Progressio is the working name of the Catholic Institute for International Relations (CIIR). CIIR is registered in the UK as a charity (number 294329) and a company (number 2002500). c Progressio 2006 ISBN 1 85287 318 3 Adan Yusuf Abokor, Steve Kibble, Mark Bradbury, Haroon Ahmed Yusuf and Georgina Barrett assert their right under the Copyright, Designs and Patents Act 1988 to be identified as the authors of this work. Cover photo: Women queuing to vote. Photo: Georgina Barrett.

About the authors:

Adan Yusuf Abokor is the Progressio country representative for Somaliland.

Steve Kibble is Progressio's advocacy coordinator on Africa/Yemen.

Mark Bradbury is a writer and analyst on Somaliland and East Africa.

Haroon Ahmed Yusuf is a programme advisor for ActionAid in Somaliland. Georgina Barrett is an International Peace and Security masters graduate with a keen interest in African affairs, particularly in Somaliland and southern Sudan.

Acknowledgements

Progressio wishes to thank the Government of Somaliland, the National Electoral Commission, the Special Protection Unit of the Somaliland Police Force, the political parties and their candidates, and in particular the people of Somaliland for their hospitality, openness and patience throughout the period that the international observers were in Somaliland. Progressio of course also extends its thanks to the international observers themselves for their commitment and good humour throughout the time they were in Somaliland. It also gratefully acknowledges the logistical support provided by international non-governmental organisations (NGOs) which made it possible for the observer team to travel throughout Somaliland. Finally Progressio wishes to thank the British Embassy in Addis Ababa for funding the international election observation team, and indeed this publication.

Note on terminology

The Catholic Institute for International Relations (CIIR) adopted Progressio as its working name in January 2006. The name CIIR has been used throughout this report because this was the name used by the organisation during the period covered by the report.

Introduction

Somalia galvanised Somaliland's government into holding the constitutional referendum in 2001. The subsequent elections and the demonstrated commitment of Somaliland's people to peaceful democratic practices of governance presents an enormous political challenge to Somalia's TFG and a dilemma for those in Somalia and the international community who oppose the aspirations of people in Somaliland for independence. At the same time the three elections demonstrate what can be feasible elsewhere in Somalia, should conditions allow.

The elections also have regional connotations. The peaceful manner in which the polls were conducted contrasts with the violence that accompanied elections in neighbouring Ethiopia in 2005. They also provide inspiration for civil activists advocating democratic change elsewhere in the region, such as in Djibouti and Eritrea.2 Some analysts regard Somaliland's experiences and achievements as a model for the new government of South Sudan.3

CIIR and the international election observers

CIIR (known in Somaliland as International Cooperation for Development - ICD4) has been active in Somaliland since 1995 (see box). In December 2002 and March 2003 it organised a small team of international observers and supported local NGOs to monitor the council and presidential elections. In July 2005, the Somaliland National Electoral Commission (NEC) formally invited CIIR/ICD to organise and facilitate a team of international election observers (IEOs) to cover the parliamentary elections.

CIIR/ICD fielded a small team to observe the preparations for the poll and the parties' campaigns and to monitor media coverage. The elections themselves were witnessed by a team of 76 observers from 20 countries, namely Kenya, Ethiopia, South Africa, Mozambique, Zimbabwe, the Philippines, Germany.

On 29 September 2005 the people of the Republic of Somaliland, an internationally-unrecognised country in north-west Somalia, elected a new parliament. These parliamentary elections, the first to be held in the Somali region since 1969, were the latest and, arguably, most important step in establishing a constitutionally-based, democratic governmental system in Somaliland.

Since breaking away from Somalia in May 1991, the people of Somaliland have sought to build a new state by charting a path away from violent conflict to a competitive and democratic political system. The process began with a constitutional plebiscite in 2001, and since 2002 all of Somaliland's key political institutions - district councils, the presidency and vice presidency, and, with these latest elections, parliament itself1 - have been subjected to popular vote. The successful staging of these three elections has important implications for Somaliland and the political entity (or entities) that emerge from the wreckage of the Somali state, and for the region in general.

These various elections have consolidated Somaliland as a territorially-defined political entity, with all the regions within its borders having elected councils and elected representatives in government. The introduction of universal suffrage and the creation of political parties that are not based on clan mark a significant attempt to move away from the kinship-based politics of the past two decades. Furthermore, the establishment of an elected parliament has the potential to restore a more equitable balance to political authority, by curbing the excesses of the executive and the increasing corruption of political life that had begun to corrode the political project in Somaliland.

The process of establishing an elected government in Somaliland has occurred in parallel with regional and international efforts to restore a national government to Somalia. This has involved the formation of the Transitional National Government in 2000 and the Transitional Federal Government (TFG) in 2004.

Since 1995, CIIR has worked alongside nascent local NGOs (LNGOs) which have played a crucial role in providing services to marginalised communities in Somaliland. Through its skill-share programme CIIR provides experienced international development workers, including those from the diasporas, who work directly with partner LNGOs. CIIR has assisted its LNGO partners by directly building their capacity to work on issues concerning education, health, disability, young people, human rights and HIV and AIDS in Somaliland, and by supporting minority groups, including youth and disability groups and women's organisations, in advocacy initiatives. Partner LNGOs work directly with government officials, other NGOs, community and religious leaders, respected elders and other members of the community and have helped guide Somaliland towards addressing issues around poverty, reconstruction and democratisation.

A week prior to the poll, the peaceful election campaign was interrupted by a serious security incident, involving a gun battle between police and a group of alleged Islamic militants in Hargeisa, Somaliland's capital. Initial fears that this marked a plot to disrupt the elections were assuaged when it was discovered that a plan to free prisoners awaiting trial for the murders of foreigners in Somaliland was behind the incident. Several of those detained by police have since been convicted of the murders, and sentenced to death.

The elections themselves and the observation mission were not affected, and the IEOs went on to visit 361 polling stations around Somaliland on the polling day itself (over one-third of the total of 982 stations).

Election day saw a turnout of over 600,000 voters. The IEOs were able to conclude that despite many procedural problems (lack of a census and registration process; breaking of the voluntary code of conduct by political parties; a lack of secrecy in some ballots; complicated ballot papers; attempts at multiple voting; and unequal representation of women) the elections were conducted in a `reasonably free and fair' fashion.5 Within days of the poll, but before the final results were confirmed, CIIR issued a report to this effect to the NEC (other missions among the IEO team also issued their own assessments of the elections).

This final report provides a more considered assessment of the elections, the results and the democratisation process in Somaliland. While the report incorporates some recommendations and insights of other IEOs, it does not necessarily reflect their opinion, and its authors assume responsibility for the analysis and for any errors of fact or interpretation.

1: Democratisation in Somaliland

had fled to neighbouring countries have returned. The relative stability sustained over the past decade has enabled international aid organisations to support local efforts to restore essential services and infrastructure, clear land mines, reintegrate displaced populations, demobilise militia and build the capacity of new civil society organisations. Human development indicators in Somaliland are generally better than in other regions of Somalia, although levels of under-five and maternal mortality, adult literacy, primary school enrolment and per capita income remain among the lowest in the world.10

Constructing a new state has brought many challenges. Between 1992 and 1996 the country was twice embroiled in civil wars. A ban on imports of Somali livestock by Saudi Arabia since 2000 has deprived the country of a key source of revenue. Demographic and economic pressures are affecting the environment and fuelling rapid urban migration, which in turn is straining the capacity of urban infrastructure. There is also evidence of growing disparities in wealth between social groups, between the east and west of the country, and between urban and rural populations. Critically, after 14 years Somaliland's sovereignty claim remains unrecognised by Somalis in Somalia or any foreign government and is contested by people in eastern Somaliland.

The lack of international recognition has deprived people in Somaliland of the type of governance support that many post-conflict countries receive. It also restricts the possibilities for developing international trade relations and encouraging inward investment. With meagre levels of international assistance, recovery has largely been achieved from the resources and resourcefulness of Somalilanders themselves. The main sources of finance have been the production and export of livestock, trade, and the remittances sent by Somalis living abroad. This has served to forge a separate identity and a feeling of self-reliance and has enabled Somalilanders to craft a political system suitable to their needs.

Unrecognised internationally, Somaliland has many of the attributes of a sovereign state, with an elected government that provides security for its population, exercises some control over its borders, retains stewardship over some public assets, levies taxes, issues currency, and formulates development policies. It has also adopted many of the symbols of statehood, including a flag, its own currency, passports and vehicle license plates. The latest phase in this process of statebuilding has involved submitting its legislature to democratic elections.

The Republic of Somaliland

The `Republic of Somaliland' was created on 18 May 1991, when leaders of the Somali National Movement (SNM) and elders of the northern Somali clans meeting at the `Grand Conference of the Northern Peoples' in Burco bowed to public pressure and revoked the 1960 Act of Union, which had joined the former colonial territories of Italian Somalia and the British Somaliland Protectorate into the Somali Republic.

The new state incorporated the territory of the former British protectorate, whose borders with Djibouti to the west, the Gulf of Aden to the north, Ethiopia to the west and Somalia to the east were delineated by international treaties between 1888 and 1897.6 This includes the former regions of north-west Somalia - Awdal, Woqooyi Galbeed, Togdheer, Sanaag and Sool.7 A sixth region - Saaxil - was created in 1996. Hargeisa became the designated capital and the seat of government. Since 1998, Somaliland's authority over eastern Sanaag and Sool regions has been contested by Puntland State of Somalia in north-east Somalia.8

The people of Somaliland are ethnic Somali and share with other Somalis a common language and religion (sunni Islam), and a traditional livelihood system based around nomadic pastoralism. There has been no census in Somaliland for almost 20 years. War, displacement, urban migration and the nomadic nature of the population make it difficult to determine the population with any accuracy. Current estimates vary greatly, from 1.7 to 3.5 million, with an urban population of between 748,000 and 1.2 million.9

The SNM leadership inherited a region devastated by a 10 year war in which tens of thousands of people were killed, half the population was displaced, the main cities of Hargeisa and Burco were destroyed, and the countryside was left littered with mines and unexploded ordinance.

Over the ensuing 14 years, the people of Somaliland have successfully managed a process of reconciliation, demobilised fighters, restored law and order and created a constitutionally based government. Somaliland has emerged as the most stable political unit within the territory of the former Somali Republic and since 1997 it has been one of the most peaceful areas in the Horn of Africa, experiencing seven years without internal conflict or military confrontation with neighbouring countries. A brief military clash with Puntland in late 2004, in which 100 people were killed, highlighted the dangers of unresolved issues. Nevertheless, the general state of security prevailing in Somaliland means that it has ceased being a refugee-producing area, and most of those who

Democratisation and state-building

The parliamentary elections of 29 September 2005 fulfilled a long term commitment of people in Somaliland to establish a democratic state. This commitment was articulated in the manifesto of the SNM and fought for in their armed struggle against the military regime of Siyad Barre. Secession in May 1991 created an opportunity to break with the past 21 years of corrupt and repressive dictatorship and to establish a democratic system of government. The SNM leadership was given the mandate to govern the country for two years as the sole party, to draft a constitution and to prepare for democratic elections. This was an ambitious and unrealistic task and one on which the SNM government, bereft of revenue and any external support, failed to deliver.

In 1992 the political consensus began to fracture and sporadic conflicts erupted in Burco and over Berbera port. As the country slipped into war, clan elders, backed by civil society activists and business people, stepped in to restore order through a series of locally managed and financed clan peace conferences (shir beeleed).11 These meetings culminated in the 1993 Borama conference, which proved to be a watershed in Somaliland's political recovery. The conference, which lasted several months, oversaw the peaceful transfer of power from the SNM government to a civilian government headed by Mohamed Haji Ibrahim Egal.12 It also produced a Transitional National Charter and Peace Charter that defined the political and institutional structures of government for a three year period, until a constitution could be adopted.

One of the important legacies of the Borama conference was the way it institutionalised the role of clans and their leaders in the system of government. The political system established in 1993, which became known as the beel (or community) system of government, fused indigenous forms of social and political organisation with westernstyle institutions of government. Described as a `dynamic hybrid of western form and traditional substance',13 government consisted of an executive president and a bicameral parliament, comprising an upper house of elders (Golaha Guurtida) and lower house of representatives (Golaha Wakiillada), whose members were nominated on a clan basis by an electoral college of elders.

This clan-based power-sharing system provided the basis of government for eight years and, despite two years of civil war between 1994 and 1996, brought a high degree of stability to Somaliland. During this period the country experienced considerable economic growth, refugees returned and were reintegrated, and much of Somaliland's urban infrastructure, municipal services and systems of education and health were restored.

The civil war was brought to an end in 1997 by a second national reconciliation conference held in Hargeisa. The conference extended the government's period of office for a further five years and re-elected Mohamed Ibrahim Egal as president. The conference also replaced the National Charter with a draft constitution which, after years of wrangling between President Egal and the parliament, was eventually subjected to a plebiscite on 31 May 2001. A major impetus for holding the referendum was the formation of Puntland State of Somalia in 1998 and the Transitional National Government of Somalia in August 2000, which both threatened Somaliland's sovereignty.14 The first article of the constitution affirms Somaliland's independence status, and its approval by the majority of the electorate in Somaliland provided a clear statement of people's aspiration to protect that.15

The adoption of the constitution also put in place the necessary steps to change the political system in Somaliland from a form of clan-based consociational government to a restricted form of multi-party democracy. The constitution introduced universal suffrage, enshrining the right of women to vote. It provides for local government and the national legislature and executive to be subjected to regular elections. But it also restricts the number of political parties to three. The restriction on parties is intended to promote the formation of national organisations that will represent a cross-section of the population and avoid the sectarian politics that occurred in 1969 when over 60 parties contested the election.

Although the public overwhelmingly endorsed the constitution, there was considerable unease about the move to multi-party democracy. Critics of the beel system blamed it for entrenching nepotism, stifling issue-based politics and excluding women from politics, while its defenders argued that it was unnecessary to change a system that had worked for 12 years - a period longer than Somalia's previous experience of democratic government. President Egal linked the adoption of the constitution to Somaliland's desire for international recognition, stating that the international community would not recognise Somaliland's independent status unless it adopted a democratic system. His critics were concerned that it was simply a ruse by his administration to extend its stay in power. Egal's sudden death in May 2002 changed the context and persuaded former opponents of the changes to participate in the electoral process.

The change from a system of selected representation to elected representation in Somaliland's political institutions has occurred in three steps. The first involved the election of 23 district councils in Somaliland's six regions in December 2002. Six political organisations fielded candidates in these elections, of which three won sufficient support to become accredited national parties: UDUB (United Democratic People's Party), Kulmiye (Unity Party) and UCID (Justice and Welfare Party).

UDUB, founded by the late president Egal and headed by his successor Dahir Riyale Kahin, was the clear winner receiving 41 per cent of the total votes, with Kulmiye second and UCID third.16 Elections for the offices of president and vice-president followed four months later, in April 2003, at which the incumbent president, Riyale Kahin, succeeded in retaining his seat, beating his closest rival Ahmed Mohamed `Silaniyo', chairman of Kulmiye, by a margin of only 214 votes.17

The elections brought a number of changes and challenges to Somaliland. Somaliland demonstrated that it had the capacity to hold peaceful elections which were deemed reasonably free and fair in the eyes of international observers, and the NEC proved its credibility in overseeing the democratic process.18 The fact that the elections were largely funded by the Somaliland government, with additional support from foreign donors,19 demonstrated a commitment by the political elite to a form of constitutionally-based democracy. Furthermore, the peaceful resolution of the tightly fought presidential elections showed that there was a determination in Somaliland to settle political differences through constitutional means rather than resorting to violence.

The elections established political parties as an alternative form of political association to the clan, although voting patterns highlighted the continuing influence of clan politics, with marked regional variations in support for the political parties. The election of district councils established, for the first time in Somalia, the structure for a decentralised form of government that is accountable and responsive to the local electorate. At the same time, the majority of votes were cast in the urban areas, which reflected the rapid process of urban settlement in Somaliland.

The formation of elected district councils that recognise the authority of the government in Hargeisa, and pay taxes to it, reinforced the internal legitimacy of Somaliland's political institutions and leaders and helped to consolidate the boundaries of Somaliland. The election of a non-Isaaq president gave credence to the claim that Somaliland was a multi-clan polity. At the same time the non-participation of most of the population of Sool and eastern Sanaag regions served to shrink the size of the polity and make Somaliland politics more inclusive.

Furthermore, the change from inclusive consensual politics to a majoritarian electoral system resulted in the exclusion of minority groups and women from elected positions. Despite the large turn-out of women voters, only two women were elected to the 379 council seats. The elections also gave the ruling party UDUB sweeping control over Somaliland's political institutions, a situation that could only be rectified through parliamentary elections.

2: Build-up to the parliamentary election

It had originally been intended to hold the presidential and parliamentary elections together in 2003. However, disagreements on the number and distribution of parliamentary seats, the delineation of constituency boundaries,20 the absence of a census and register of voters, and procrastination by the sitting parliamentarians and the cabinet, delayed the introduction of the necessary electoral law for two years.21 The elections were postponed twice more until a formula was agreed for distributing the 82 parliamentary seats to Somaliland's six regions: Woqooyi Galbeed (Hargeisa region) was given 20 seats, Awdal 13, Saaxil 10, Togdheer 15, Sanaag 12, and Sool 12. The parties presented 246 candidates for the 82 seats,22 of which nine were `reserved seats' (in the contested regions of Sool and eastern Sanaag and Buhodle district in Togdheer) to be apportioned to parties on the basis of their overall regional vote.

As with the district and presidential elections, the parliamentary elections were organised and managed by the NEC with external technical support. This was a major undertaking, with 1.3 million ballot papers to be printed; 1,500 ballot boxes (bags) to be distributed; 982 polling stations to be identified and equipped; 4,000 polling station staff, 6,000 party agents, 3,000 police and 700 domestic observers to be trained; and 76 international observers to be coordinated. The government again contributed to the cost of the elections. But having financed two elections, and due to declining revenues and increased military expenditure on the territorial dispute with Puntland, it was only able to provide 30 per cent of the projected costs; the balance was born by international donors.23

In order to ensure a smooth, orderly and transparent process, the NEC was asked by the European Union to identify an international partner through whom funding could be channeled. The Hargeisa-based Academy for Peace and Development, an affiliate of WSP International, was chosen. It assembled an international team to provide legal, technical and logistical expertise and to develop a database to manage the information generated by the process.24

As in the previous elections, the parties signed a code of conduct prepared by the NEC covering the conduct of the campaign, the use of public finance, compliance with electoral law and measures to deal with existing gaps in the law. The media similarly signed a voluntary code of conduct for their coverage of the elections. In addition, an Election Monitoring Board (EMB) was established with representatives from civil society to independently monitor the election, mediate disputes and provide a formal public assessment of the elections.

The candidates

The democratisation process in Somaliland was intended to mark a progression from clan-based politics to party politics. This has not occurred. Although clan and religious-based parties are proscribed by the constitution, the district and presidential elections both illustrated the continuing influence of lineage politics in the multiparty system. The system chosen for the parliamentary election, whereby the electorate voted for the candidate rather than the party, served to reinforce this. If the presidential election was primarily about the parties and party leaders, then the parliamentary election was more about establishing clan representation in parliament. Although the NEC scrutinised the qualifications of the candidates, according to established criteria of age, citizenship and education, it was in the interests of the parties to have candidates who could deliver the votes of their constituencies. The clan leadership therefore played a key role in selecting the candidates and in financing their campaigns.

The motivations of the clans were various. Better access to government resources, jobs, and better services were all incentives to see their candidates elected. However, as government resources are limited, with revenues of only US$20 million per year, to increase the prestige and political relevance of the clan was just as important.25

Another key factor in candidate selection was the ambition of the candidates themselves. Some invested a significant amount of their own resources in their campaign, while others relied on their clan. Some refused to stand for the party that their clan supported, which split the clan. As some joined or switched parties at the last minute, their party loyalty was uncertain. It was difficult for some party candidates to fight on the platform of a party that they had only recently joined.

There was a great deal of tactical selection of candidates and voting. In some cases the parties sought to undermine support for opposition candidates by selecting a close relative as their own candidate. Many of the ministers promoted their own candidates from within their own clans. One minister is even alleged to have supported three candidates for the three parties from within his clan.

The selection of candidates required clan leaders to assess the numbers of votes that they could muster, as it would seem to be self-defeating to have too many candidates from one sub-lineage. However, it did happen that some sub-clans had several candidates and because the vote was spread too thinly they did not gain any seats.

The candidates and their clans were therefore the driving forces in the campaigns, rather than the parties from whom they received very little financial support. It was reported that some candidates spent as much as US$200,000 on renting premises for their supporters to meet, on campaign advertising, on qat, and in some cases on buying support by paying off debts of voters.26 A considerable number of the candidates came from the diaspora, with money and support from abroad. Running such a campaign required personal wealth or resources within the clan, which discriminated against potential candidates who were less well off, those from `minority' clans, and women.

There were very few examples of candidates who stood outside the territory of their clan on a so-called `national ticket'. One example was the owner of the Ambassador Hotel, who could draw on his personal wealth to run his campaign. But most stood for election in the regions where their clans are populous and campaigned in the districts where their clans are a majority. Prior to the election the candidates and their parties moved large numbers of people to vote in the region of their clan, particularly out of Hargeisa. As clan territories extend across the border to Ethiopia, it is reported that the candidates and parties campaigned there and organised significant numbers of people living there to vote in Somaliland.

Women candidates

Inevitably women candidates were one of the casualties of this kinship politics. Women in Somaliland have slowly been making inroads into representative politics. One criticism of the beel system of government was that it discriminated against the participation and representation of women in politics; no women had held decision-making roles in the national peace conferences.

With universal suffrage enshrined in the constitution, the adoption of electoral democracy gave women in Somaliland the right to vote and hold public office. Since 2002 there have been three women members of the cabinet, but of the 379 councillors elected in the district elections only two were women. A proposal to amend the electoral law and grant women a quota in parliament failed to win parliamentary approval in October 2004, leaving women with no alternative but to stand for election. Unlike the male candidates the women were selected by the parties rather than proposed by the clan, but as women could not guarantee to win the vote of their clan there was no incentive for the parties to select female candidates. Therefore, out of the 246 candidates in the parliamentary elections, only seven were women.27

The election campaign

The election campaign started on 30 August. Campaigning was peaceful, if a little disorganised. One fatal incident was reported, on the day before the poll, involving a dispute within one party over the allocation of fuel. When campaigning became too frenetic in Hargeisa, the parties were allocated specific days for holding rallies. There were allegations of government misuse of funds in support of UDUB and of unfair use of media time on state-run television and radio (see the appendix for a summary of the conduct of the media), and criticisms of government for appointing new ministers of state to curry favour with clans.28 Ministers in particular were criticised for using government vehicles and other resources in support of their preferred candidates. The EMB was sufficiently concerned at the government's actions that they issued several press releases and the NEC upheld these complaints. As a result media coverage became more equitable, but it proved difficult to control the use of government funds.

In terms of their policies there was little substantive difference between the parties. All held similar positions on Somaliland's independence and the dispute with Puntland and all espoused a liberal economy. The thrust of the parties' campaigns, therefore, revolved around the delivery of social and economic progress. For individual candidates the key issues were around trust and leadership and their ability to deliver benefits for their clan constituencies.

The CIIR pre-election observation team concluded that, despite some obvious problems of commission and omission, the conditions were such that a reasonably free and fair election was possible.

The large number of candidates in the elections and the active participation of the clans did give a sense of broad public participation in the elections. The nature of political debate ensured that these elections linked local district and regional politics to national politics. Rather than Hargeisa dominating the national debate, local politics had an influence on national-level politics, in a way that was not dissimilar to the clan conferences in the 1990s. Conversely, the results of the elections will impact on political relations at a local level.

3: Conduct of the election

Like the previous two elections, the election day passed peacefully. People turned out in large numbers to vote from early morning. Attempts at double voting, underage voting and some transporting of people by the parties was observed in most regions, although not in sufficient numbers to affect the outcome. In parts of Hargeisa, there were some tensions reported among the large crowds queuing to vote, but on the whole the public was disciplined, the security services maintained order, and voting procedures were largely followed.

There were some incidents involving violence - one of which in Burco the day before polling involved loss of human life. This was an internal party dispute over the allocation of fuel to candidates. However other incidents appear often to have been due to inexperience. Other problems related to the complicated and cumbersome ballot paper. This was in certain instances compounded by lengthy procedures for a largely illiterate population in a situation where there was no census and therefore no registration process. Checks on those who voted were varied and in some cases persons who appeared younger than 16 were apparently able to vote.

Equally police took action to detain those who attempted to vote more than once - which seemed to suggest the system of ultra-violet light checking was mostly working. It also suggests over-enthusiasm on the part of some political party supporters or possibly those seeking to illegally increase their clan representation.

The question of appropriate enforcement for what often appeared to be a game for young women and men is one that needs to be pursued. Observers in Hargeisa were present when police arrested two young women and took them to the police station. Some of those arrested were released in mid-afternoon - whether they tried to reoffend is not known. Observers in certain places noted that such offenders had their shoes and belts removed and were made to sit outside or in a group, as a sign of community disapproval.

In hot weather tempers did get raised on occasion, although the day was largely marked by enthusiasm and tolerance. On some occasions those who had not in fact voted were deemed by the inexperienced staff to have done so - perhaps due to the ink being passed from person to person inadvertently. On occasion in Hargeisa NEC staff were called in to arbitrate.

Queues were very long, especially in the morning and latter part of the afternoon. In Hargeisa, the majority of voters appeared to be mostly women and young people, while elsewhere it appeared queues were evenly split between male and female.

In some parts of Togdheer, heavy rain impeded both the conduct of the poll and the ability of the international election observers (IEO) team to observe as many polling stations as it would have wished. There were some polling stations, especially in Hargeisa, where law and order on occasion appeared to be on the point of breaking down. The deployment of additional security staff seemed to calm the situation. In one station NEC replaced the chairman where inefficiency and various allegations appeared to be causing problems.

Overall, however, the IEO team was satisfied that, despite some difficulties with lack of secrecy, in part due to high illiteracy, the manner in which the poll was conducted was reasonably free and fair (see box on next page).

3: Conduct of the election

Positives

- Rural and many urban stations had calm and organised queues.
- Huge voter turnout, with enthusiastic participation.
- No or very little intimidation of voters.
- The counting process was transparent and meticulous in most places, with few discrepancies.
- Widespread recognition of the value of the elections and the democratisation process.
- Voters mainly understood the ballot paper and could identify the candidate of their choice.
- A substantial number of women participated as polling station staff, domestic observers and party agents.
- Many of those attempting to multiple vote were turned away.
- Widespread commitment to abide by electoral law by voters and staff.
- Party agents were good at following the process and represented all parties at each station - usually with two people per party.
- Illiterate, remote nomadic and elderly voters were prioritised and assisted to vote in a transparent and well intentioned way.
- Good distribution of ballot materials and procedures.
- In the majority of cases the equipment and materials functioned well.
- Most polling stations opened on time, or thereabouts.
- Equal access to polling stations was allowed to both men and women.
- The `no weapons' policy was adhered to.
- Good voter registration.
- Mostly secret voting (except those that needed assistance).
- Generally good organisation by local NEC offices, including supply of extra ballot papers on demand.
- Security for international election observers was well organised.
- Well trained chairmen and polling station staff in most cases.
- Awareness and support of women candidates, in principle.

Negatives

- Lack of female candidates.
- Attempts at voter fraud and multiple voting.
- Inadequate number of polling stations overall.
- Some polling stations were too small for effective voting process, were not in neutral locations and/or lacked security in rural areas.
- Lack of census/voter register had implications for voter identification.
- Secrecy of the vote was compromised in many cases due to illiteracy or lack of voter confidence, particularly of women.
- Some instances of organised transport by political parties were reported. Reports of big trucks ferrying voters to polling stations.
- Some rural polling stations lacked local observers and/or effective staff training.
- Some polling stations lacked adequate lighting.
- Red pens were not distributed to all polling stations (Sheik district).
- Reported cases of ultra-violet light not working at some polling stations.
- Poor usage of ink or wrong ink used at some polling stations because of lack of effective training on how to use the invisible ink or decant it effectively.
- Transference of ink between voters potentially prevented legitimate voting in some cases.
- Lack of clarity about opening and closing times for polling stations, particularly when people were still queueing after 6pm.
- Fatigue of all staff, in particular the polling station chairman, due to pressure of logistical issues, long working hours and large number of responsibilities.
- Reported cases of voters from across national borders.
- Reported cases of observer teams and members of the Special Protection Unit of the Somaliland Police Force causing disruption.
- Reported cases of party agents doing the duties of polling station staff.
- Campaign materials were evident near the polling station in a minority of cases.
- Lack of familiarisation of polling stations by party agents - reported by an observer and NEC headquartters.
- Communication problems between all levels of election administration on polling day and beforehand.
- In a minority of cases, queues were disorderly.
- Women often received less voter education than men, especially in rural areas.

Conduct of the election

4: The role of the international election observers

The IEO mission was truly international in its composition, thus reinforcing its overall credibility as a voice of the international community. Particular positives were the inclusion of many women, several members of the Somaliland diaspora, and observers from Africa itself - namely Ethiopia, Kenya, Mozambique, South Africa and Zimbabwe.

The logistical coordination of the 76-strong mission was also effective, particularly as it was undertaken under significant resource and time constraints. All IEOs were given appropriate training and briefing about their roles and responsibilities and the international standards by which they were to conduct their observation. This was vital to ensure that observers adhered to the principles of election observation and did not inappropriately engage in the election process. Equally they were briefed or came with prior knowledge of the contextual situation. The regional focus of the organisation ensured broad observation across most of Somaliland, reinforcing accuracy of assessments.

The mission covered the whole election process including the pre- and post-election period in-country. This included an assessment of the campaigning processes, adherence to the party and media codes of conduct, and mechanisms for complaints and upholding of the electoral law.

On the ground, there were some issues concerning the interaction between local people and the IEOs. Voters and polling station officials were often aware of the mission's overall role, but on occasion seemed unclear as to where the boundary between observer and participant lay. Guards from the Special Protection Unit (SPU) of the Somaliland Police Force, who provided security for the IEOs, would have benefited from more training about their role during the observation period. While they did not directly interfere with the process, on occasion they failed to understand that their presence within the polling booths or their actions did impact on the process. Some problems arose because of the NEC's failure to clarify effectively and in a timely manner their expectations of the roles and responsibilities of their own body and that of the CIIR mission. As a result, CIIR was often left to take responsibility, financially and practically, for elements of the processes which the NEC should have been responsible for.

Poor communication and information-sharing by the NEC made the logistical organisation of the IEO teams more difficult and complex than it should have been. It put a great strain on CIIR's resources and therefore blurred the lines of responsibility.

Without up-to-date and accurate maps, particularly of the rural regions, IEO teams were heavily dependent on local knowledge when planning their polling station routes. More accurate information (from the NEC) about polling station locations and the distance between them, even without maps, might have facilitated this preparation.

Recommendations for future election observation missions

CIIR's role balanced the need for effective coordination of IEO teams with the institutional independence required by organisations participating in the IEO teams. If future election observation missions are conducted in a similar manner it is highly recommended to continue to have a single coordinating body.

A greater number of observers would increase coverage of the country and reduce the number of polling stations that each team visits. This would enable IEOs to spend more time at each station and enhance coverage and analysis. Equally, as the mission focused on the period leading up to polling day, polling day itself and the immediate aftermath, this left little scope for a continuous, directly observed assessment of the situation. Both points require greater financial commitment from donors and greater logistical coordination and resources.

4: The role of the international election observers

5: Recommendations of the international election observers

- While overall the elections were conducteed in a reasonably free and fair manner, there is scope for improvement, and a number of recommendations have been made. These include:

- greater civic education and training for candidates, election staff and voters
- stronger measures to encourage female candidates
- a higher degree of preparation and establishment of a permanent electoral commission
- more resource support from the international community.

It is also recommended that a national census be held as soon as possible. This would enable a voter registration system to be established (and would in turn enable accurate estimations of turn-out). It could also help eliminate the biggest problem: attempts at multiple voting. These were observed on frequent occasions by the IEO teams. However, it is pleasing to note they were often detected, and were by most accounts far less numerous than in 2002's local elections and the presidential election in 2003.

Recommendations to the NEC

Where Somaliland's own electoral organisation is concerned, more polling stations will be needed in future, particularly in rural areas where people are expected to travel great distances. Improved equipment (eg lighting) needs to be provided to the polling stations. Communication links between NEC headquarters, local and district level NEC offices and with IEO teams need to be improved to ensure clarity on roles and responsibilities and to facilitate logistical arrangements. The NEC needs to consider additional ways of assisting illiterate voters during polling day so that secrecy of voting is not compromised.

The internal capacity of the NEC itself needs to be increased, with the support of external donors and practitioners where appropriate, to ensure greater organisation, effectiveness and transparency of their work. This would also reduce their dependence on external organisations to undertake tasks. Such capacity building should include staff training, timetabling, voter education (particularly of female voters) and preparation of ballot papers and candidate lists.

The NEC should maintain local/regional NEC offices beyond election periods. At present they are scaled back and closed down until the next election/referendum. However, to maintain genuine capacity development, local training and political engagement, these offices needed to be fully functioning. This, again, relies heavily on the sourcing of funds.

The NEC needs `more teeth'. It needs to be able to hold people to account for their actions when they violate the electoral law or codes of conduct. Perhaps most importantly, the NEC needs to develop a system of voter registration and/or to campaign for a full country census of Somaliland.

Main points

- Increase number of polling stations and improve accessibility in rural areas.
- Need for greater civic education and polling staff/candidate training on all aspects of voting process.
- Need to begin preparations for future elections earlier including funding.
- Need to take message back to donor/international community that without their support and commitment Phase Two of the European Commission's Somali democratisation programme (covering all Somalia but beginning in Somaliland) is unlikely to happen.
- Need for a national census and a voter registration system which will help solve many difficulties.
- Need for a permanent electoral commission with a dedicated section dealing with domestic and international observers.
- Need to ensure that more women are involved in the election as a whole.
- Improve campaign chances of women by affirmative action, including concrete examples, such as provision of vehicles for women candidates.
- Investigate how quota systems for women (and minorities) have worked elsewhere (eg Uganda).
- Strict adherence to electoral law and code of conduct would expedite the elections. Identification and multiple voting
- Ideally, full voter registration should take place, and voters be issued with ID cards.
- Review options for marking fingers and ballot paper. The use of invisible ink appeared to work technically, but it was not obvious to voters, especially young adventurous ones, that there was any marking - hence attempts at voting again were perhaps motivated by curiosity about the system of marking.
- Show voters the ink under the light on arrival, to discourage those who are attempting to double vote because of the novelty of invisible ink.

5: Recommendations of the international election observers

- Standardise the inking procedure, with specific instructions on where to apply ink, and on applying ink on the way in to the polling station to give time for the ink to dry.
- Consistent and stricter sanctions on those attempting to vote more than once, such as detention, and perhaps sanctioning parties as well as individuals.
- Notices to be posted about these penalties in polling stations to deter parties/would-be multiple voters.
- NEC to transmit messages about the consequences of multiple voting throughout its campaign. Staffing
- More gender balance of polling station chairpersons.
- Chair over-burdened; she/he should take a coordinating role only.
- Second person other than chair to assist illiterate people in marking ballots.
- Consider reducing party agents to one per party at any one time - some polling stations were very small and overcrowded.
- Review criteria for selecting polling stations.
- Consider having a day and a night shift for all staff, or other measures to reduce fatigue.
- Ensure food and drinks available for polling station staff.
- Examine options in relation to qat chewing in polling stations which can be intimidating and not helpful for performance of duties.

Polling station logistics

- Need for more appropriate buildings for polling stations, especially in terms of size and lighting (voting could take place outside in rural areas - although this is not feasible during the rainy season).
- Separate entrance and exit doors on polling stations, where feasible.
- Need for increased supplies of pens with larger tips and lighting equipment; also a magnifying glass for each polling station.
- Pens should be fastened to the table in the voting booth.
- Black felt tip pens more useful than red especially when some regional ballot papers were pink.
- Two voting booths to be in place, as per electoral law.
- Ballot papers should be reduced in size.
- Replace serial numbers on ballot papers with random numbers or bar codes, to avoid possibility of tracing voters through register (funding would be required).
- Ideally, ballot bags should not be used again. (The ballot bags used were chosen because they could be readily imported from Ethiopia, but they proved impractical.) If they are, ballot bags should be bigger; frames should be provided; they should be centrally located in the line of sight of observers; and seals for the ballot bags should be distinguished by different colours.
- Voter turn-out in this election should be the basis for distribution of ballot papers in the next election if there has not been a census and registration process.
- Increase efforts to have more ballot papers on hand and readily available when requested.
- Diagrammatic instructions posted inside the voting booths would help voters and ensure greater secrecy.
- Soldiers and the SPU need greater guidance on their role - in particular, about not engaging in the voting process in any way and staying outside the polling stations.
- Training is needed for polling staff on transport procedures for the ballot boxes to go to district/regional level once voting has been completed and votes counted.

Education, training and information

- More civic education (of course this has funding implications) - which until now has mostly been done by civil society when funds were available.
- Civic education and awareness campaign for electorate on individual choice with greater emphasis on multiple voting being both wrong and illegal.
- Provision of sample ballots and voter education material outside polling stations to assist voters.
- Clearer training for polling station staff on the following points: setting up of polling stations; inking procedure; how to seal the ballot bag; the counting process, eg use of record book; return of ballots to the district office; and culture of respect towards voters.
- Male candidates using women's faces as symbols should be discouraged as this leads to possible confusion as to the gender of the candidate. Closing and ballot counting procedures
- More clarity about closing procedures, with the NEC taking overall responsibility at national level, and the chair announcing imminent closing at polling station level. This would comply with electoral provisions that those in the queue at 6pm should be allowed to vote, but not those joining the queue after that. Transport
- More transport to be made available to the NEC for supplying material, replenishing ballot papers and transporting ballot boxes.
- Enforcement of prohibition of vehicle movement; fining parties for transporting voters on election day.
- However, consideration should be given to the possible need to provide transport for disabled and remote nomadic voters.

Recommendations to the international community

- Donors, including the new United Nations Democracy Fund, should recognise the requirement for sustained funding for:
- A permanent NEC with full time staff and sustained training programmes.
- A nationwide census to aid voter registration, ballot paper distribution and the reduction of multiple voting in time for the planned 2007 local government elections.

5: Recommendations of the international election observers

- Logistical support including transport, short term polling stations building purchase from the private sector, additional staff and polling station materials (eg ballot boxes).
- Support for women's participation in the civic and electoral process.
- Support for parties' reflections on lessons learned and options for changes and improvements - including the possibility of exposure and exchange visits.
- Investigating how best to respond to ideas emanating from civil society (see section below). Donors might also investigate possibilities of providing information to parliamentarians and parties on how to develop coherent policy initiatives and form sustainable policy-oriented parties (including clear processes for promoting internal party discipline). Recommendations to political parties
- Engage with party supporters, local civil society and international donors on how to develop coherent policy initiatives and form sustainable policy-oriented parties.
- Examine ways of increasing women's participation in the civic, electoral and parliamentary processes.
- Undertake with others reflections on lessons learned and options for changes and improvements - including the possibility of exposure and exchange visits. Recommendations arising from dialogue with civil society
- Civil society groups need practical training materials, manuals and posters to further voter education.
- Ability to learn from other countries such as Uganda on electoral practice.
- Training for civil society in lobbying and advocacy techniques.
- Coordination of election training activities between civil society, political parties, NEC, government and parliament. (This could lead to a joint plan of action.)
- Greater involvement of civil society in post-election activities.
- Funding to civil society well before the elections.

Reflections on the recommendations

Women's participation

Despite the relative progress in women's participation in Somaliland, women still remain largely excluded as representatives in parliament. Voters in the parliamentary elections were encouraged, for example, to support male candidates, particularly if in the same clan.

Although many of the party agents and local observers were women - as were the majority of voters - the political parties did not do enough to promote women candidates. The strong influence of clan support for particular candidates, combined with the fact that clans are male led and male dominated, means men are put forward as candidates. In addition, with a female candidate, the issue of clan loyalty is complicated in terms of whether she is representing her family clan or the clan of her husband.

The draft electoral law submitted by consultants in early 2005 suggested that a quota or other form of positive discrimination would be needed to address the issue of women's participation and representation in the electoral process. However, the parliamentary committee which oversaw planning for the election refused to even discuss the idea.

Women were not engaged in the drafting process of the new multiparty system. Efforts should now be made to review the electoral laws and structures with the promotion of women's participation in mind. This may, for example, lead to quotas for women candidates or a framework to increase their participation. But providing the means for women to participate is not on its own enough. Women must want to be candidates, and must feel that they can compete on equal terms - and must genuinely be able to do so. This will require awareness raising, advocacy work and training of women so that they are effective as candidates and potential MPs.

Women voters need to be given the freedom to vote as they wish and not be forced into clan/family lines. This issue is made even more difficult to address because of the illiteracy levels among women. As a result, in many cases, when they voted their votes were read out and shown to people. This may have ensured transparency but it failed to give women the voting secrecy which may have enabled them to vote more freely.

Training should be given to political parties and MPs to make them aware of the need to secure women's votes and also integrate women into their own parties and management and policy structures. If they could see the benefits of having women candidates it might secure greater support for women candidates.

Greater efforts need to be made to educate women and girls in rural areas where they have less access to the media and resources available in Hargeisa and other urban centres. Radio could play a key role in this process and the capacity here needs to be increased.

Transport

Travelling is banned on polling days for security reasons and in order to minimise attempted multiple voting. There is, however, a clear need for better transport for voters, particularly those who are elderly, disabled or live in rural areas.

Transport by political parties is not uncommon in other countries and does not necessarily determine which way people vote in a secret ballot. However, if political parties or candidates are allowed to provide transport, there is the danger that those with more money and access to greater resources, especially those from the governing party at the time of the election, might be given an unfair advantage. Here both electoral law and its implementation are crucial. An alternative possibility is to increase the number of polling stations to reduce the need for transport on polling day.

Ballot papers

Distribution of ballot papers was based on voting patterns in the last election but was still difficult to gauge as people moved around the country before polling day in order to vote for their favoured candidate (and so gain clan advantage in particular areas). The issue of the breakdown of communication between polling stations and the district levels where additional ballot papers were needed needs to be addressed.

Parties also expressed concerns that taking large numbers of ballots to the border polling stations increased the risk of fraud.

Originally there were to be pictures of candidates to assist illiterate voters but they were not clear so symbols were introduced (not without controversy). Some male candidates used women's faces as symbols. This should be discouraged as it could lead to confusion as to the gender of the candidate (although there is no evidence that this was the intended outcome).

Training

The NEC had only from 2 July (when it received money to carry out training) to train people in their different roles. This was a serious constraint and needs to be rectified in future. In particular, there were problems when the number of polling stations was suddenly increased after staff training had been completed, making it very difficult to arrange everything in these additional places on time.

25,000 sample papers were issued for training purposes, but seem to have disappeared which gave some cause for concern, especially when fake ballot papers were intercepted at Hargeisa airport.

There is a need for increased party agent training and also training of others on the role of party agents so that they are not asked to facilitate the voting process other than within their defined role. Similarly, soldiers and the SPU need to be properly informed about their role in the process. They often became drawn into the voting process in some way and did not adhere to the electoral law by staying outside the polling stations (although it is not obvious that any intimidation occurred).

Invisible ink was chosen for voter identification as a result of the widespread use by Somaliland women of henna (which is difficult to remove). The decision was taken only the month before the election and training was necessarily limited on use of the ink, when to apply it, what to look out for in terms of ink being transferred and indeed the novelty factor. As a result, for example, although polling station staff were told to apply the ink upon entry to the stations, some applied it upon exit. In addition to staff training, voter education will be necessary to help resolve these and other problems.

Chairs of the polling stations were overworked, with responsibilities including assisting the illiterate and disabled voters. Diagrammatic instructions, like those printed in the voter training manuals, would - if clearly posted inside the voting booths - help voters in making more secret and independent voting decisions and reduce the chair's workload. It would also help if in future the chair focuses more on coordination and supervision, leaving the more administrative tasks to a deputy - although this obviously would require more resources, training and staff.

Ongoing outside support

A national census and a voter registration process, bearing in mind the impending local elections scheduled for 2007, would both need considerable external funding. Outside assistance in terms of long term work - rather than the usual short term consultancies - is also needed to help build and maintain structures, systems and skills in good governance after the elections.

6: Analysis of the election results

Counting the votes

As happened in the previous presidential election, disputes between the parties emerged once the ballot counting started, with calls for recounts and the disqualification of ballots in some areas. One reason for this is that the process for counting and reconciling figures was not followed precisely according to the election law. The law required the ballots to be counted in the polling stations overnight, and the results to be immediately sent to the regional office of the NEC and then transmitted to the national NEC office in Hargeisa. Given that most polling stations finished counting in one night, it should have been feasible to tabulate all the results in Hargeisa within four days. Instead it was not until 14 October that the final results were announced, two full weeks after polling day.

Turn-out

The results of the 2005 parliamentary election were announced by the NEC on 14 October, and were confirmed by the Supreme Court on 1 November.29 A total of 670,320 valid votes were counted in the parliamentary election. This was lower than some had speculated and less than half the 1.3 million ballot papers that were printed for the election. Nevertheless, the turn-out was considerably higher than in the previous two elections;30 over 182,000 more votes were counted in the parliamentary election compared to the presidential election. The higher turn-out for the parliamentary election can partially be accounted for by the increase in polling stations from 900 in the presidential election to 982 for the parliamentary election. Seasonal factors also meant that pastoralists were grazing in Somaliland during the parliamentary election. A substantial number of people may have crossed from Ethiopia to Somaliland for the election, which may also account for regional variations in voting, although hard evidence for this does not exist. Voting patterns

After three elections it is possible to identify some regional patterns of voting (see table 1 and chart 1). In each election the western regions of Woqooyi Galbeed (Hargeisa region), Awdal and Saaxil have accounted for over 60 per cent of the votes cast, reflecting the concentration of Somaliland's population in the west. The fact that the largest number of votes were cast in Woqooyi Galbeed further attests to the concentration of the population in this region and the growth of the capital Hargeisa since the war. These regional patterns of voting also reflect historic socio-economic differences between Somaliland's regions. Better transport and infrastructure, shorter distances for voters to travel, better media coverage, more intense campaigning and voter education and a more sedentary population also facilitated a higher turn-out in the west. The smaller number of votes in Somaliland's three eastern regions of Togdheer, Sanaag and Sool can, in part, be accounted for by the smaller population and the fact that the population is more nomadic than in the west of the country.

The lower poll in the east is also a consequence of the non-participation of parts of eastern Sanaag and Sool regions in the elections. The very low number of votes counted in Sool region, ranging from one to three per cent of the total vote, illustrates this. Nevertheless, in line with the rest of the country, there has been an increase in the numbers of votes counted in Sool in the course of three elections. This may represent an incremental increase in participation by people in Sool, albeit at a time when representation in parliament is declining.

Other patterns worth noting are differences in the regional votes between the three elections. First, there has been a decline in the total vote in Woqooyi Galbeed compared to other regions, from 42 per cent of the total vote in the district elections to 35 per cent in the recent parliamentary election. The decline probably reflects the large movement of people from Hargeisa to vote in other regions prior to polling day. Based on figures from previous elections, as many as 35,000 people may have left Hargeisa to vote in other regions. This movement happened peacefully and required considerable organisation by the candidates, their clans and the parties.

6: Analysis of the election results

Table 1: Regional votes in Somaliland's three elections

Region      District 2002 % Presidential 2003 % Parliamentary 2005 %
W/Galbeed   186,383   42    208,864        43   253,229          35
Awdal       100,495   23    65,934         14   133,026          24
Saaxil      27,234    6     30,537         6    52,479           7
Togdheer    66,598    15    115,064        24   121,751          17
Sanaag      53,096    12    57,938         12   89,286           15
Sool        6,261     1     9,702           2   20,557           3
Total       440,067   100   488,039        100  670,328          100

Sources: Bradbury et al, 2003; NEC, 2005.

Second, there is a significant change in the scale of voting in Awdal and Togdheer regions between the three elections. In Awdal region the vote has swung from 23 per cent of the total vote to 14 per cent and back up to 24 per cent in successive elections. While the overall vote in the parliamentary election has increased compared to the presidential election, the increase has been by far the largest in Awdal region, which records 67,092 more votes than in the presidential election; this compares to an increase of 44,365 in Woqooyi Galbeed. As reports on the previous elections noted,31 Togdheer is almost the mirror opposite of Awdal. There the vote has swung from 15 per cent of the total vote in the district elections to 24 per cent in the presidential election and back to 17 per cent in the parliamentary election. These two regions are politically significant, given the importance of the vote in Awdal for UDUB and in Togdheer for Kulmiye. The large swings in the scale of the vote cannot be accounted for by seasonal variations in the timing of the elections, or the movement of people from Hargeisa or from Ethiopia which should have affected Awdal and Togdheer equally.

At the very least, the large variation in the scale of the voting in these two regions reinforces the need for voter registration.

The political parties

The ruling party UDUB repeated its successes in the district and parliamentary elections by gaining the largest number of votes, returning 33 MPs, with Kulmiye second with 28 and UCID third with 21 (see table 2). UDUB is therefore the largest party in parliament, but with the opposition having 49 seats it does not command an overall majority. This makes Somaliland the only place in Africa where parliament is not controlled by the government.

All three parties saw their vote increase (see table 3), but the proportion of the popular vote for each party changed (see chart 2). In terms of the popular vote, UCID has been the main gainer in the parliamentary elections, increasing its overall vote by 103,118, compared to an increase of 55,854 for UDUB and 22,813 for Kulmiye. This is a significant development for UCID which came a poor third in the district elections.32 Its growing importance is apparent from the fact that it won parliamentary seats in all regions and beat Kulmiye into third place in Saaxil. One of the reasons for the increase in UCID's popularity has been a shift in support to them from UDUB by the Habr Yunis clan. Another is that candidates who were unable to get onto UDUB's list chose to join UCID instead.

Chart 1: Regional distribution of votes in Somaliland's three elections

Table 2: Parliamentary election results

Party Total votes % of vote Parliamentary seats
UDUB      261,449   39          33
Kulmiye   228,328   34          28
UCID      180,551   27          21
Total     670,328   100         82

Source: NEC, 2005.

In terms of the popular vote, Kulmiye appears to have lost ground in the parliamentary elections, with UCID making marginally more gains from Kulmiye than UDUB. While Kulmiye gained support in Woqooyi Galbeed, it appears to have lost support to the other parties in Saaxil, Togdheer, Sanaag and Sool regions. In Togdheer its support declined by over 15,000 votes, while UCID's increased by over 21,000 votes. The comparative decline in the popular vote for Kulmiye probably reflects the strength of the Habr Yunis clan which increased its number of parliamentary seats (see below). However, Kulmiye holds a majority of parliamentary seats from Togdheer, and perhaps more importantly has the majority from Woqooyi Galbeed, including the capital.

Compared to the presidential elections UDUB lost some support in the parliamentary elections in Woqooyi Galbeed and Togdheer, but gained in Awdal. The largest single increase in the regional vote for a party was for UDUB in Awdal, where it received 31,344 more votes than in the presidential election. The importance of this high vote in Awdal vote is clear, given that UDUB won a majority of seats in only Awdal and Sool (see table 4).

Party representation in parliament

UDUB holds the largest number of seats in parliament, but by a small margin of only five seats over Kulmiye. In contrast a combined opposition can command a sizeable majority of 16 seats over UDUB (see table 4). This has removed the monopoly that UDUB has held over Somaliland's political institutions since the district elections and means that parliament can, if it chooses, hold the executive to account.

As political parties were not represented in parliament prior to these elections, it is difficult to link changes in clan representation in the parties and in parliament. The results, however, do point to a bedrock of support for UDUB from the Habr Yunis and Gadabursi, for Kulmiye from the Habr Jeclo' and Habr Awal (mostly Sacad Muse) and for UCID from the Habr Yunis (see table 5).

Winners and losers

As noted, although multi-party democracy in Somaliland was intended to move Somaliland away from clan-based politics, the influence of clan politics on the elections remains apparent. The parliamentary elections did not produce any dramatic changes in popular support for the parties, but the clan composition of the parliament has changed (see table 6). The adoption of a majoritarian voting system has produced winners and losers.

The most obvious change is that Isaaq representation in parliament has increased by nine seats to 57, while the representation of the Harti and `minorities' has fallen. The increase in Samaroon (Gadabursi) representation from 11 to 13 seats was predictable given the regional distribution of seats agreed by parliament for this election.

The biggest losers have been the Harti (Dhulbahunte and Warsengeli) people of Sool and eastern Sanaag regions whose representation in parliament has fallen from 14 to 10 members. The attitudes of the Harti towards the election have been mixed. Some participated in the election, while others rejected it. The results which leave parliament dominated by the Dir (Isaaq and Samaroon) will only increase their sense of marginalisation within Somaliland.

Table 3: Votes for parties by region in the presidential and parliamentary elections

UDUB 2003 UDUB 2005 Kulmiye 2003 Kulmiye 2005 UCID 2003 UCID 2005
W/Galbeed  79,328   81,552   81,585    95,881    47,951    75,796
Awdal      43,347   74,691   16,607    26,837    5,976     31,498
Saaxil     17,554   21,793   10,271    12,355    2,712     18,331
Togdheer   38,105   39,529   63,506    47,639    13,453    34,583
Sanaag     23,359   34,727   27,830    36,652    6,749     17,907
Sool       3,715    9,157    5,524     8,964     463       2,436
National Level 187      -    192        -        129        - 
Total     205,595  261,449 205,515    228,328    77,433   180,551

The Ciise have also seen their seats reduced from five to one, but largely as a result of five Ciise candidates withdrawing from the election a week before polling day. The `minorities', who gained representation in Somaliland's second parliament, have lost all bar one of their seats in the legislature.

Within the majority Isaaq clan family, the three largest clans have all seen an increase in their number of seats. Together, the Habr Awal (Sacad Muse and Ciise Muse), Garhajis (Habr Yunis and Ciidagale) and Habr Jeclo' hold 51 seats. On the other hand, the smaller Isaaq clans have both seen a reduction in their number of seats: Arab (five), Ayub (one). Within the Isaaq the most significant change has been an increase in the representation for the Habr Yunis in parliament (up from seven to 17 seats); together with the two Ciidagale seats the Garhajis have over 20 per cent of the seats in parliament. They have representation from across Somaliland, with the exception of Awdal region, and are potentially the single largest `clan block' of votes in parliament. Even in Hargeisa three Habr Yunis candidates were elected, which is surprising given that they are more populous in Togdheer. The shift in Habr Yunis allegiance to UCID has made it into a serious third party. In 1994 the perception by the Habr Yunis that they were under-represented in parliament was one of the grievances behind the civil war in Somaliland. The equitable representation between the three major Isaaq clans should mitigate clan-based ructions in Somaliland.

Other changes are apparent at the level of sub-clans, with some of the larger lineages and politically stronger lineages losing ground to smaller lineages. This is partly due to the larger lineage fielding too many candidates.33 The impact of these changes, if any, will only become apparent over time.

Table 4: Seats won by the political parties by region

UDUB         Kulmiye UCID Total All Parties
Hargeisa        6       8     6      20
Awdal           7       3     3      13
Saaxil          4       2     4      10
Togdheer        5       6     4      15
Sanaag          5       5     2      12
Sool            6       4     2      12
Total Seats     33      28    21     82

Table 5: Distribution of seats by clan and party

Clan           UDUB Kulmiye UCID Total Seats
Sacad Muse       2   5      2    9
Ciise Muse       3   1      3    7
Habr Yunis       8   1      8    17
Ciidagale        2               2
Habr Jelco'      5   9      2    16
Arab             2   3           5
Ayb              1               1
Gadabursi        7   3      3    13
Ciise            1               1
Warsengeli       2   2           4
Dhulbahunte      2   3      1    6
Hawiye-Fiqishini 1               1
Total Seats      33  28     21   82

Source: interviews, Hargeisa 2005.

Table 6: Change in clan family composition in parliament

Clan families Old parliament New parliament
Isaaq             48          57
Samaroon          10          13
Ciise             5           1
Dhulbahunte       9           6
Warsengeli        5           4
Hawiye/Fiqishinin 1           1
Minorities        4           0

Source: interviews, Hargeisa 2005.

The new parliamentarians

Only 18 parliamentarians stood for re-election, of which 14 were re-elected. The elections have therefore brought a wholesale change in the members of parliament, as well as party and clan representation.

In general the new MPs are younger than in the old parliament, with 56 under the age of 50 and 12 under the age of 40, and only one over the age of 60. This compares to only five under the age of 40, and 16 over the age of 60 in the old parliament. But, given that many of the old parliamentarians became MPs in 1993, the average age of the current intake is probably similar to that of 12 years previously.

In terms of professional background there are slightly fewer MPs in the new parliament who were former members of the security services, or former politicians, and slightly more professionals and civil society actors. There are also more new members from the diaspora.

Women had everything to gain from the parliamentary elections: with no elected women in any previous Somali parliament, the presence of even one would be progress. In the end two of the seven women that stood in the elections gained seats. One was elected as a Kulmiye candidate in Awdal, and the other, from eastern Sanaag, got in on a reserved seat. The lack of wider success has been a difficult lesson for those advocating for women's participation in politics. It has convinced them that while the clan system remains so influential women will not get elected to parliament.

They have concluded that it will be better to advocate for a quota system, which has given women in Puntland and Somalia a larger representation in their parliaments. In their assessment of the elections civil society groups stated: [In] a clan-based society struggling to transform itself into a multi-party democracy a quota for women is the only solution to advance the participation of women who essentially contributed to peace building and economic reconstruction and rehabilitation of this country.34

The road to a popularly and directly elected government in Somaliland has been long and difficult. Since unilaterally declaring independence from Somalia in 1991, the people have been through civil war, they have lost the main market for the key livestock trade, they have had to cope with the reintegration of huge numbers of refugees and displaced people, and they have had to rebuild their tattered infrastructure. While the lack of international recognition and limited support for reconstruction has been detrimental in many ways, it has given people a certain freedom to craft a model of government suitable to their needs. This has combined indigenous forms of social and political organisation - famously described as a pastoral democracy35 - within a liberal democratic framework. Furthermore, given the minimalist role that government plays in people's affairs, the Somaliland state is not something that has been imposed from above, but is rooted in popular consciousness. While Somaliland's independence is contested externally and internally, for the majority of the people the desire to be treated separately from Somalia is very real. The staging of district, presidential and parliamentary elections was an expression of the people's desire and ability to manage their own affairs. The parliamentary election was a critical event in the transition to democratic government in Somaliland. But it is not the end of the process, and Somaliland faces many challenges in putting democracy into practice.

The house of elders

First, there is the matter of the upper house of elders (the Guurti). This remains an unelected legislative institution. Its mandate is due to expire in 2006 and a decision will need to be made on whether it should become an elected chamber or remain an appointed one. As people in Somaliland experiment with democratic multi-party politics, they will be challenged to maintain and incorporate the positive attributes of a pastoral democracy - consensus building, mediation, arbitration - within the system of government, while minimising its negative influences. The Guurti provides a mechanism for this. Although the chamber has been criticised for becoming too institutionalised and reliant on government to be able to act as the objective mediator in Somaliland's national politics, it remains a unique institution that has been at the heart of clan-based power sharing and consensual politics in Somaliland, linking modern political institutions to the traditional political organisation and, by extension, inter-communal politics to national politics. The public recognises the role that it played in mobilising the population in the war against Siyad Barre and in shepherding Somaliland through the minefields of post-war politics and state-building. There are many sources of legitimacy and authority in societies and although the authority of the Guurti is not based on a popular vote it is no less legitimate for that.

There is a general consensus among the parties, the NEC and the public that the Guurti should remain a nominated body. And the country cannot afford another election. The Guurti can be a place where imbalances in political representation as a result of elections can be restored, with seats being made available for minority clans and possibly even women. Meanwhile, following the 2005 elections, the Guurti will certainly have an important role to play in assisting the new parliament to settle in.

A functional parliament

A measure of Somaliland's democracy will be the way in which Somaliland's elected institutions function and are utilised by the parties and the public. The first few months of the new parliament are likely to be a turbulent time, as its members and the parties negotiate their relationship with each other and with the executive. Somaliland will be the only country in Africa where parliament is not controlled by the government. The majority of the MPs will be new, and while many of them will be better educated than their predecessors, they will lack experience and be unfamiliar with the functions of parliament. Achieving political consensus has been the cornerstone of stability in Somaliland, to the extent that uncomfortable compromises have been made at times. In the new parliament, the opposition are likely to form an alliance and challenge the government on a number of fronts.

The opposition agenda for parliament includes:36

- Revoking the emergency laws.
- A review of the constitution, to curb the powers of the executive, to review the size of parliament, and to review the restrictions on political parties. This might include a proposal to create a post of prime minister.
- Impeach the president. Although openly discussed in the elections, opposition party leaders have cooled this fervour. There may be a push for early presidential elections, although this is likely to be rejected on grounds of cost.
- Measures to reduce the size of the cabinet from 50 to between 12 or 18 ministers. It is likely that the parliament will exercise its powers to reject cabinet appointments and may push for the removal of certain unpopular ministers.
- Measures to increase fiscal accountability and transparency in the executive, through greater control and oversight over the national budget, a review of foreign investment contracts, a review of fishing concessions, a review of the management of Berbera port, and the establishment of a commission to tackle corruption.

7: Beyond the elections

- Measures to open up and stimulate the economy, including plans for the leasing of Berbera port.
- A review of the media law.
- An open debate on Somaliland's relationship with Somalia, and the status of the contested eastern regions.
- A review of the security sector budget, with the aim of cutting it and putting more resources into social services.
- Voter registration and census, which is seen as essential not only for holding elections but as part of the state-building process (by defining and counting one's citizens).
- A review of electoral law, to consolidate and iron out contradictions in the existing legislation.
- Strengthen local government laws for the decentralisation of government.
- A review of the role of the Guurti, and its appointment.
- Renewal of the mandate of the NEC, which has another year on its mandate.

The opening of the new parliament on 28 November did not set a good precedent. The task of electing a new speaker and deputy speakers for the new parliament resulted in scuffles within parliament, and a postponement of the next session for several days. The crisis was resolved by the Guurti who mediated between the parties; the opposition's candidates were duly elected as the speaker and his two deputies.

The practice of government

Having committed itself to an elected government, the government needs to demonstrate its respect for civil liberties, human rights standards and the rule of law that are expected in a democratic society. Since 2002, Somaliland's reputation for this has been called into question by some high profile legal cases, creeping corruption, and an increasing investment in internal security.37 Immediately after the elections, the government demonstrated again its intolerance of anyone that it does not agree with, by declaring the European Union delegate to Somaliland persona non grata. Such arbitrary action will not win Somaliland foreign friends.

The political parties

The new parliament will be an important test for the system of multi-party politics. The parties are weak institutions that, to date, have shown little life outside the election campaigns. They appear as vehicles for their leaders, their internal constitutions are weak, and there is little democracy practised within them. During the campaign the parties did not bring their candidates together to formulate policy and they missed the opportunity to get their message across to the public. This will make it much harder for the parties to enforce discipline.

The parties' links with and control over members elected in their name to the district councils has been tenuous. The regional structures that existed disappeared once the parliamentary election campaign started. In the presidential election the parties operated as united political forces, with funding and campaigning centralised. A major constraint facing the parties has been resources. In the parliamentary election they received only a limited amount of public funding from the government for core salary costs, and were reliant on the patronage of clan and businesses. The leaders of the parties did not themselves stand in the election, and they are likely to find it difficult to enforce adherence among their MPs to the party or its policies. For the candidates the parties were a legal mechanism for entering parliament. The sitting MPs may seek to reform the parties, by insisting that they become members of the parties' executive committees. Some may decide to join other parties. One of the few forms of leverage that the parties have over their MPs is for the party to refuse to renominate a candidate who does not follow the party line.

The parties will also be under pressure from women and civil society organisations to review their policy on women candidates, and ensure that changes are made to their structures, policies and political agendas that will affirm the participation of women and minorities in politics.38

A debate continues in Somaliland on the validity of a three party system in a plural democracy. Some argue that the restricted system is a sensible solution to clanbased politics, to prevent social cleavages or parties becoming a reflection of the clan. Others argue that the restriction on the number of parties is a direct contradiction of the right to free association. The constitution also effectively gives the parties eternal life. While the parties are unlikely to push for a change, there is likely to be popular pressure to review the restriction and allow for an increased number of parties or a mechanism that would enable the de-selection of a party that fails to deliver.

Civil society

Another challenge will be how the relationship between civil society groups, the political parties and government develops. Civil groups in Somaliland are actively organising on issues of gender representation, human rights and social service provision. Although parliament includes several MPs who were formerly active in civil society, it remains to be seen whether it proves to be more amenable to external advocacy or will actively seek opinion from civil society groups.

Democratisation and state-building In 1999, President Egal argued that democratisation would facilitate international recognition of Somaliland. However, while the parties campaigned on a pro-independence platform and many voters themselves saw voting as a patriotic act, these elections were not put on for external purposes. They were about changing the internal balance of power. Nevertheless, the parliamentary election was also another step in building an independent state and it has implications for the status of Somaliland and its relationship with Somalia.

The parliamentary election serves to reinforce the contrast between Somaliland's elected administration and the Transitional Federal Government in Somalia, which is struggling to sustain a semblance of parliamentary consensus. Reactions by people in Somalia to the parliamentary election in Somaliland were mixed. Somaliland received messages of support from southern civil society groups after the election and in November 2005 the prime minister of the TFG, Mohamed Geedi, made the extraordinary statement, in an interview with the BBC, that `if the international community recognises Somaliland then we will have nothing against that'. However, the statement caused a furore, with the prime minister accused of treason,39 and was rescinded.

Having staged three elections, the commitment of the Somaliland people and the political elite to a democratic form of politics cannot easily be questioned or ignored. To do so would make a mockery of the West's commitment to support democracy. As one parliamentary candidate remarked: Something is wrong. In Somalia you just need to kill 100 people to be recognised by the international community as a `player'. But you do not get any recognition if several thousand people vote for you here.40

A lack of support for democratic Somaliland would not go unnoticed in neighbouring countries. And to ignore what has been achieved in a democratic Islamic country would also send the wrong message to Somalia and to countries in the region and the Middle East.

The international community has, to date, shown its support for democratisation in Somaliland by funding the elections. And following the elections, Somaliland received messages of congratulations from several countries and international bodies. These include the Arab League, whose representative visited Hargeisa in October 2005, and the US government which issued a press release from the State Department that made favourable reference to the poll. The United Nations through the Secretary General's Special Representative for Somalia, Ambassador Francoise Lonseny Fall, commended people in Somaliland for the progress they have made towards security and democracy. But words alone will be insufficient. The institutions in place to sustain a democratic system in Somaliland need assistance: these include the NEC, parliament, the judiciary, and the government itself. It will also be important for people in Somaliland to see the benefits of democracy, by investment in the country's infrastructure and services.

In the Somali National Peace Conference in Mbgathi, the strategy of regional and international mediators was to `park' the issue of Somaliland, in order to protect the stability in that region. The message from Somaliland's leadership is that the international community should of course support a resolution to the crisis in the south, but in a way that does not hold Somaliland hostage to developments there. Moves by the African Union earlier in 2005 to investigate Somaliland's claim for independence indicate that other African governments are not averse to giving this serious consideration. In the meantime, the message from the elections is that people in Somaliland are getting on with building a state and hope that the international community will accept and support them in this.

Conclusion

Somaliland has been engaged since 1991 on a journey to build systems of legitimate and accountable governance with some form of social contract with civil society. Lack of international recognition has given Somalilanders the opportunity to build their own system. Their history of conflict resolution has involved a bottom-up approach to building societies from local communities upwards, gradually widening the arena of political agreement and political consensus.

The government signed the Universal Declaration of Human Rights, saw peace restored, demobilised former combatants, brought social and economic rehabilitation, and oversaw the drafting of a constitution based on universal suffrage, decentralisation and multi-partyism. More recently Somaliland has been in transition, seeking to move from traditional institutions of clan governance to western-style governance structures. The future is by no means determined, but despite some setbacks the road to democracy has at least been mapped out - although there are some concerns over corruption, commitment to human rights standards, fair representation for women and minorities, and indeed how Somaliland is responding to the war on terrorism.

These latest elections were, yet again, carried out peacefully - in contrast with the situation in some of Somaliland's neighbouring territories. It is apparent that the process in Somaliland has had exemplary if unintended consequences in Mogadishu and Puntland, and may be a useful lesson for the region's people in pushing for a voice in their governance.

Despite some major problems of resources, time and organisation, the NEC (helped by their extremely hardworking consultants) ran a good election. The diaspora also played a visible role in this election and contributed extensive experience of the democratisation process. However, there were some problems. As with the 2002 elections, transparency was rated more highly than secrecy of the ballot. Problems related to a lack of census and registration process; a largely illiterate population and very complicated ballot papers with symbols for all candidates; governing party use of money, vehicles, fuel, and airtime; attempted multiple voting; and above all, the very unequal representation of women.

Nevertheless, the conduct of the election demonstrates the commitment of Somalilanders to expressing their democratic wishes. The question now arises: how much does Somaliland have to change to come to terms with its own democratisation process? Following the referendum of 2001, local elections of 2002, presidential elections of 2003 and now the parliamentary elections of 2005, many are still assessing how these democratic gains and intentions are working out against the background of clan, patriarchal and business interests (and indeed some tensions emanating from political Islam). In a sense this is a wider story about how the mix of a traditional (and therefore understood) structure and a liberal democratic one changes, and at what speed, and about who controls and wants to control the process. The mix of clan system and autocracy means the government's authority is weak and dependent on its management of clan relations and the patronage of Somaliland's big businessmen.

Given that this was the first parliamentary election in 36 years (and the first time women have been democratically elected to a Somali parliament), Somaliland has some claim to be making a little progress on representation of women. But can a patriarchal clan system, with the strengths and weaknesses that that provides, now listen and respond to the voice of women, and recognise not only the rightness of the case but also the economic power that they wield? The idea of set quotas/reserved seats for women in parliament is being increasingly vociferously raised.

Now that the formal building blocks of democracy are in place, how do the people of Somaliland ensure that their voice is organised, represented and responded to? Civil society outside clan structures is busy organising itself in terms of gender representation, provision of social services, tracking budgets, human rights practice.

How will political parties and government respond to stronger and better organised civil society? How do they see coalition-building, internal discipline, holding the executive to account? Do we yet see signs of differentiations between the parties, their internal dynamics as well as programmes and policies? What steps need to be taken to ensure an independent and trained judiciary? How will the government deal with the economically powerful diaspora and its links to uncontrolled urbanisation? How will the donors react?

Economic development is often said to be the key, given that currently the clan is the support system and that the government has not shown itself capable of providing support for society. Until that situation changes, is a move away from clan lines possible? Can civic education enable voters to pick their own candidates/leaders without sticking to clan lines? Some of the key actors are beginning to move the hybrid form along to reflect more clearly the developing Somaliland - women's groups, civil society, urban youth, some of the business sector. How the more traditional elements - exemplified perhaps by the ruling party UDUB and a number of clan leaders - react to such movement will in reality shape the Somaliland postelection path.

Appendix: Media coverage of the election

CIIR retained the services of a respected local journalist, Adam Musse Jibril, to monitor election coverage by print, radio and television media in Somaliland. As he noted, `One of the key elements for realising peaceful, free and fair elections depends on the neutral, impartial and positive role of the local media, both state-run and private.'

Echoing the code of conduct signed by the three political parties - Article 26 of which affirms that the political parties shall have equal access to the stateowned media during the election campaign, and prohibits abuse of state resources by any party - a media code of conduct was signed by local media representatives. The document provided a legal and moral basis for monitoring media fairness, its impartiality and objectivity.

The media monitor assumed his role with the official nationwide launch of campaigning on 30 August, and continued until 27 September, two days before the elections. Included in the monitoring were:

- The Mandeeq media group comprising: the Somali Daily News, an English language daily paper; Horn of Africa, an English language weekly paper; and Al-qarni- Ifriiqi, an Arabic language weekly paper.
- The Jumhuuriya media group comprising: Jumhuuriya, a Somali language daily paper; and The Republican, an English language weekly paper.
- The Haatuf media group comprising: Haatuf, a Somali language daily paper; Somaliland Times, an English language weekly paper; and Al-Haatuf Al-arabi, an Arabic language weekly paper.
- Somaliland National TV (SNTV) which is a government-run institution, broadcasting mainly in Somali.
- Hargeisa Radio which broadcasts primarily in Somali with additional programmes in English, Arabic and Amharic. Programming is confined to news and commentary, reflecting government viewpoints.
- Somaliland TV (SLTV), a private-sector enterprise, and according to the media monitor, a more professional and efficiently run organisation than its state-owned competitor.
- The BBC Somali service.

It must be noted that the influence of all forms of national media, including the two television stations, is confined mainly to Hargeisa.

Early media coverage

In the pre-election period the monitor noted that the three political parties, UCID, UDUB and Kulmiye, started campaigning vigorously, utilising extensive resources. For example, one UDUB candidate was said to have mobilised 80 buses and trucks in the early days of the campaign. After three days, it became clear that the parties would be unable to sustain this momentum and cross-party agreement was reached that activity would be scaled down.

Early media coverage seemed balanced, including coverage from state media. The monitor noted that Abdilahi Geel-jire of Kulmiye phoned Somaliland National TV to thank them for their unbiased coverage of his party's public rally on the second day of the campaign - coverage described as in line with the media code of conduct. At the same time, Omar Abdi Fidin, a member of the executive committee of UCID party, said it was a good beginning - `the government-run media has given larger space to the opposition than was expected'. Not all agreed that there was complete neutrality, but the general feeling (in the words of the monitor) was that in the first three days of campaigning, `all stakeholders have shown reasonable levels of tolerance, broadmindedness and patience'.

The broadcast media

Thereafter, however, the picture changed, with SNTV in particular adopting a far more pro-government stance. The monitor's assessment was that `SNTV not only tilted towards UDUB but became a forum for government propaganda for campaigning'. He instances that in the programme `Candidate's Forum' (Masraxa Murashaxa), Abdulqadir Ismail Jirdde of UDUB was given 55 minutes airtime compared to Sheikh Abdulaziz Samaale of Kulmiye, who received 13 minutes. Fifteen minutes had been nominally allocated by the NEC for each party candidate.

On 13 September, SNTV devoted the entire evening programme to UDUB. As a result, the Election Monitoring Board (EMB) issued a strong statement on the following day, warning ministers (without mentioning them by name) to stop `provocations' and warning the SNTV authorities to desist from violating the code of conduct. However the same programme was broadcast in the same way the following evening and a further warning produced an `intemperate reaction' from the director of SNTV. The NEC responded with an official letter to the director, copied to the Somaliland president and the three chairs of the political parties, stating that the director of SNTV was not abiding by the code of conduct. This did result in some improvement. SLTV are thought to favour the opposition and Kulmiye in particular, but their coverage was markedly more neutral, `giving space to the ruling party'.

Radio Hargeisa did hold to the code of conduct and give each political party some airtime, but this was not entirely equal. This was particularly the case in the programme `Be acquainted with your candidate' (Baro Murshaxaaga) in which candidates from different regions, parties and clan affiliations were interviewed. On 13 September - in a manner strongly reminiscent of SNTV coverage - Abdulqadir Ismail Jirdde of UDUB was given 46 minutes airtime, while Sheikh Abdulaziz Samaale of Kulmiye was allowed only 13 minutes, less than the required 15 minutes. Another case in point was that on 16 September, Ms Ikran Haji Daud, a Kulmiye candidate, got only 12 minutes airtime, whereas Ms Hodan Abdi Husein of UDUB got 21 minutes.

In the opinion of the monitor, the BBC Somali service offered notably fair and impartial coverage. He notes that `as a result the BBC Somali section has acquired a further excellent reputation amongst Somalilanders'.

Newspapers

Both Jumhuuriya and Haatuf are private newspapers inclined towards the opposition and critical of the government, but `playing a positive role in the democratic process'. Their coverage of the UDUB campaign was reasonable, although critical of the government's partisan role in the campaign, and also of the use of national resources in support of the ruling party.

General comments

There was a general failure to give equal broadcast airtime or equitable print coverage to all parties, reflecting the fact that different media organisations owe their allegiance to different political camps. The Jumhuuriya and Haatuf media groups and Somaliland TV tended to favour the opposition, while the state-owned media, such as the Mandeeq media group and Somaliland National TV, generally leaned towards the `government party'.

The media code of conduct, though signed by all parties, was not fully respected. However, it created legal and moral grounds for dealing with those who attempted to avoid fulfilling their agreed obligations, thus reducing tensions and contributing to a climate of relative tolerance. The local media in general played a positive influence that was reflected in the public's enthusiastic attitude towards the elections, and the high turn-out.

However, media coverage of women's participation was `at the minimum scale' despite each female candidate being interviewed by Radio Hargeisa and Somaliland TV. Both SNTV and SLTV publicised a debate on women's rights, organised by local NGO the East African Human Rights Initiative, and funded by the Africa Educational Trust.

Recommendations

Local media staff are under-trained and most lack higher education. Further educational and professional training and guidance on acceptable practice are needed to achieve and maintain impartiality and fairness. In particular, the independent private media organisations need capacity building assistance. The new parliamentarians will also need training in order to raise their legal and political consciousness so that they are aware of the importance to the democratisation process of a free and impartial media.

NOTES

1 The September 2005 elections were to the house of representatives. The upper house of parliament - the house of elders or Guurti - is not at present an elected body (see chapter 7 for more on the Guurti).
2 Advocates of political change in Djibouti have been closely following developments in Somaliland believing that it provides a model for Djibouti (interview, Hargeisa, October 2005).
3 Iqbal Jhazbhay, South African election observer.
4 CIIR was known in certain countries (including Somaliland) as International Cooperation for Development (ICD). From January 2006 CIIR has adopted the working name Progressio.
5 International Election Observation team interim report on the Somaliland elections to the House of Representatives 29 September 2005, issued 7 October 2005, available at www.progressio.org.uk. The findings and recommendations of the interim report have been incorporated in this final report.
6 Somaliland Ministry of Foreign Affairs (August 2002) The case for Somaliland's international recognition as an independent state, Somaliland Government, Hargeisa.
7 The five regions had been designated districts under the Somaliland protectorate.
8 For more in-depth histories of Somaliland see: Bradbury, M (1997) Somaliland country report, CIIR, London; WSP-International (2005) Rebuilding Somaliland: Issues and possibilities, Red Sea Press, Hargeisa/Lawrenceville, NJ.
9 The Somaliland Ministry of Planning estimates the population to be three million, giving a population density of 22 people per square kilometre. See Ministry of National Planning and Coordination Somaliland in figures 2004, Hargeisa. The voter turn-out in the district elections suggest the population may be nearer to 1.7-1.8 million in the areas that voted.
10 UNDP (2001) Somalia Human Development Report, 2001, United Nations Development Programme, Nairobi.
11 Some 33 clan peace conferences took place in Somaliland between February 1991 and 1996.
12 See Bradbury (1997) for a description of the Borama conference.
13 See WSP International (2005) for a description of this system.
14 Puntland State of Somalia was created in 1998 as a non-secessionist state in north-east Somalia and is based on a confederation of Harti clans that include the Dhulbahunte and Warsengeli people in eastern Somaliland. Puntland consequently claims sovereignty over eastern Sanaag and Sool regions of Somaliland. The Transitional National Government of Somalia claimed sovereignty over the whole of Somalia including Somaliland.
15 While it is not disputed that the great majority of the public endorsed the constitution, the administration's claim that 1.18 million people voted and 97.9% approved the constitution is highly questionable, given the limited voting that took place in Sool and eastern Sanaag regions and in light of the smaller turn-outs in the various subsequent elections.
16 For a description of the district and presidential elections see Bradbury, M, Abokor, A, and Yusuf, H (2003) `Somaliland: Choosing politics over violence' in Review of African Political Economy No.97:455-478, ROAPE Publications Ltd, 2003; and International Crisis Group (July 2003) Somaliland: Democratisation and its discontents, Africa Report No 66, Nairobi/Brussels.
17 This final figure approved by the Supreme Court was an increase on the margin of 80 votes originally announced by the NEC.
18 Abokor, A, et al (2003) `Very much a Somaliland-run election': A report of the Somaliland local elections of December 2002, CIIR, London.
19 Several European donor governments and the US government provided matching funds for the district elections and the presidential elections, and the European Commission supported the district and parliamentary elections.
20 Since 1991, 20 new districts and one new region (Saaxil) have been created by the government in Somaliland due to patronage politics. Many of them have not been formally accepted by parliament.
21 The elections were postponed in 2003 when MPs from Gadabursi walked out when they calculated that the regional distribution of seats would leave them with fewer seats to contest than they currently enjoyed.
22 At the last minute five Issa candidates from Zaylac area withdrew from the elections when they were refused a quota of seats in parliament.
23 In addition to financing the costs of the NEC and technical support to it, donors funded the training of election staff, voter education through local NGOs, and the training of political parties.
24 WSP International's support for the Somaliland elections is part of a democratisation project for the whole of Somalia.
25 Interviews, Hargeisa and Gabiley, September and October 2005.
26 Interviews, Hargeisa and Gabiley, September and October 2005.
27 UCID 3, UDUB 2, Kulmiye 2.
28 International Election Observation team interim report on the Somaliland elections to the House of Representatives 29 September 2005 (see note 5).
29 The constitution allows for a period before the final confirmation during which parties can raise objections.
30 See Bradbury, M, Abokor, A, and Yusuf, H (2003) `Somaliland: Choosing politics over violence' in Review of African Political Economy No.97:455-478, ROAPE Publications Ltd, 2003.
31 Bradbury et al (2003).
32 Bradbury et al (2003).
33 There is some evidence that the longer lineages (laandheere), often considered the more powerful lineages, may have lost out to the shorter lineages (laangaab) due to the former fielding too many candidates.
34 NEGAAD Umbrella et al (18 October 2005) `A joint appeal of civil society groups towards improving the political position of women in Somaliland', Hargeisa.
35 For an exposition of this refer to the classic monograph by Lewis, I (1961) A pastoral democracy: A study of pastoralism and politics among the northern Somali of the Horn of Africa, Oxford University Press, London.
36 Interviews, Hargeisa, October 2005.
37 African Rights (May 2003) Somaliland: Shadows of the past as human rights deteriorate, Discussion paper no 11, Hargeisa.
38 NEGAAD Umbrella et al (18 October 2005) `A joint appeal of civil society groups towards improving the political position of women in Somaliland', Hargeisa.
39 Abdi Ismael Samatar (2005 November) Somali prime minister: An act of betrayal, venality, and incompetence! University of Minnesota.
40 Interview, October 2005.

Since breaking away from Somalia in May 1991, the people of Somaliland have sought to build a new state by charting a path away from violent conflict towards a democratic political system.

As the latest step in this process, parliamentary elections - the first in the Somali region since 1969 - were held in Somaliland in September 2005. This report, drawing on the findings of a team of international election observers coordinated by Progressio, describes the conduct of the elections, analyses the results, and examines their significance for the democratisation process in Somaliland.

Progressio is an international development organisation working for sustainable development and the eradication of poverty