legal reform that IREC does not believe require such reform – for instance, the use of modern technology for transmitting results.
3.8 Assessment of the functional efficiency of the ECK and its capacity to discharge its mandate

Three basic problems confront those organising elections: credibility, sustainability and efficiency. Credibility is the litmus test of an electoral structure: unless trust in the system is developed, the whole electoral exercise loses significance. A non-credible system does not confer democratic legitimacy upon those elected. Credibility is also a universal problem of electoral organisations, although it is frequently forgotten in consolidated democracies where it is taken for granted. The ECK’s credibility was severely dented in the 2007 elections, and one of the main challenges that the Commission will face is how to restore it.

Sustainability can be defined as the ability to conduct an electoral process with domestic human and financial resources. This is not a major issue in Kenya, and the support that the ECK has been receiving from external sources is limited and could easily be substituted by domestic resources.

The last of the basic problems – efficiency – is a universal problem confronting all types of organisations. It may not be very polite to discuss the efficiency of entities that perform good deeds, particularly those that should be cornerstones of the democratic process. However, this is an inescapable issue: an electoral organisation must strive for the most cost-effective way to achieve a defined level of credibility.

These three challenges are interrelated in a somewhat perverse way. Credibility is the essential consideration, and sustainability and efficiency are frequently sacrificed on the altar of credibility. Inefficient solutions are frequently adopted because they seem to contribute to credibility. For instance, expensive printing of ballot papers abroad is adopted because it contributes in the short run to credibility. But such inefficiencies then become entrenched as cultural givens after two or three elections.

It is extremely difficult to assess the ECK’s efficiency without conducting a specialised study. However, we shall attempt to evaluate three areas which provide reasonable pictures of the overall efficiency of the organisation. The first such area will be the cost of the electoral process in Kenya, as compared with that in other countries. Secondly, we shall make an evaluation of the efficiency of registration, since we have fairly complete data on the subject. Registration is a useful example, as it is one of the few electoral activities conducted between general elections. Thirdly, we shall examine one specific – and important - case of decision-making in the organisation which we have been able to reconstruct in detail. The case provides an opportunity to look, even if only superficially, at planning and innovation in the organisation.

Comparative costs of elections: Common-sense and experience would lead one to expect the cost of elections in Kenya to be comparatively low. The following factors should support this expectation:

* Kenya conducts the three elections (presidential, parliamentary and civic) simultaneously.

* Kenya has significant experience in elections; this tends to correlate with a decrease in costs. It has conducted four general elections since 1992, and a considerable part of its key personnel has been involved in the organisation of elections for ten years or longer.

* The costs of temporary personnel are reasonably low.

* Registration is a one-step process and voters’ cards, which do not incorporate photographs or sophisticated security elements, are not expensive.

For a rough estimate of the cost of the electoral process throughout a five-year cycle, we have used data from three financial years: 2005/6, which can be considered as a normal year without general elections, and 2006/7 and 2007/8, which correspond to financial years during which practically all the activities related to the 2007 general elections took place. We have added the cost of the Commission itself, which is a direct charge on the Consolidated Fund and does not appear in the budget estimates. The amounts used are probably conservative, since they do not include potential upward revision of the 2007/8 estimates, which might still occur. The costs are as follows:
Commissioners’ salaries (approx. KSh 100m/yr x 5 yrs) KSh 500m
Budget 2005/6 (KSh 1,936m/yr x 3 non-election yrs) KSh 5,808m
Budget 2006/7 (includes registration & other expenses) KSh 4,872m
Budget 2007/8 (most of the election expenditures) KSh 8,195m
TOTAL KSh 19,375m

The total cost of this full electoral cycle is thus KSh 19,375m, approximately equivalent to US$ 290m. It should be noted that this amount does not represent the full cost of the elections since the estimates do not include:

* the cost of security (provided through a different budget)

* the contributions made by foreign donors to the ECK or to civil society organisations (CSOs) involved in voter education

* contributions to the funding of political parties, as is the case in some countries with which comparisons are made

* the full cost of building a voter register, since the register used for the 2002 elections was the basis and registration expenditures only cover additions to that register between 2003 and 2007

The cost of the election (without the additional costs mentioned above) is thus US$ 20.4 per registered voter (or US$ 29 per cast ballot). If we do not consider the whole electoral cycle, but limit the calculations to the costs sustained in 2006/7 and 2007/8, in which period the general elections took place, the cost per registered voter is US$ 13.74. How does this compare with other countries?

Comparison is somewhat difficult, because the data from other countries may not correspond with the items we have considered when estimating Kenyan electoral costs. However, these quotations from the article, “How much do elections cost?” in the ACE Encyclopedia (produced by the Administration and Cost of Elections [ACE] Electoral Knowledge Network) go some way to providing an answer:

“Low electoral costs, approximately US$ 1 to US$ 3 per elector, tend to manifest in countries with longer electoral experience: the United States and most Western European countries; Chile (US$ 1.2), Costa Rica (US$ 1.8) and Brazil (US$ 2.3) in Latin America; Benin (US$ 1.6), Botswana (US$ 2.7), Ghana (US$ 0.7) and Senegal (US$ 1.2) in Africa; India (US$ 1) and Pakistan (US$ 0.5) in Asia; and Australia (US$ 3.2).”

“In most countries that have less multi-party experience, costs tend to be higher, even taking into consideration elections that have taken place as part of peacekeeping operations, where the cost per elector is highest: Mexico (US$ 5.9), El Salvador (US$ 4.1) and Paraguay (US$ 3.7) can be mentioned in Latin America; Lesotho (US$ 6.9), Liberia (US$ 6.1) and Uganda (US$ 3.7) in Africa; and Russia (US$ 3.7) in Eastern Europe.”

The costs in Kenya are comparable only to very special cases of post-conflict elections like Angola, Afghanistan or Cambodia. They are even higher than those observed in cases like Bosnia-Herzegovina under the Dayton Accords (US$ 8). The high cost does not speak highly of the ECK’s efficiency.

The efficiency of the registration process and activities in the non-election period: Registration of voters constitutes a significant part of the expenses of any EMB. In the case of Kenya, they can be estimated at around 30% of the total electoral cost. As is discussed at length in annex 3.A the ECK has established a system that incorporates the cost of continuous registration while maintaining those of periodic registration. Although the ECK has established a network of 71 district offices, only some 2% of voters opt to register there. The remaining 98% do so at the annual registration drives.

While it is true that the ECK registered 1,767,000 voters during the two registration drives in 2007, the cost per registered voter was KSh 1,233 (around US$ 18), which is extraordinarily high. Although it is difficult to make precise estimates, the cost per registered voter of the 1,078,000 voters registered in the registration drives or through continuous registration in 2003 through 2006 was much higher, probably by as much as 50%.

As mentioned in annex 3.A, the productivity of the district offices is very low in non-election periods. In the first four months of 2008, the network of district offices conducted only 553 transactions: 129 new registrations, 110 transfers, 23 cases of persons missing in the register, 275 deceased and 16 changes in particulars. There was not much else to do in those offices during that period, so we can conservatively estimate that half of the cost of these offices is related to these 553 transactions. Since the cost of these offices for that period amounted to close to KSh 100m (according to the 2007/8 estimates, the annual budget of the district offices is KSh 402m), the cost of each of those transactions comes to KSh 90,000. While this is an extreme example, given that the period analysed was immediately after the general elections, the cost is staggering.

As discussed in annex 3.A, the best alternative for future voter registration is to move from continuous registration to a system that would relate voter registration to other population databases. One of the suggested alternatives would be to register voters when they request their national IDs. It is important to mention that this is not a new proposal. In its report on the 2002 general elections (p. 91), the ECK asserted that voter registration “would be made easier if voter registration was incorporated into the issuance of National Identity Cards. The cost would be reduced and the ECK’s goal of registering all eligible Kenyan voters would be realised. There is need to … [i]ntroduce major amendments to the process of registering persons to include the registration of voters [and to] [f]ormulate a long term policy on centralised registration of persons for issuance of a single identification card.” However, nothing was done and the statement could well be repeated in its entirety in the forthcoming report on the 2007 elections.

Decision-making in the ECK: This section provides a detailed analysis of the way in which the ECK handled a major issue: the counting and tallying of votes at the constituency level. The problems of counting and tallying had already been identified in 2005, and it is worth quoting from the ECK’s Report on the National Referendum Evaluation Workshop (held in March 2006):

“To speed up the tallying process, the Commission should consider procuring computers for this purpose. Those to do the tallying must be computer literate … There should be a mechanism for verification during counting and tallying of votes … [A]dvance planning for communication protocol to be utilized in the 2007 general elections [should] commence immediately” (p. 26).

‘The plenary stressed that the plans to decentralize the computerization process … should be aggressively pursued. In addition, the plan to procure/hire 210 computers for the constituency level should commence immediately in readiness for the 2007 general elections. … It was suggested that there is need to brief the commissioners on ICT activities regularly to enable them [to] understand the technological advancements and make informed decisions” (p. 36).

Nothing much seems to have been done in the succeeding few months, and the topic of computer use for tallying at constituency level re-emerged only in the second meeting of the steering committee on assistance to the 2007 general elections, held on 25 October 2006. The Committee approved the use of assistance funds to procure computers and requested the ECK to submit a list of the required equipment. (An additional use was mentioned in a later ECK funding proposal to UNDP – “voter registration education”.)

Seventy-two days after the October approval, on 5 January 2007, the ECK chairman sent the United Nations Development Programme (UNDP) the specifications for 210 laptops and 210 high-capacity printers, without any description or justification for the proposed use of the equipment. Within the next ten days, UNDP prepared the Invitation to Bid and other arrangements and the Invitation was published on 15 January 2007, calling for proposals to be submitted by 6 February 2007. The proposals received were analysed and, as the ECK specifications were somewhat exacting (2 x 12-hour internal batteries, TV output, and both a touchpad and a pointing nub), 32 out of 33 proposals for the laptops and 53 out of 55 proposals for the printers were rejected.

As the procurement was for more than US$ 1 million, it required the approval of UNDP’s Advisory Committee on Procurement (ACP) in New York. The ACP initially rejected the proposal, indicating concerns about the large proportion of rejected proposals and the lack of justification for the special requirements. Having received explanations, however, the ACP authorised the procurement on 26 April 2007 and the procurement order to the single approved supplier was issued by UNDP on 2 May 2007 for delivery within one month.

Well before that, the ECK had been receiving advice from two external experts, Michael Yard (IFES) and Paul Anderson (UNDP), both emphasising the importance of improving the effectiveness of the transmission and the counting/tallying of results and suggesting a procedure based on the direct transmission of results from the polling stations – a procedure that had been successfully used in the 2005 referendum.

In late May, the ECK IT manager, Ayub Imbira, proposed a revised solution that “eliminates the PO [presiding officer] from the communication of results to ECK HQs and only leaves it to the RO [returning officer]. The following shall be the procedure of communicating the elections’ results to ECK computer servers (for the ‘call centre’).

* The RO receives results from the PO and resolves any disputes.

* The RO, with his/her staff, shall capture these data into the computer provided to them by ECK. Each election’s results are captured individually for each polling station. The RO will also tally the results as has been traditionally done. The emerging totals for each candidate in each election, among other computations e.g. number of cast votes per station, number of spoilt votes, shall be computed by the computer system and comparisons made to manual tallies. Differences should be resolved there and then.

* The data is then transmitted to ECK HQs through a secure telecommunication link.

* The computers servers at ECK HQs will receive these data, accept it into the database and re-transmit a copy of it (for each polling station) back to the RO. This is a handshaking process through which ECK HQs is acknowledging the receipt of the data.

* These results can then be made public.

* A failure in the system for any constituency will lead us to using the traditional system that uses voice telephony. Failures in the system may arise from telecommunication link failure, RO’s computer system failure, inability by the RO and his/her staff to use the computer system and scepticism in the use of the system.”

As can be seen, the system proposed is simple enough and would practically eliminate arithmetical errors. Mr Imbira proceeded to enumerate the main advantages of the proposed solution in relation to the traditional way of doing things:

* “Data is transmitted over a secure telecommunication link.

* Handshaking by the ECK HQs computer system and the ROs’ computer systems provides more authentic results.

* ECK HQs will receive and store elections’ results for each polling station contrary to the tradition of only capturing the final tally for the constituency.

* Duly signed elections’ results per polling station will be scanned and archived.

* Partial elections’ results for a constituency can be made public before all the

counting centres have submitted their results.” A further advantage of the proposed solution is that requirements would be quite modest. The proposal lists them as:

• one computer for each constituency

* software for capturing and tallying the results

* secure telecommunications links (wireless links using GPRS - Safaricom and Celtel or CDMA - Telkom Wireless preferred)

* electricity supply back-up generators

* at least one ICT competent staff member (among the RO’s clerks/staff in each constituency)

It should be noted that the necessary software is not particularly complex. Furthermore, IFES had offered to develop it within the framework of the support provided to the ECK. Similarly, since the assistant returning officers (AROs) and clerks had not yet been recruited, it would have been possible to recruit persons with a basic knowledge of computers and, in that case, their training in the use of user-friendly software would not have taken more than a day or so. The proposal was submitted to the ECK Plenary by the Chairman of the Research and Technical Committee on 14 June 2007 and adopted without amendment. It should be noted that the laptops had been delivered to the ECK three days earlier.

Both Mr. Imbira, the ECK IT manager, and Suleman Chege, now ECK Secretary but at the time an ECK Deputy Secretary, testified that they learned of the decision of the Plenary only by 8 August 2007. According to Mr. Imbira’s written statement, “the tender for elections results tallying software was finalized on 23rd October 2007 and a purchase order raised”. As to “the General Packet Radio Service (GPRS) and the Virtual Private Network (VPN), it was necessary to wait until 30th November”. It seems that the requirement of some computer literacy for the recruitment of AROs and/or clerks, as well as the eventual addition of one or two days so that they could be trained in software use, were not included by the ECK units in charge of recruitment. In the event, the computers and printers were stored in the ECK warehouse until mid-December 2007, when they began to be distributed to the constituency offices. In some cases, they arrived only a few days before election day. It should not come as a surprise that, when the returning officers gathered in Nairobi on 22 December 2007, some complained about the problems they would have in using the computers, and the decision of 14 June 2007 was revoked in circumstances that are unclear. Since the ECK consistently failed to provide IREC with minutes of its Plenary meetings, IREC could not establish the status of this decision.

This analysis, although far from complete, provides some glimpses into the internal functioning of the ECK. In any organisation, when a decision is made by the top decision-making organ, it is supposed to be implemented without further ado by the rest of the organisation. Clearly, this was not the case with the computerisation of the results transmission and tallying process.

The need had been perceived after the 2005 referendum, but nothing seems to have been done for several months, in spite of the approaching 2007 elections. The advice of external experts was not taken into account. However, an adequate solution was developed internally and approved in time. When the proposal was adopted by the Plenary on 14 June 2007, enough time remained to ensure that computers could be used for tallying results at constituency level in the December elections: this required only the full support of all the relevant units in the organisation – procurement, personnel, training, etc. But nothing seems to have worked efficiently. There were delays in informing the officials concerned that the proposal had been approved, and all the necessary processes seem to have proceeded at a snail’s pace.

If the proposal had been implemented, it would have been possible to issue partial results for all constituencies in an adequately programmed sequence, which would probably have eliminated most of the tension build-up created by the use of a slow, error-prone and old- fashioned approach to the tallying and relay of election results.
Conclusions/findings

The overall picture of the ECK depicted by these three partial analyses does not show an efficient organisation which plans in advance and implements crucial decisions expeditiously and transparently. The image is the opposite: a traditional organisation, with inadequate flows of information, averse to even minimal risks and to the use of technology, functioning in a compartmentalised fashion. Whatever the decisions to be made concerning the electoral system and other aspects of the electoral process, one thing is sure: the ECK’s internal management processes deserve a thorough overhaul.
Recommendations
On the ECK’s legal framework

IREC recommends that urgent consideration be given to drafting and enacting more detailed provisions on the ECK’s institutional aspects, taking into account the relative complexity of the organisation and also its responsibilities, which have increased exponentially since the early 1990s. These provisions could be added to the existing National Assembly and Presidential Elections Act (as sections 3-3B have been) or contained in an Electoral Commission of Kenya Bill or in a consolidated electoral law, such as the Electoral Bill currently being drafted by the Kenya Law Reform Commission (KLRC).
On the composition of the ECK and the appointment of commissioners

IREC recommends that the requirement for a broad consultative process prior to the appointment of ordinary members and the chairman of the ECK be given legislative grounding; consultation should include political parties and civil society in its broadest sense. It is not essential to any purpose that the actual appointment be made by the President. Alternative means should be considered, including appointment by Parliament.

IREC recommends that the maximum number of commissioners be reduced to such a number as are functionally able to do the work. The currently bloated structure at the top should be trimmed radically.

IREC recommends that expiry of the terms of office of ECK members not coincide with election years. Ideally, a fully composed commission should be in office for two years prior to the conduct of general or presidential elections.

IREC recommends that the ECK be made accountable to Parliament, without prejudice to its status as an independent body; this should affect the channels by which it establishes and seeks approval for its budgetary requests.
On the ECK’s operational procedures

IREC recommends a review of the ECK’s administrative procedures, with a view to introducing as much certainty – and as little discretion – as possible in key operational areas so as to ensure uniformity of performance throughout the electoral process, from polling station level up to the various teams at the national tallying centre. These procedures should be codified, well-known to staff and form the basis of training.

IREC recommends that the ECK review its overall training/briefing procedures in order to improve on their utility for adult learners. Training should, as a standard and not good-to-do measure, include simulations with the tools of work that the election staff are being prepared to use in the forthcoming elections. Commissioners and senior staff should all receive basic training in election management, such as Building Resources in Democracy Governance and Elections (BRIDGE) or Basic Election Administration Training (BEAT), at the earliest possible opportunity after joining the ECK and also receive such refresher training as may be necessary in the course of their service to the ECK.

IREC recommends that, in the recruitment of temporary election officers, commissioners should themselves vet only returning officer recruitment; they should establish clear procedures for the recruitment of other staff and ensure these are strictly implemented.
On the structure of the ECK Secretariat

IREC recommends an urgent re-examination of the roles of commissioners vis-à-vis those of staff, with a view to establishing a clearer commission-management separation of roles. Once clear lines of authority and responsibility have been established, investment in staff training should be increased with a view to vesting in the Secretariat much of the institutional memory needed to conduct a genuine election. Greater, if not all, implementation responsibility should be delegated to the Secretariat while greater, if not exclusive, policy-making responsibility should remain with the Commission.

To ensure accountability, IREC recommends the establishment of clear lines of individual responsibility for service delivery among both commissioners and staff.

IREC recommends that the ECK operating structure be rationalised to reduce time devoted to committees. Some committees can be consolidated or even abolished altogether. Their procedures should also be clarified, clear terms of reference should be drawn up for each committee and minutes should be kept for institutional memory.

IREC recommends that the district offices be disbanded. Instead, the ECK should decentralise only down to the provincial level, since eight well-equipped provincial offices would be able to function both at election time and between elections.. The provincial staff should include the right mix of skills, including information and communication technologies (ICT) literacy.

IREC recommends that the ECK revise its secretariat structure with a view to introducing performance management across the board. Staff should be on three-to-five-year contracts, renewable on the basis of performance. They should have clear job descriptions and the space to perform their roles on the basis of the responsibilities set out therein. Annual performance appraisals and performance-improvement measures should also be put in place.

IREC recommends that the Commission Secretary, among other qualifications, be an experienced election manager, competitively recruited from the open job market and have the status of Permanent Secretary.
On ECK funding modalities

IREC recommends that the issue of the ECK’s expenses being a charge on the Consolidated Fund be considered carefully in the constitutional review process, which is expected to begin soon. Any measures agreed in that process to deal with the question of the financial independence of constitutional commissions would, of necessity, apply to the ECK.
On the ECK’s advisory role

IREC recommends that in the constitutional review debate, enumeration of the ECK’s roles should include advising the government, Parliament and other stakeholders on electoral law reform.

Though the ECK’s role will be advisory, IREC recommends that its advice henceforth be taken more seriously than has been the case in the past and that the relevant institutions charged with law reform proceed to deliberate such proposals expeditiously with both the ECK and other stakeholders, with a view to incorporating them into the law.
On the ECK’s functional efficiency

IREC recommends that a new or transformed ECK undertake a thorough management systems review as a matter of urgency. Such a review should aim to tailor the institution to implement critical decisions expeditiously and transparently.