EUROPEAN UNION
DELEGATION OF THE EUROPEAN COMMISSION
IN THE REPUBLIC OF KENYA
Head of Delegation

1 Introduction

The primary objective of the EC’s Somalia Country Strategy Paper (and Cotonou Agreement) is to support the move towards the establishment of democracy and good governance in Somalia. In much of the country this is still a long way off, however in Somaliland the period of transition and the movement towards elected administrations is now well under way.

Somaliland (SL) is in a process of transition from a traditional system of governance with a Parliament and executive being appointed by traditional leaders who represent clanic groupings to a more democratic system. The traditional system was centred on large clan conferences selecting leaders and providing the mandate for government. In early June 2001 the people of SL voted in a referendum on a new constitution. The constitution was overwhelmingly accepted, with the first public vote to have taken place in Somalia for 35 years, setting a valuable precedent.

The referendum was conducted almost entirely without external assistance and clearly had shortfalls. Elections are a much more complex, and within fragile post conflict societies, can result in renewed conflict. On this basis and in line with the July EU Council Conclusions on Somalia, the EC joined forces with other donors in order to initiate activities to assist this transition to democracy.

The EC fielded a short consultancy to look into the feasibility of electoral support, assisting the Somaliland Electoral Commission (SEC) in drawing up together a work plan to guide their future activities. The positive feedback from this initial consultancy highly encouraged the fielding of a project. At the same time many of the consultant’s recommendations were adopted by the SEC and the election date notably changed from the 27th of October to the 15th December to accommodate the necessary work.

1.1 The process

The local elections would generate elected Councillors who in turn will appoint Mayors for the 23 districts. The importance of these elections is that the currently registered political organizations would compete at this level to become political parties. To do so, parties must solicit at least 20% of the votes in four of the six regions; the first three organizations with the highest percentages of votes would become parties eligible to contest presidential elections. These difficult selection criteria were designed to prevent clan-dominated parties. It can be assumed that any organisation meeting these criteria can be seen as cross-clan supported.

The most formidable challenge facing the electoral commission was the registration of voters. There was an absolute minimum chance to do this in time given the absence of any form of registration of citizens at regional level, the distribution of people over large rural areas and beyond the borders of Somaliland and indeed Somalia (criteria being born in Somaliland or of Somaliland parents).

In addition, there are various problems facing political organisations hoping to enter the competition where time was a crucial factor. Chief of these being little or no experience in setting-up parties, and more dangerously, no funds or guidance in fund raising, which expose the situation to possible external manipulation.

1.2 Intervention timings

- SEC requested support in May 2002;
- A consultant (Dr Franciso Cobos Flores) was fielded in early September 2002;
- 25 September, Member States Meeting to review consultant’s findings;
- 27 September consultant’s report accepted;
- 11 October Member States meeting to review GTZ project proposal;
- 26 October GTZ technical assistance mission (TAM) in support of SEC fielded (2 electoral specialists);
- 15 December, local elections were held;
- 31 December project activities ended.

1.3 Donor contributions (details to be confirmed by GTZ):

EC			€450,000
Denmark		€125,000
Finland	      €10,000
Netherlands		€50,000
UK			£60,000
Switzerland		€50,000

The SEC comprises 7 members, selected in the following manner:3 by the President, 2 by the upper House of Elders (Gurti) and 2 by opposition parties.

The composition of the SEC was then authorised by a two-thirds majority in Parliament and the Gurti. The seven members are employed each with a salary of $300 per month. In addition there are five administrative staff, each of whom was selected by the civil service commission. None are former government employees. The team are all well motivated for the task, with one of the commission members being a woman.

1.5 Local elections

It was anticipated that counsellors for the 23 districts are elected (Mayors are subsequently elected by the counsellors), with the districts divided into five categories, depending on their populations. This has been done as follows:

In the capital Hargeisa, 25 individuals were to be elected; in category A (primary urban centres), 21 individuals; in category B, 17 individuals; in category C, 13; in category D, 9.

It became clear, given the prevailing security violations in Sool region (disputed with neighbouring Puntland) and the District of Buhodley that elections could not be conducted in all 23 districts. The decision to cancel elections in three districts was taken by SEC and is commendable.

2 Lead up to Election Day

Riyale (the current president of Somaliland) went out of his way in his pre-election speech, 13/12/02, to differentiate "parties from the nation", this in response to criticisms levelled against elements of his own party UDUB who continued to use the trappings of office to campaign. The vice president issued a directive at this time for all administration vehicles to be turned over to the electoral commission to further facilitate the running of the local elections.

2.1 Transparency of SEC accounting and government contribution. The arrangements for accountability within the SEC were limited and accounting basic, yet donors required sound accounting practices and this requirement was considerably enhanced by the TAM.

Simple cash flow problems were encountered, particularly as the project worked on a reimbursable basis. This worked, but limited the ability of the project to become fully engaged to the level originally anticipated, resulting in an anticipated under spending (to be confirmed in detail in the final financial report by GTZ). In order to enable smooth running on election day and to ensure the project could remain engaged to the level anticipated, a 50% advance was agreed for election day costs. However the point was emphasized that these are their elections and it is the duty of the administration to ensure sufficient funds are available. In terms of management the SEC contracted an external auditor to perform a review of their accounts.

2.2 New management structure

The TAM had put in place a management structure, to take on the majority of work; this unit reported directly to the SEC, but worked under the guidance of the expatriate team. There were difficulties as SEC members did not respect the line of command and often dealt with the field directly, which at times caused misunderstandings. However, by being in place much of the day-to-day practical work was sufficiently covered.

The management unit gave a presentation on the preparations, which were in place, and the activity plans for the actual day of elections. The presentation was of good quality and clearly demonstrated that the TAM had been able to instil a decent level of understanding. It is worth mentioning that through the project 3,200 election workers have been trained, 400 domestic observers and basic but effective civic education has been conducted.

It is questionable if the unit will remain in place for the coming elections in 2003 as the SEC claim they do not have a sufficient budget for the work. The TAM strongly advised that this is addressed urgently.

2.3 Number of voting days

It was planned to conduct only one voting day, however, the TAM stated that statistically it was highly unlikely that all voters would be able to vote in such a limited time. The electoral law states that one-day only was possible. It was agreed in a meeting with all political parities that one-day would be maintained, despite the constraints.

2.4 Registration

Attempting registration showed the frailties of the voter card system. Registration was only done in urban areas, with around 330,000 people being registered. The figures are not yet compiled and there were clear irregularities in the process. These were recognized and lessons have been learnt from the process. On polling day all people were allowed to vote irrespective of whether they had a registration card or not (and then marked with indelible ink), this was the only feasible solution given the circumstances.

2.5 Security

It was recognised that security deteriorated in Sool region and as a result elections were cancelled in 2 of the 4 districts and Buhodley District. 2,283 police provided security on polling day and to escort and collect sensitive materials to all the voting locations (ballot boxes, sheets and indelible ink).

3 Election day

All polling stations opened at 7.00 am and were to close at 6.00 p.m. During the day it became apparent in the primary urban areas that closure at 6.00 would not be possible, leading to a decision for polling stations to remain open until 12.00 or until lines of voters were finished. However in reality most stations were closed by 8.00 pm as the voters were no longer lining up or ballot papers were completed (although in several cases re-supply was provided from strategic reserve stocks).

In no cases did any of the internationals that visited polling stations witness any violence, systematic fraud, open campaigning or blatant organizational and logistics failure. In total over 100 polling stations, in Burao, Sheik, Berbera, Hargeisa, Gabiley, Kalabyd, Erigavo, Boroma, Togwagale and many villages in-between were visited by internationals.

In all polling stations visited, police were present, orderly lines maintained, ballot papers, boxes and indelible ink were present. The atmosphere was highly positive and the majority of adults voted. Many interviews were conducted, and without exception voters were happy with the way they had been treated. Understanding of the process was relatively high in urban centres, reflecting the focus of the civic education activities, however in rural areas not all people actually understood why they were voting, with a number thinking it was the presidential elections.

3.1 Polling station staff.

Without exception every staff member interviewed had received two days training, allowing basic principles and mode of operation to be functional. There were some clear shortfalls, but in the opinion of all internationals present these were not manipulations of the process but a reflection of Somali culture and the limited knowledge within society of the democratic process. Only in one polling station visited was there any discontent of the process – a result of several voters attempting to vote twice. They were however discovered – in itself a reflection of the fact that the process was functional.

3.2 Observers

In every station visited the contesting parties had representatives observing the process. These representatives remained in place for the whole day. Again, when asked they all felt the process was functional and accountable.

In around 35 – 40% of stations domestic observers were present from COSONGO and NAGAD (the main local NGO umbrella agencies). These observers had received one day’s training organised by the project. In a number of cases they were clearly well informed, in others not.

Cases of both the party and domestic observers becoming actively engaged in organising the election day were reported. This was, in all cases, harmless but is a contravention of normal election procedure.

3.3 Secrecy of the vote

The biggest weakness observed was the fact that secrecy of vote was not functional. This was clearly a reflection of reality. The majority of voters are illiterate. Political parties are known by their names, but not so much by their logo. This meant that people were totally confused by the ballot papers. People seemed more than willing to express who they were going to vote for if this was "required" to speed the process. This reflects on the general understanding of what the elections were all about but, as the results below illustrate, people did vote for all the parties running so overt coercion does not appear an issue. The ballot papers were then completed by the election staff or the representative of the selected party. If voters were literate they were permitted to vote behind curtains.

3.4 Eligibility to vote

As the registration process was only partially conducted cards were not utilised as the price of submission. Essentially everyone who wanted to vote was allowed. In no cases (except for the under age) was anyone turned back. In all cases those who had voted were marked with indelible ink, which was not possible to remove.

3.5 Role of Media

ICD/CIIR provided journalist training specifically in relation to the elections. The training was conducted in a 3 day workshop and was partially funded by the project as well as ActionAid and PENHA. Radio, TV and print journalists were given training, building on the journalist training conducted by the BBC in 2001 (EC funded).

It appears that the overall output was a success, with media of all types trying to present an unbiased view. The example of Radio Hargeisa on election day was outstanding with Journalists keeping a running commentary on the day’s events, with constant reminders for people to be peaceful and calm. At many polling stations the radio was being broadcast to those waiting in line. All Somalis interviewed in Hargeisa were impressed by the radio coverage.

In addition, two documentaries for local consumption will be produced of the lead-up to the elections and the events of the day. These were partially funded by the project and will provide useful civic education tools in the future.

3.6 Participation of Women

Women’s participation was impressive. At each polling station large numbers of women were present, while around 20% of staff working were women. The percentage of women voters was much higher in the afternoon (primarily reflecting women’s higher work load and the fact that many men chew Quat after lunch). The number of women candidates does appear to be very limited, however hard facts on this are not available, with only single figures being reported to have been included on party lists of candidates.

3.7 Review of polling day

A meeting was conducted on the 17th in which all internationals who were directly or indirectly involved in the events of the day met. In all cases each group reported positively on the day’s proceedings, being impressed with the level of organisation and commitment shown by all involved.

Those present in meeting:

Paul Simkin (Chair) EC,	 Paul Crook, EC
Manfret Gers, TAM / GTZ, Elis Suedjem, TAM/GTZ
Emman Sandahl, (Student) Nick Selsey, TAM/GTZ
Adan Adoker, ICD,  Steve Kibble, CIIR London
Mark Bradbury, ICD,  Hussein Bulhan	APD
Hugh Fenton, DRC, Henrik Jespersen,	Danish Embassy, Nairobi.
Mat Bryden, War Torn Societies	
Owen Richards, British Embassy, Addis Ababa.			
	

The results took longer than anticipated to be compiled. Counting was scheduled to take place at each polling station. This did not happen, partially due to security concerns and simple problems of no lighting. At the same time the counting process employed was painfully slow. There is a need to review the procedure and ensure that each polling station is of a sufficient size, in a safe area and has simple things such as lighting, water and toilets.

	Regions							
 Political org Hargeisa  Sahil Togdher Awdal Sanag Sool TOTAL	Percentage
1. UDUB  70,989  13,502 18,330 58,939 16,574 1,055 179,389 40.76%
2. Kulmiye 29,923 5,309 17,476 13,679 13,701 3,070 83,158  8.90%
3. UCID  30,676   2,900 4,821  7,422  3,401  224   49,444  11.24%
   Sahan 14,748   2,054  15,234 4,499  11,356 51   47,942  10.89%
   Hormod 29,104  1,188  1,454  7,229  1,409 154   40,538 9.21%
   Asad   10,943  2,281  9,283  8,727  6,655 1,707 39,596 9.00%
   TOTAL 186,383 27,234 66,598 100,495 53,096 6,261 440,067	100.00%

There is an east-west drift with UDUB growing weaker as you move eastward. This was expected but is not so dramatic as to give cause for concern with power blocks opening along clanic lines.

During the referendum it is reported that 1.18 million votes were cast. These were at the time estimated to be highly inflated figures. The table below demonstrates a variance of over 700,000. It is true to say that an estimated 70% of eligible voters cast their vote during the local elections, which would give a voting population (all over 16) of around 628,000; even if 50% managed to vote (low estimate) the voting population would be in the region 880,000 or slightly higher since elections were not conducted in every district. Based on these figures Somaliland’s population could be very vaguely estimated at around 1.7 – 1.8 million (taking 50% under the age of 16 and a voter turnout of 50% with some allowance for non voting districts).

Region votes Referendum valid votes Local election valid votes Variance
 	 	 	 
Awdal	   180,473	 100,495	79,978
Hargeisa 402,558	 186,383	216,175
Sahil	   110,406	 27,234	83,172
Togdheer 224,159	 66,598	157,561
Sanaag   162,078	 53,096	108,982
Sool	   103,185	 6,261  	96,924
Total	   1,182,859 440,067    742,792

How the political future of the three parties will go forward will be of note and requires further work on the development of party politics (something USAID started to develop through the International Republican Institute). In a more sophisticated system we could be looking at transferable voting or other such techniques. Presently we wait to see how personalities look to deliver what support they have to the surviving three parties. The system must look at whether the losing organisations can stay in existence and work in the local government systems.

Questions now have to be raised as to the ability of the newly incumbent local councils to take forward issues of local government. The understanding regarding the role of these elections is also reflected in people's ability to take forward local issues and develop the community/government linkages required to draw out issue-based politics in the future. There is a clear need for training so as those elected actually understand their own responsibilities.

The domestic observers group was a welcome addition, but as demonstrated by the report produced by COSONGO the product is of questionable quality. It would be unreasonable to expect a high quality result, given the fact that observers were only given one day’s training and only very limited financial recourses were made available. There is clearly potential for much more comprehensive work in this area, building on the limited, but positive, success achieved to date.

A longer term problem remains in that district boundaries are not clearly demarcated resulting in the closure of one polling station. Although not a major problem at this moment, it will present difficulties, which will grow as the democratic system progresses.

There continues to be a need for voter and civic education, particularly in rural areas where radio coverage does not reach. This in conjunction with the high percentage of illiteracy needs to be taken into account for planning of future elections. Without tackling these problems it is unlikely voter secrecy will be achieved.

In urban areas it was clear that the number of polling stations or period of voting (only one day) was not sufficient, with significant voters turned away, while others were forced to stand in line for hours at a time. Since these were the first elections for the vast majority of the populace, they have not presented a disincentive to democracy, however it is highly unlikely that the same level of motivation will remain in place in the future unless the problems are tackled.

6 Conclusion

It is clear that the process was of as high a quality as is realistic within the prevailing environment. It is highly unlikely this would have been achieved without donor support and technical assistance.

The electoral commission released a communiqué on 16/12/02 which is both bold and open - Item three, for example, states "The early realisation, by the Commission members, of their multiple shortcomings and their subsequent bold stance to seek advice......”Such an attitude is noteworthy as donors engaged to bring in the GTZ team.

It was emphasized that if the SEC expect to receive assistance for the next election phase they will need to clearly prepare requests and have them circulated in the near future. This is a priority. It was also made clear that the EC would not consider further funding beyond the existing project (a point made from the outset of EC dialogue with concerned parties).

The Government will provide a high level of resources to the presidential elections (now scheduled to be held on 30 March 2003) simply because they are considered more important and will be watched closely by the international community. These elections are constitutionally supposed to take place by February, however it is likely that March or April will be the date recommended by SEC and approved by Parliament.

Many of the lessons learnt and experience passed on by the TAM will provide a basis against which future elections can be organised. Sadly it is unlikely that the same level of discipline and organisation will be possible without external and impartial advice being on hand.

The combined support provided to the elections in many respects was the highest profile project undertaken to date in Somaliland, and yet the most risky. It is a commendable success to the project team and the Electoral Commission that the local elections passed off without incident (possibly one of the most peaceful elections in Africa for the past 20 years). The project has clearly paved the way to helping with the growth of democracy in Somalia. It is however only one small step and considerable support is required for the future, in terms of leaders training, civic education and continued advice and assistance.