IRI recognizes and applauds the efforts by the government of Somaliland, the NEC, the political parties, and the people of Somaliland in making these elections a success. Regardless of all the difficulties that occurred, in the September 2005 parliamentary election, the people of Somaliland were able to cast their ballots in a peaceful manner and without significant intimidation.

It was inspiring for IRI observers to see the enthusiastic faces of the citizens of Somaliland on Election Day, braving the scorching sun and pouring rain to make sure that their votes were counted – for political stability, peace, and development.

International Republican Institute

Somaliland September 29, 2005 Parliamentary Election Assessment Report


Table of Contents

Executive Summary

    1. Background Information
    2. Legal and Administrative Framework
    3. Pre-Election Period
    4. Election Day
    5. Post-Election Period and Results
    6. Findings and Recommendations…
    7. Conclusion

Appendix A: Voting Results in 2005 Presidential Elections

Appendix B: Voting Results in 2003 Presidential Elections

Appendix C: Voting Results in 2002 Local Government Elections

Appendix D: Voting Trends

Executive Summary


The International Republican Institute (IRI) has conducted programs in Somaliland since 2002 with the support of the U.S. Agency for International Development (USAID), the U.S. Department of State, and the National Endowment for Democracy (NED).  IRI’s Somaliland programs have included capacity-building workshops and consultations with Somaliland’s three political parties and numerous leaders of youth and women’s advocacy organizations.  In the run-up to the September 29, 2005, parliamentary elections, IRI also supported trainings for political party poll-watchers and the parties’ national and regional campaign committees, as well as the production of an instruction manual for election officials, party agents, and security officers.

In the fall of 2005, IRI received funding from USAID to dispatch a seven-member election assessment mission to Somaliland, led by former U.S. Ambassador to Djibouti Lange Schermerhorn, who also participated in the September 2002 assessment mission that marked the beginning of IRI’s Somaliland program.  The delegation worked closely with the Catholic Institute for International Relations (CIIR), which was designated by Somaliland’s National Electoral Commission (NEC) to coordinate the more than 70 international election observers.  On Election Day, members of IRI’s delegation split into seven teams, joining with other international observers, and traveled to four of Somaliland’s six regions.  In the locations they visited, IRI delegation members observed the opening of polling stations, voting procedures, ballot tabulation and the reporting of results.  In all, IRI representatives visited 94 polling stations, roughly 10 percent of Somaliland’s 985 voting locations.


The IRI team applauds Somaliland on the conduct of this election, a significant accomplishment for the territory.  Institute observers witnessed no major violence or unrest on voting day.  Lines of 200-300 voters at the opening of polling stations indicated the great excitement surrounding the election.  Widespread procedural difficulties were observed, but the IRI team found little evidence of violence or intimidation and was satisfied that most voters were able to cast their ballots without undue interference.

The NEC should be commended for its efforts to facilitate voting amid logistically difficult circumstances and tight security constraints, and with very limited resources.  Expectations for the NEC were great prior to the election and IRI believes that the commission met these expectations.  All of the polling stations IRI observed were well-equipped with election materials.

During the pre-election period, IRI noted several violations of the Political Party and Media Codes of Conduct.  Personal attacks on candidates standing for office were reported during the campaign period and the ruling party was harshly criticized for using state resources to support its media campaign and to transport high-profile campaigners.

On Election Day, IRI recorded multiple sightings of poll workers wearing political party paraphernalia and polling stations with campaign flags or posters prominently displayed.

In IRI’s assessment, candidate campaigns in Somaliland generally are in the early stages of development.  In this election, the political parties did not effectively promote issue based platforms to voters during the campaigns, and candidates appeared to be selected mostly based on clan allegiances rather than accomplishments or ideology.  Campaigning was largely conducted along clan lines.  From interviews with voters, IRI found that many could not associate candidates with particular issues and were uninformed about the role and importance of parliament as an institution.

Efforts to vote multiple times were common and grew in number throughout Election Day, particularly among young people.  IRI observers noted a number of well-organized efforts to move voters between polling stations to facilitate multiple voting.  Somaliland electoral law restricts vehicle movement on Election Day only to those vehicles authorized by NEC; in all regions, IRI visited, observers spotted large numbers of trucks and other vehicles full of voters.

Those attempting to vote multiple times slowed the process for those who were queuing to cast ballots legitimately.  Such persons created a chaotic atmosphere as the lines grew in length and the police grew busier ejecting them from the polling stations.  Although in the overwhelming majority of cases attempted multiple voters were caught through the use of indelible ink and ultra-violet lamps, in some isolated cases, IRI observed poll workers irregularly checking voters’ fingers for ink.  Additionally, the concept that multiple voting is illegal and immoral did not appear to resonate with the many voters who seemed to find great amusement in trying to vote more than once.

IRI also observed that in many polling stations, especially those in urban centers, insufficient security personnel were provided to control unruly crowds, particularly later in the day when voters became concerned that they would not be able to vote before polls closed.  Finally, IRI noted that in certain districts, notably some in Awdal Region, there was a large increase in the number of voters since the 2003 presidential election, though not on a scale large enough to significantly influence the nationwide result.  Unfortunately, the NEC was not given sufficient power to investigate cases such as this and to evaluate whether certain results were unacceptable.

IRI’s most urgent recommendation is that central voter registration be created in Somaliland to ensure that multiple voting is more difficult in future elections.  In addition, although women turned out in large numbers to vote, women should be more prominently included in the campaign processes in future elections.  Among the seven-woman candidates, only two won parliamentary seats.  Support for women candidates needs to be bolstered in future campaigns.

I.         Background Information

Somaliland is a self-declared republic located in the northwestern region of the former Republic of Somalia.  It shares borders with Djibouti, Ethiopia and Somalia and the capital, Hargeisa, is home to an estimated population of 500,000 Somalilanders.  Although an official census has not been taken in Somaliland since independence in 1960, it is estimated that nearly 3.5 million people live within its borders.  Somaliland’s largest source of income, other than remittances from Somalilanders living abroad, is its livestock; most Somalilanders are nomadic or semi-nomadic pastoralists.  In addition, virtually all Somalilanders are Sunni Muslims.

Somali society is organized around the clan system. Unlike most African countries, Somaliland is ethnically, culturally, and religiously homogenous. However, divisions exist within a society based on ancestral lineages, which are the basis for clans and sub-clans.  The largest clan in Somaliland is the Issaq clan, which is concentrated around Somaliland’s relatively densely-populated central plateau.  Other regions of Somaliland are dominated by less-populous clans, such as the Gadabursi (western Somaliland, along the Ethiopian border) and the Dhulbhante and Warsengeli (eastern Somaliland, near Puntland).


From 1884 to 1960, Somaliland existed within its current borders as the British Somaliland Protectorate.  In June 1960, both Somaliland and its neighbor to the south, Italian Somalia, were granted independence by their colonizers.  After existing for one week as an internationally-recognized independent state, Somaliland entered into a formal union with the former Italian colony to form the Republic of Somalia.  The union was troubled from the beginning, and the relationship between the two partners deteriorated rapidly after Mohamed Siad Barre, an army commander from a rival clan of Somaliland’s dominant Issaq clan, rose to power in a 1969 coup d’etat.

Under Siad Barre’s dictatorship, northwestern Somalia was subjected to violent oppression and infiltrated by his feared intelligence services.  In response, dissidents in Somaliland re-energized the Somali National Movement (SNM), which had been formed during Somaliland’s independence struggle, in an attempt to overthrow Siad Barre and to liberate Somaliland.  In 1988, Siad ordered the city of Hargeisa bombed, in an effort to crush the SNM.  However, this type of brutal suppression only served to strengthen Somaliland’s resistance and desire for freedom. While the South descended into anarchy and was overrun by warlords, after the 1991 collapse of Siad Barre’s regime, Somaliland began to build the institutions of an independent democratic state.


After the collapse of the Republic of Somalia, the SNM faced the onerous task of uniting and stabilizing Somaliland.  In April 1991, a conference of clan elders convened by the SNM in Burao issued a declaration of independence and a national charter.  The charter called for the SNM to act as Somaliland’s government during a two-year transitional period, at the end of which a constitution would be drafted and approved by popular referendum.  Under this constitution, Somaliland would transfer power to a civilian, democratically-elected government.  However, less than a year after the national charter was signed, the SNM dissolved into belligerent factions, plunging Somaliland into a nine-month civil war.  Finally, in October 1992, a ceasefire was implemented.

In January 1996, a second national conference was convened by clan elders in the western town of Borama.  This five-month conference produced a new national charter, a peace accord, and a new civilian government under the leadership of Mohamed Ibrahim Egal, a former prime minister of Somalia.  This conference also produced a bicameral parliament that embodied the republic’s quest to achieve stability by combining traditional, clan-based, decision-making systems with Western-style political institutions.  The lower house, known as the House of Representatives, was to be the primary legislation-crafting body; the upper house, known as the guurti or House of Elders, was to serve as a locus of moral authority, responsible for maintaining peace and equitably resolving conflicts among the clans.

Despite the breakthroughs achieved by the Borama conference, tensions among the clans over issues of power-sharing led to a resumption of civil conflict from 1994 to 1996.  In December 1996, a third national conference was held in Hargeisa that ended the conflict, addressed the grievances of certain clans, and re-appointed Egal to a five-year term as president.  The conference resulted in Somaliland’s first sustained period of peace, reconstruction, and economic growth, continuing to the present day.

During his second term, Egal finally followed through on his commitment to produce Somaliland’s first constitution.  In May 2001, voters in Somaliland voted in a referendum to approve the new constitution.  The document received the support of 97 percent of the electorate, a result that was widely seen as a mandate for an independent Somaliland.  Under the new constitution, Somaliland’s first municipal elections were to be held in December 2001, and presidential elections in March 2002.  However, delays in passing an election law and establishing an election commission necessitated a one-year delay.

In a tragic turn of events, President Egal died suddenly in May 2002, while under medical treatment in South Africa.  As provided for in the constitution, Egal was succeeded by Vice President Dahir Rayale Kahin, and the municipal and presidential elections were held as scheduled.  Rayale’s UDUB (Union of Democrats) party was victorious in the December 2002 municipal elections, Somaliland’s first multiparty poll. In the March 2003 presidential elections, Rayale defeated his closest challenger by a margin of only 80 votes out of nearly half-a-million cast.  After mounting an unsuccessful court challenge, the opposition KULMIYE (Solidarity) party accepted the results and conceded defeat in the election.

After the presidential polls, one major challenge remained in completing Somaliland’s transition to electoral democracy: parliamentary elections.  Under Somaliland’s constitution, the lower House of Representatives is directly elected by the people of Somaliland while the guurti is chosen by clan leaders.  However, most members of the House of Representatives had held their seats since the institution was established in 1993 without ever having to run for election.  Since these members had been nominated by clans rather than political parties, Somaliland remained a virtual one-party state, and the two opposition parties, despite winning more than half of the votes in the presidential election, held no national elected office or cabinet position.

Political Parties

Before the ratification of the 2001 constitution, multi-party politics did not exist in Somaliland.  There was a high level of concern among Somaliland’s leaders that introducing an open environment for political parties would lead to clan warfare by other means, as each clan and sub-clan formed its own party.  In order to prevent such an outcome, the 2001 constitution and the electoral law stipulated the following:

  • Somalilanders were free to form political parties to contest in the first municipal elections;
  • However, only the top three vote-getting parties in the municipal elections would be allowed to compete in the presidential election and all subsequent elections.

The three parties who won the right to continue political activities as a result of their performance in the 2002 municipal elections were KULMIYE, UCID (Party of Justice and Welfare) and UDUB.

KULMIYE (Solidarity) is Somaliland’s largest opposition party.  Its chairman, A.M. Sillanyo, is a former commander of the SNM who held several ministerial posts in President Egal’s administration.  The party’s top ranks include many war veterans who were active in the campaign against Siyad Barre for Somaliland’s independence.  KULMIYE has based its appeal on the patriotic reputation of its leaders and on its efforts to reach out to women and youth.

UCID (Party of Justice and Welfare) is Somaliland’s second largest opposition party.  Its chairman, Faisal Warabe, is a Somaliland-born Finnish civil engineer.  Unlike the other two parties, UCID was formed to promote an issue-based agenda: the party advocates the ideals of liberal democracy and the introduction of a European-style welfare state.  In the first two elections, UCID placed a distant third, with 11 percent and 16 percent of the vote, respectively.  However, UCID has positioned itself as a “swing” party by tempting the other two parties with the prospect of a coalition, which would break the stalemate between KULMIYE and UDUB.  UCID has performed best with urban voters, especially those in the capital area.

UDUB (Union of Democrats) has held the presidency since the party’s formation in 2000 and is officially chaired by Somaliland’s current President, Dahir Rayale Kahin.  Rayale served as an intelligence officer in Siyad Barre’s government and as Vice President of Somaliland under Egal.  UDUB promotes itself as the party of stability and has based its appeal on the accomplishments of Presidents Egal and Rayale, chief among which is the implementation of peaceful democratic elections.  UDUB’s main burdens have been the identification of the party with certain unpopular government policies and ministers, and a perception that UDUB has not done enough to reach out to women and young voters.  In the past two elections, there were also serious accusations by the opposition parties that UDUB was given unfair access to state-owned media and government resources.  UDUB has performed best in the president’s home region of Awdal and in Sahil, the region from which the Vice President hails.

 II. Legal and Administrative Framework 

The following documents compose the framework for the September 29, 2005 parliamentary election:

  • The constitution of the Republic of Somaliland, which was ratified by popular referendum in 2001;
  • The National Electoral Law, which was approved by Somaliland’s parliament in April 2005; and
  • The Political Parties Code of Conduct, which was signed by all three parties, the NEC, and the Vice President in July 2005.

Constitution of the Republic of Somaliland

Somaliland’s constitution establishes a presidential system of government with separate legislative, executive and judicial branches, subject to a range of checks and balances.   Members of parliament serve five-year terms, beginning from the date when the Supreme Court declares the election results.  The President is required to announce the election of a new House of Representatives at least one month before the outgoing House’s term of office expires.  There is a provision in the constitution that the House of Representatives may serve past its term of office in the event of “dire circumstances,” which are defined as “a wide war, internal instability, serious natural disasters, such as earthquakes, epidemic diseases or serious famines.” Such circumstances are to be determined by the House of Elders on the proposal of the cabinet.  Each newly elected House of Representatives is to be inaugurated and convened by the President within 30 days after the electoral results are declared by the Supreme Court.

House of Representatives Election Law 

Drafting Process

Prior to the 2005 parliamentary elections, the sitting parliament had been selected on the basis of clan before the ratification of the 2001 constitution.  The sitting parliament, made up of unelected clan representatives, was responsible for drafting the law under which Somaliland’s first post-independence parliamentary elections would be held.  Many observers noted that this was akin to requiring that members of parliament draft their own “death warrant.”  Not surprisingly, the process was quite contentious, with considerable delays on behalf of the lame-duck members.

The House of Representatives finally passed an Election Bill in January 2005, but the bill contained two “poison pills” making it virtually impossible to hold elections.  First, the bill required that all voters be registered before Election Day, a difficult process that would take many months, perhaps even a year or two.  Second, the bill stipulated that voting be held in all areas of Somaliland, an impossibility given that parts of Sool region were under the control of militias based in neighboring Puntland.

Despite intense lobbying by Somaliland’s civil society and political parties, the bill was passed by veto-proof two-thirds majorities in both houses of parliament and sent to the President to be ratified.  The President exercised his final option by sending the bill to the Supreme Court to decide on its constitutionality.  In February 2005, the court ruled that the bill violated the constitution by rendering it technically impossible to hold the election within the mandated period.  The court ordered the House of Representatives to redraft the bill.  In April 2005, a new law was passed without the controversial clauses.

Voters and Candidates

Under the new Election Law, citizens of Somaliland are eligible to vote if they are 16 years of age or older in the year of an election and are not incarcerated.  Since there is not yet a system of voter registration in Somaliland, voters are free to cast their ballots in whatever location is most convenient to them.

According to the Election Law, candidates are nominated by majority vote of the national executive committee of each political party, upon receiving the recommendations of the party’s regional committees.  Each candidate is required to pay a non-refundable deposit of 1 million Somaliland shillings (approximately US$150) to the Ministry of Finance.  All candidates’ names have to be submitted to the NEC at least 60 days before the election.  The NEC is to verify the qualifications of the candidates and publish a final list at least 45 days before the election.

Distribution of Seats

The allocation of seats between regions was one of the most contentious issues in the drafting of the second and final draft of the Election Law.  Leaders of all three political parties agreed that for the sake of expediting the process, and in the absence of reliable census figures, the proportional distribution of seats should be identical to that of Somaliland’s last parliamentary election, held in 1960.  Since the original parliament had 33 seats and the current one has 82, the parties agreed that each region should see its seats multiplied by a factor of 2.5.  This resulted in the following distribution:

Hargeisa Region 20 seats
Togdheer Region 15 seats
Awdal region 13 seats
Sanaag Region 12 seats
Sool Region 12 seats
Sahil Region 10 seats

The Voting System and the Ballot

In the municipal elections of 2002, Somaliland employed a proportional representation system in which voters cast their votes for specific parties and positions were filled on the basis of sequential party lists.  However, the 2005 Electoral Law established a unique voting system that combined proportional representation with voting for specific candidates, incorporating elements of an open party list system.  The format of the ballot was as follows:

# Name Symbol Vote # Name Symbol Vote # Name Symbol Vote
1 A —— 5 E —— 9 I ——
2 B —— 6 F —— 10 J ——
3 C —— 7 G —— 11 K ——
4 D —— 8 H —— 12 L ——

Since a different set of candidates ran in each of Somaliland six regions, six different ballots were printed.  In order to assist illiterate voters, each candidate was assigned a different pictorial symbol.

Voters were allowed to choose only one candidate and mark their vote next to his or her name.  However, each vote was counted twice.  The first count was to determine the total votes received for each party.  These totals were used to calculate the number of seats each party won in that region.  For instance, in the above illustration, the region has four open seats.  If UCID wins 50 percent of the votes, KULMIYE 25 percent and UDUB 25 percent, then UCID wins two seats, and the other two parties win one seat each.

The second count established the number of votes received by each candidate, which was used to determine which candidates won election.  Let’s say the results were as follows:

Name Total Votes Name Total Votes Name Total Votes
A 45 E 23 I 37
B 76 F 4 J 19
C 52 G 31 K 28
D 27 H 42 L 16

In this case, the two seats allocated to UCID (which won 50 percent of the vote) go to the top two vote-getters, candidates B and C, the one seat allocated to KULMIYE goes to candidate H and the remaining seat goes to UDUB’s candidate I.


The official campaign period began 30 days before Election Day and ended 24 hours before Election Day.  The Electoral Law addressed one of the main concerns of the 2002 and 2003 elections: the alleged use by UDUB of public resources to conduct its campaigns.  In the 2005 campaign, all parties were to be given equal access to state-owned media and public assembly grounds, to be allocated in consultation with the NEC.  All parties were prohibited from using public property to advance themselves.  However, article 23 of the law also gave significant power to town mayors in regulating campaign events.  In order to hold a rally in a particular town, a party had to inform the mayor at least 48 hours in advance.  If two or more parties requested to hold rallies on the same day, the mayor could approve one party’s rally and order the other party or parties to rally on different days.  No two parties could hold rallies in the same town on the same day.  The mayor could also ban rallies altogether if he was “satisfied that they might damage the health, morals or public order, etc.”

Political Parties Code of Conduct

After the Election Law was passed, a substantive discussion began among the three political parties, the NEC and the government aimed at producing a code of conduct to which the parties would adhere during the pre-campaign, campaign, election, and postelection periods.  The discussion was also meant to produce an agreement on how to address loopholes in the Electoral Law.  This dialogue supported by the War-Torn Societies Project (WSP) and organized by the Hargeisa-based Academy for Peace and Development (APD), took place through meetings held twice a week over the course of several months.

On July 27, 2005, the three political parties (along with the NEC and the Vice President on behalf of the government) signed a final version of the code of conduct.  By signing the Political Parties Code of Conduct, all parties promised that during the campaign period they would: respect the freedom of others to campaign and to disseminate ideas through banners, billboards and printed materials; respect the rights of the press and election monitors; recognize the authority of the NEC in conducting the election and follow its lawful orders; ensure peaceful and orderly polling and respect voters’ freedom to vote without interference; establish and maintain lines of communication between the parties and the NEC; ensure the security of election officials and avoid interfering with their duties; cooperate with election officials, observers and monitors; ensure secrecy in voting; and refrain from occupying polling stations or other illegal activities to procure votes.  Perhaps most importantly, the code of conduct established a framework for enforcing the electoral law’s stipulation of equal access to state-owned media during the official campaign period, and established an Election Board of Monitors (EBM), under the management of the NEC, to identify and investigate violations of the law and/or the code of conduct, and to report its findings to the NEC.

The parties also agreed not to: engage in activities to jeopardize peace and stability; use defamatory language that might provoke violence; intimidate candidates or members of other parties; disrupt, destroy, or prevent the campaign activities of other parties; destroy or prevent the distribution of other parties’ campaign materials; or carry weapons to any public campaign event.

In negotiating the code of conduct, the parties also identified deficient areas of the electoral law and agreed upon ways to improve or clarify the law.  For example, the law failed to say what would happen if two or more candidates received the same amount of votes but there was only one seat to be allocated.  The parties agreed that such a situation would be resolved through a public drawing of lots by NEC officials.  To give another example, the law made no provision for the death or incapacitation of a candidate after the candidate list was published.  The parties agreed that if this were to occur 14 or more days before the election, the relevant party would have the right to nominate a replacement, and all votes cast for the deceased or incapacitated candidate would go to the replacement.  If the death/incapacitation were to occur closer to the election, the party would not be able to replace the candidate, but all votes cast for him would be awarded to the party in the first count.  The Electoral Law stipulated that all three parties would have equal access to state-owned media during the official campaign period.  The code of conduct also provided the necessary structure for enforcing this stipulation.  The code of conduct further augmented the Electoral Law by providing a mechanism to enforce the law’s prohibition on the use of state resources to benefit particular parties or candidates.

 III.  The Pre-Election Period

Democracy Has Its Price

As it began preparing for the 2005 election, the NEC faced significant financial challenges.  Somaliland’s government, whose total annual budget is roughly US$25 million, could only spare about US$500,000 for the election.  To meet the anticipated shortfall, a group of interested donors formed a Somali Democratization Program Steering Committee, which met periodically in Nairobi to coordinate international support for the election.  The committee’s membership included representatives from the European Commission, WSP International, and the governments of the United Kingdom, Switzerland, Norway, Denmark, Italy, and the United States.  IRI participated in meetings of the steering committee, but was not a formal member.

None of the countries represented on the steering committee recognized Somaliland’s independence, or the government of Somaliland as a legitimate entity.  The solution to this dilemma was to classify support as the first phase of a Somali Democratization Project that would also attempt to promote democratic processes in Puntland and the rest of southern Somalia.  Funds would be filtered through an international nongovernmental organization (NGO) that would monitor and assist in the implementation of the elections.  The WSP was chosen to be this implementing partner.  In addition to frequent visits by its Nairobi staff, WSP hired three international consultants (from Denmark, El Salvador and South Africa) who were stationed in Hargeisa for several months before Election Day and were intimately involved in the implementation of election procedures.  About 70 percent of the total budget for the election was delivered through these mechanisms.

Nomination of Candidates: The Parties Lose Power and the Clans Get Involved

The switch from a clan-selected parliament to a popularly elected one caused much anxiety among the clans.  In the previous system, each clan was guaranteed a certain level of representation in the legislature, but in the new system clans would have to compete with one another for scarce positions.  This change was meant not only to introduce a greater degree of democracy to the legislature, but also to facilitate the growth of a strong clan-transcending political party system in a country where the concept is still quite new.

Also, because the ranked party list system of the 2002 municipal elections had been abandoned, in this election it would be the voters, not the parties, that decided the order in which candidates would take their seats.  This decision removed the power of the parties to decide in which order their candidates would join parliament.  It also made it imperative for the clans to convince political parties to nominate their favored candidates so as to guarantee representation in parliament.

In the 2003 presidential election, political parties counted on their presidential candidates to bring with them formidable clan blocs from their home regions.  However, in this election, all parties had to run their candidates against rival party candidates from the same clans, even in their base regions.  Furthermore, the scarcity of funds for party campaigns meant that at the local level, candidates had to rely on their own clans for financial backing and manpower.  Thus, the nomination process was less about choosing candidates who would be loyal to their parties and more about choosing candidates who could command the largest ethnic numbers and strongest financial support from their clans.  As a result, the parties ceded much of their control over the nomination process to clan leaders.

According to the Election Law, all names of nominated candidates were to be submitted by July 15, 2005, exactly two months before Election Day.  As the deadline approached, the three political parties scrambled to complete the lists of names they hoped would bring them victory.

Exclusion of Women

Among the greatest losers in the nomination process were women candidates.  Women’s advocacy groups welcomed parliamentary elections as women’s first real chance to gain national elected office.  During the drafting of the Electoral Law they pushed heavily for quotas to be established guaranteeing parliamentary seats for women.  When they were rebuffed by the parliament, they turned their energy toward obtaining guarantees from the parties themselves that a large number of women would be nominated as candidates.

However, with clan leaders exercising great control over the nomination process, it was inevitable that women would be sidelined.  First, tradition-minded clan elders did not see women as appropriate standard-bearers for the clan.  Second, there was the issue of trust: in Somali society, women’s loyalties are perceived as being divided between the clan of their fathers and that of their husbands.  In the end, only seven women were nominated for parliamentary seats, along with 239 men.  Beyond the influence of the clan, IRI also heard reports from party leaders and other stakeholders that many of the most promising and influential women in Somaliland, including the leaders of prominent women’s groups, declined to stand for election, thereby further decreasing the chances for women to pick up parliamentary seats.   It appears that the largest factor in these women’s decision not to run was their lack of access to resources and the difficulty of fundraising outside the clan system.

National Party Campaigns

The official campaign season began on August 29, 2005, exactly one month before the election.  As agreed under the Political Parties Code of Conduct, this 30-day period was divided equally; each day of a repeating, three-day cycle was set aside for campaigning by one of the three parties.

By all reports, the campaign season was generally peaceful.  The division of days among the parties was not strictly observed, especially outside the major cities, but it was followed closely enough to avoid direct confrontations between campaigners of different parties.  IRI received no reports that national or local authorities had prevented parties from conducting legal campaign activities, and all parties were satisfied that they had been able to campaign freely without undue interference.

Several months before the campaign season began, all three parties set up two-tiered campaign organizations at the national and regional levels.  At the national level, each party’s national campaign committee (NCC) would raise money, designate broad themes for the party campaign, direct resources to campaign activities and organize rallies, trips and broadcasts.  At the regional level, regional campaign committees (RCCs) would assist and provide advice to candidates on their campaigns.  They would organize locallevel campaign events and ensure that the NCC’s messages and themes were promoted at the grassroots.  From June to August, IRI conducted a series of training workshops with members of these committees, instructing them on techniques of message development, communications, fundraising, and campaign planning.

However, with the start of the campaign period, many of these structures were neglected or failed to function.  At the national level, party leaders proved unwilling to cede control over the campaign to a committee that was peripheral to central party organs.  As a result, most campaign decisions were made by the party leaders and implemented by the parties’ national executive committees, and the NCCs (whose membership often overlapped with that of the executive committees) were relegated to minor roles.

As with previous campaigns, efforts were minimal by the political parties to differentiate themselves on the basis of competing policy platforms or positions on contentious issues of national importance.  Where issues were raised at rallies and in broadcasts, they were for the most part issues on which all three parties agreed, such as the quest for recognition, support for Somaliland’s sovereignty, and condemnation of the invasion of Eastern Somaliland by various Puntland-based militias.

Most campaigning focused on introducing each party’s candidates and publicizing their qualifications and clan backgrounds.  In addition, national party campaigns focused on spreading negative perceptions about rival parties and their leaders; attempts were made to link UDUB with unpopular policies and cabinet ministers; KULMIYE leaders were accused of seeking to overthrow the government and of being in secret negotiations with politicians in southern Somalia; UCID was attacked for being a party of émigrés, out of touch with day-to-day realities in Somaliland.  There was little talk of key bread-andbutter issues facing Somalilanders, such as employment, the provision of water and electricity and economic growth.  Beyond vague condemnations by the opposition parties of the government’s corruption and incompetence, few specific proposals for institutional reform were raised.

This lack of platform-based campaigning did not reflect a disinterest among party leaders in key issues facing Somaliland.  To the contrary, in interviews with party leaders, IRI found that they had very clear ideas, positions and priorities on these issues.  This was especially true of the opposition parties; leaders of the party in power were notably less willing to discuss the party’s positions and its post-election legislative agenda with IRI.  It would appear that, regardless of the existence of different visions among the party leaders, party campaigners understood that this election would not be about policy issues and platforms.  At the local and regional level, this election was seen as a competition among clans to maintain their levels of representation in parliament despite moving away from a system in which these levels were guaranteed.

At the national level, the election was also an opportunity for the parties to establish and solidify public perception of their overall strengths:

  • UDUB marketed itself as the party whose leaders had directed Somaliland through years of peace, democratization, and economic growth. The party promoted itself as the only one that could be trusted to lead Somaliland.
  • KULMIYE sought to establish itself as the patriotic alternative to UDUB- led by liberation heroes of great prestige- that could restore integrity to government, and move it toward greater openness and transparency. The party also aimed to maintain its strong following among women and youth.
  • UCID presented itself as a party of intellectuals and professionals, bringing expertise from the West that could accelerate Somaliland’s development and modernization. It also sought to establish itself as a middle ground option that would draw on voters disillusioned with UDUB but unable to make the leap to KULMIYE.

 Grassroots Campaigning

Party campaign structures at the regional level proved even weaker and less relevant than those at the national level.  All three parties made clear their expectation that candidates would fund their own campaigns with little or no direct financial support from their parties.  At least one party expressed to IRI pride in the fact that its candidates would be entirely self-supporting.  However, the absence of party support for local campaigns removed a major incentive for candidates to take direction or advice from party officials.  In some cases, the parties recruited and nominated candidates willing to use their own resources to campaign; these financially independent aspirants preferred to design their own campaign strategies without consulting their party.  In other cases, less-wealthy candidates appealed to clan leaders to endorse and fund their campaigns; in these instances, the clans took over roles that would normally be played by parties: organizing events, coordinating get-out-the-vote efforts, etc.

Local campaigns paid scant attention to issues and platforms.  Candidates focused their campaigns on mobilizing voters from friendly clans, mostly by engaging in traditional social forums such as afternoon qat-chewing sessions with elders and distributing small amounts of cash and other commodities to voters.  During the campaign season, qat merchants in Hargeisa reported that their business was booming as candidates bought enormous quantities to distribute to voters.  As in the nomination process, the considerable power of the clans in local campaigns proved a major disadvantage for women candidates.  Unable to engage in important male-dominated networking channels like qat-chewing sessions, women candidates were unable to secure the essential clan backing they would need to win.

Media Environment

Somaliland has a vibrant media sector that enjoys a reasonable degree of freedom, despite government attempts to apply restrictions.  In 2004, Somaliland’s government tried and failed to secure the passage of a repressive media law based on that of neighboring Ethiopia, and in recent years there have been several cases in which security forces have detained journalists in response to unfavorable news coverage. Nevertheless, Somaliland’s independent news organs have not been deterred from offering frank critiques of ongoing political events, and they provided useful news and commentary throughout the campaign period.

Somaliland’s print media includes both independent newspapers, such as the English language Somaliland Times and the Somali-language Jamhuuriya, as well as state-owned publications such as Mandeeq.  Newspapers are printed in Somali, English and Arabic.  The print media has a limited circulation, given difficulties of distribution to remote areas, so the readership of newspapers and other publications is largely limited to major urban areas.

There is only one radio station in Somaliland, the state-owned Radio Hargeisa.  During the campaign season, Radio Hargeisa was commended for offering airtime to all three parties as stipulated by the Political Parties Code of Conduct, but independent monitors from the Institute for Cooperative Development (ICD) noted that on many programs ruling party representatives were given much longer blocs of time to present their arguments.  Although it is technically legal to establish independent radio stations, the Ministry of Information has refused to issue any permits, arguing that a free environment for radio stations could lead to the broadcast of clan propaganda and hate speech.  Many Somalilanders get their news from the BBC Somali Service, which reaches the entire country.  Media monitors reported that the BBC coverage of the election was evenhanded, though since the station covers Somali affairs throughout the Horn of Africa, only a small amount of airtime each day was devoted to the campaigns in Somaliland.

Somaliland has two television stations. Somaliland Television (SLTV) is an independent station that has broadcast news and entertainment programs for several years.  Earlier this year, the government established a state-run channel, Somaliland National Television (SLNTV).  Both have a limited number of viewers given that very few Somalilanders have access to television sets.

Though all media organs submitted themselves to both the Political Party and Media Owners’ Codes of Conduct, the coverage offered by most media houses was judged to be less than impartial.  Government-owned media organs, especially SLNTV, were strongly criticized for broadcasting government propaganda rather than providing even-handed coverage.  On the other hand, independent media organs such as SLTV and the Haatuf Media Group were also criticized for tilting their coverage toward the opposition parties.  Depending on one’s perspective, it can be said that all points of view were represented in media coverage of the election, but seldom within the same media outlets.

Other Challenges 

A number of events put the strength of the code of conduct and its enforcement mechanisms to the test during the campaign season.  Though its enforcement powers and legal status were vague, the EBM was active during the campaign in “naming and shaming” violators of the Electoral Law and the Political Parties’ Code of Conduct.  For instance, the EBM rebuked SLNTV, the state-owned broadcaster, for devoting nearly all its coverage to ruling party campaign events, despite the code of conduct’s mandate of equal access to the media by all parties.  EBM criticism of the use of state-owned vehicles for campaigning by cabinet ministers resulted in a significant reduction in this abuse, which had marred past elections.  The APD reports that for campaigning in Burao, ministers were compelled to rent cars.  The EBM also catalogued and publicized the frequent instances in which party campaigners violated the code of conduct’s prohibition of inflammatory statements and personal attacks.

During the final weeks and months before the election, Somaliland also found itself forced to adjust to several unexpected events:

    • A territorial dispute erupted among three regions (Hargeisa, Awdal, and Sahil) over the status of six polling stations in an area north of Hargeisa. Leaders from each of the three regions claimed that this territory belonged to them, so votes from the disputed polling stations should go to their regional pool.  The Ministry of the Interior suggested that four polling stations should go to Awdal (not surprisingly, the home region of the minister), one to Hargeisa, and one should be shared among the three regions.  However, NEC saw this solution as unworkable and ultimately closed all six polling stations.
    • About a week before Election Day, police seized 150 forged ballot papers at the Hargeisa airport. The ballots had all been marked for a particular UCID candidate in Awdal region.  The candidate was detained but quickly released at the urging of NEC members, who pointed out the lack of evidence directly linking the candidate to the forged ballots.  UCID leaders argued that the forgeries were simply an attempt to distribute more sample ballot copies for civic education purposes, since the government had not provided enough.  However, NEC officials noted that the forgers had carefully removed the word “SAMPLE” that was emblazoned across legitimate sample ballots.  In the end, the candidate in question won a parliamentary seat and has not faced further government action.
    • In the early morning hours of September 23, Somaliland security forces, acting on information from an undercover agent, raided a house on the outskirts of Hargeisa. A battle ensued between police and a group of gunmen, some of whom were alleged to have ties to international terrorist groups.  Several of the suspects were killed in the gunfight and four others immediately taken into custody.  In the following days, several other suspects were arrested while trying to flee Somaliland.  In the house, police discovered a large stockpile of weapons, including machine guns and anti-tank mines.  After interrogating the suspects, the government concluded that they had been planning attacks linked to the ongoing trial of the killers of four foreign aid workers in late 2003 and early 2004.

 IV.  Election Day

On September 29, 2005, a seven-member IRI delegation observed voting in four of Somaliland’s six provinces.  In total, the delegation visited 94 polling stations, nearly 10 percent of the total polling stations in Somaliland.

Opening of Polls

According to the NEC training manual for election officials, all polling staff were to arrive at the polling station by 5:00 AM, by which time all necessary materials should have been received by the station chairman.  Over the next hour, the chairman was to instruct all staff on their duties, including how to apply the clear, indelible ink to voters’ fingers, and how to use the ultraviolet lamp to check for inked fingers.  The chairman was then to ensure that the ballot box was empty, seal it in the presence of all election workers, and record the serial number of the seal in his register.

Not all staff arrived at 5:00 AM, which would have given them a full hour to prepare for opening.  Only four out of the seven opening stations observed had a full complement of staff by 6:00 AM, when polls nationwide were scheduled to open.  Of the seven polling stations where IRI observed opening procedures, only two opened at the appointed time.  Among the 94 polling stations IRI observed during Election Day, 72 presiding officers reported that their stations had opened on time; it is likely that the true number is considerably lower.  Presiding officers in all 94 polling stations indicated that they had received all the necessary election materials in a timely manner, so this cannot be blamed for the late start.

In many stations, there seemed to be some confusion regarding administrative procedures.  About half of the presiding officers appeared confident in their roles, knew how to instruct their colleagues and ably walked the other officials through the opening processes. At some polling stations, chairmen failed to record important information such as the number of ballot papers received and the serial numbers of the ballot box seals.

In order to facilitate the transportation of materials, the NEC provided all polling stations with sealable canvas bags from Ethiopia rather than ballot boxes.  However, polling stations were not provided with any means of supporting these bags.  While in all stations IRI observed the polling chairman confirmed that the ballot bag was empty at the beginning of the day, insufficient provisions were made for the stabilization of the bag. Trying to find a way in which to hold the bag up was a cause for much improvisation; some chairman attempted to tie the bag to the ceiling, while others tried to suspend it between two chairs, etc.  After polls opened, many voters visibly struggled with the bag and especially with the opening.  A large number had to be assisted by polling officials or party agents.

In many polling stations, including station number 589 in Togdheer where electricity was only provided in the evening or not at all, opening procedures and early voting took place in near-total darkness, making it difficult for voters to read the ballot.

In all seven opening stations, party agents and other observers were present to watch opening procedures.  The agents conducted their duties professionally and refrained from interfering in the process.

At all polling stations they visited at the beginning of the day, IRI observers noted long queues of enthusiastic men and women, many of whom were reported to have camped out overnight.  Roughly 250 people were in line before the opening of the polls at Dila A polling station in Baki district.  At some stations, security personnel were barely sufficient to control unruly crowds when polls finally opened, and some scuffles were observed.  At one station in Burao, IRI observed that women voters had separated themselves into lines according to party affiliation and were wearing headbands with their parties’ logos.


After observing voting procedures across Somaliland, IRI observers were satisfied that, generally speaking, Somaliland’s NEC did a commendable job in organizing and executing Somaliland’s parliamentary elections.  Most polling station officials and party agents carried out their duties diligently and with great dedication under extremely challenging circumstances.  IRI observers noted a number of problems both minor and more serious, but in nearly all such cases these problems could be attributed to lack of experience, insufficient training, scarce resources, or exhaustion.  IRI observers saw no outright evidence that election officials were complicit in ballot fraud or intimidation of voters.  In addition, all difficulties should be viewed in light of the fact that this was the first direct parliamentary elections to be held anywhere in the former Republic of Somalia in more than 35 years.

Voting procedures

Members of the polling station staff included a chairman, a secretary, several ballot scrutinizers, and several security officers to maintain order outside the polling station.  In addition, each political party was allowed to send up to two accredited poll watchers to observe voting procedures in every station.  Women were represented on polling station staffs in many districts.

According to Somaliland’s Election Law and the NEC training manual, the voting process was to take place as follows: voters form two queues outside the polling station (one each for men and women), assisted by security officers.  As each voter enters the polling station, a polling official checks his or her fingers with an ultraviolet lamp to ensure that he or she has not yet voted, and then applies indelible ink to his or her pinky finger.  The voter then proceeds to the registration table, where his or her name is entered into the voter registration book by the secretary.  Afterward, the chairman stamps a ballot paper, hands it to the voter and explains how to vote.  The voter then enters the booth, marks the ballot with a red pen, folds it, exits the booth, and places it in the ballot bag before leaving the station.

Voter Registration

Somaliland does not have a national voter register, and a proper census has never been performed.  Furthermore, citizens of Somaliland have no standard form of identification.  In many areas of the country the people are largely nomadic, so it is difficult to document their place of residence.  Therefore, voters coming to cast their ballots on Election Day were not expected to prove their identities; it was sufficient for them to claim to be a citizen of Somaliland above the age of 16.  However, each voter’s name was entered in a registration book.  After the books are compiled, the NEC intends for this to form the basis of a national voter register that they hope to complete in time for the next election.

 Access to polling stations and voting facilities 

In about 90 percent of the stations that IRI observers visited, voters had no problems accessing the polling station and voting facilities. Most stations were established in easyto-find buildings (such as schools or public offices) at the center of villages and urban neighborhoods.  Most stations had ample space for the voting exercise to take place.  However, in some rural areas where suitable buildings were scarce, and at some Hargeisa polling stations, voting was held in cramped rooms with little space to maneuver and little possibility for voting in secret.  On the other hand, in some areas where there were no suitable buildings available, such as in Borama municipality of Awdal region, the NEC organized tents that were well ventilated and spacious.

Campaign activities and materials near polling stations

IRI observers noted only a few isolated cases where campaign materials were visible in or near polling stations.  In Togdheer’s station number 635, for instance, a polling official was wearing an UDUB T-shirt.  Also, in many areas party flags had not been removed from buildings surrounding the polling stations.  However, in general, the Electoral Law was sufficiently followed in this regard.  Where they were visible, the placement of party logos and emblems appeared to be inadvertent and IRI observers did not evaluate this as a deliberate move to influence or intimidate voters. The polling stations were generally free of outright canvassing or campaign materials.

Atmosphere at the polling station 

Eighty of the 94 polling stations IRI visited were described as either “generally orderly and calm” or “somewhat disorganized but calm.”  Voters generally queued in a somewhat organized manner; polling officials maintained order within the station, and the number of security officers was sufficient to take care of any instability that occurred.  However, an additional nine stations in three of the regions that IRI visited were described by observers as “chaotic,” and three were even described as “unstable and dangerous.”  The calmest region observed by IRI was Sahil, where no stations were described by IRI observers as chaotic or unstable.

For the most part, the chaotic or unstable stations were ones that IRI visited later in the day, when voters were becoming anxious to vote before the station closed.  In some places, such as station number 605 in Togdheer, queues had nearly disintegrated; voters crowded around the entrance, pushing, shoving, scaling walls and banging on windows and doors.  In station number 601 in Togdheer, voters managed to break open a second door.  It was apparent in many places that insufficient security personnel had been provided to keep the situation under control.  In certain stations, such as station number 257 in Hargeisa, security officers used sticks and whips to keep voters away from the entrance.   

IRI observers witnessed isolated cases of violence on Election Day. For instance, at stations number 85, 102 and 113 in Awdal, an IRI observer witnessed eruptions of fistfighting and stone-throwing among voters.  In station number 102 in Awdal and station number 267 in Hargeisa, voting had to be suspended temporarily in order for the situation to be brought under control.  In number 267, officials locked themselves inside the station as they waited for additional police to arrive.  However, in most places where violence was observed, it was quickly quelled by the security officers deployed at every polling station. No fatalities or serious injuries were reported.

IRI observed various instances in which security personnel at polling stations interfered inappropriately in the voting process.  Many did not adhere to the electoral law by staying outside the polling stations, although IRI saw no obvious instances of intimidation by these officers. In some instances, security officers exceeded reasonable limits and used excessive force to maintain order at the polling stations.  In other cases, they were not able to prevent the dissolution of queues or queue-jumping by people (mostly young men) attempting to vote multiple times, whether through neglect or insufficient personnel.

Multiple voting 

Attempts at multiple voting were rampant across all the regions where IRI observed.  As the day progressed, almost every station IRI visited experienced increasing numbers of voters attempting to enter polling stations a second time.  However, in the majority of polling stations these people were caught and expelled after a polling official used the ultra-violet light to see voters’ inked fingers.  Regardless of these widespread attempts at multiple voting, IRI saw no evidence that large numbers of repeat voters were being allowed into polling stations, with possible exceptions in Baki district.

However, in more than 10 percent of the stations that IRI visited, observers noted that election officials were not consistently checking voters’ fingers for ink, so it is likely that some voters were allowed to vote more than once.  In many places, especially in Awdal  and Sahil regions, 20-30 percent of the officials observed did not check for ink stains on people showing up to vote. This may have aggravated the already serious problem of multiple voting.

The attempts by would-be multiple voters led to a carnival-like atmosphere, in which voters would run out of the station grinning and giggling after they had been caught.  It appeared to IRI observers that most repeat voting was not attributable to organized attempts by parties or politicians at ballot fraud.  Rather, it is IRI’s assessment that young (perhaps first-time) voters were seeking to prolong the excitement of the voting exercise by doing it more than once and testing the new colorless ink system for flaws.  Some repeat voters even appeared to be curious to see what their inked fingers would look like under the ultra-violet light.  IRI saw no evidence that these voters were being paid by any individual or group.

The reaction of election officials at different polling stations to this phenomenon varied widely.  In some stations, the multiple voters were calmly refused entry.  In others, they were angrily chased away, yelled at or even handled roughly by security officers.  IRI observers reported that at certain stations in Sahil, security officers detained multiple voters and forced them to sit in one area for several hours.  Observers from another delegation told IRI that at some polling stations, multiple voters had their shoes and belts confiscated before they were detained.  This seemed to be an effective deterrent to other would-be multiple voters.

The large numbers of people attempting to vote multiple times had a clear and adverse impact on order at stations observed late in the day.  IRI observers believe that these voters contributed to the long queues still in place at 6:00 PM, and therefore probably resulted in large numbers of voter disenfranchisement when many stations did not allow those queued by 6:00 PM to vote.

IRI observers also witnessed attempts at underage voting.  Whereas the legal voting age is 16 years and above, there were numerous cases where visibly underage people attempted to vote, and in some isolated cases they were allowed to do so. In Hargeisa’s station number 256, for example, a 15-year-old admitted to having been allowed to vote.

Secrecy in voting 

There was a commendable effort by election officials to set up polling stations in such a way that voters enjoyed secrecy. As much as possible, voters were left to mark their ballots peacefully without any interference except in cases where they needed assistance due to illiteracy or disability.  Despite the fact that article 37 of Somaliland’s Election Law stipulates that every polling station shall have two voting booths, in many cases only one was provided. For the most part, this was because some stations were too small to accommodate two booths.  Many of these stations were so small that it was difficult for voters to maneuver their way past the officials, party agents and waiting voters to get to the voting booth, so many voters elected not to vote in secret.

There also seemed to be a deficiency in the training of election officials on the construction and use of voting booths; there was no standard type of booth and in some cases they were just curtains taped to walls.  In addition, in many stations the booth was little used and the chairman did not direct voters to use it.

Assistance to illiterate and disabled voters

People with disabilities received appropriate assistance in all cases that IRI observed, and observers received no reports of any person who was disenfranchised due to disability.  In station number 605 in Togdheer, IRI observers saw an elderly blind woman being respectfully led into the polling station through a back door by election officials, allowing her to bypass a long and unruly queue.

In many parts of the country, especially rural areas, it was apparent that illiterate voters made up a large proportion of the electorate.  Although the system of assigning a pictorial symbol to each candidate was intended to allow illiterate voters to vote without assistance, in a large number of cases, voters were not familiar with their candidates’ symbols, while other voters complained that the symbols on the ballot were either too small or unrecognizable.  Whenever illiterate voters were not able to complete their ballots, the station chairman offered assistance in a manner that was acceptably transparent.  The voter would tell him which candidate she supported, he would tick the appropriate box, and before allowing the voter to cast the ballot he would display it to the party agents so they could see that he had followed the voter’s direction.

In some rural areas, IRI observers noticed that candidates had distributed slips of paper to their supporters containing their campaign symbol; the voter would show the paper to the chairman and he would immediately know which candidate they supported.

Transportation of voters

Article 27 of the Election Law states that vehicle movement on Election Day is prohibited without written permission by the NEC.  This vehicle ban was designed to ensure a high level of security, to discourage multiple voting, and to remove an unfair advantage from candidates and parties who can afford to transport their voters to the polls.  However, in all four regions IRI observers visited, this provision was blatantly disregarded.  On Election Day, politicians, clan leaders and other individuals transported thousands of voters to polling stations.  IRI observers saw scores of trucks, pickup trucks, buses and small personal cars, many emblazoned with campaign posters and party logos, carrying voters to and from polling stations.

IRI has no evidence to indicate whether these vehicles were facilitating multiple voting or simply carrying voters to polling stations from remote areas.  However, observers from the institute did note that at some stations the number of multiple voters being turned away increased significantly after the arrival of these vehicles.  Further, at station number 605 in Burao, an IRI observer photographed a minibus transporting voters between stations that bore a vehicle permit (number 194) issued by the NEC to KULMIYE.

Party Agents

In all stations that IRI visited, all three political parties were represented by accredited poll-watchers; in most stations, each party had two agents present.  IRI observed that the vast majority of these agents appeared to have been well trained and were conducting their responsibilities professionally without interfering in the voting process.  In some smaller polling stations, the presence of a full complement of six party agents caused the station to become overcrowded, making it difficult for voters to maneuver through the different steps of the balloting process.  

In crowded polling stations, party agents often assisted in small tasks such as showing voters how to insert their ballots into the cumbersome ballot bags.  Though such involvement by party agents is technically disallowed, IRI was satisfied that these party agents were trying to be helpful and did not intend to manipulate or intimidate voters.  However, IRI noted a few cases in which a party agent was assigned inappropriate duties; in Togdheer’s station number 606, a KULMIYE party agent was tasked to check voters’ fingers for ink with the ultra-violet lamp; similarly, in Baki district, a party agent was assisting illiterate voters by marking their ballots for them, a job the Electoral Law designates solely for polling station chairmen.

Other administrative issues

Despite some confusion about procedures, in general, polling officials appeared welltrained and confident in performing their tasks. They handled the voting process with diligence and patience.  They treated voters fairly and respectfully, save for a few cases where men were given preference in voting over women.  At many polling stations, facilities were not sufficient to handle efficiently the large number of voters, and Somalilanders found themselves waiting for long hours, sometimes under a punishing sun or pouring rain, before voting. In Togdheer’s station number 585, for example, a woman reported to IRI that she had joined the queue at 2:00 AM and had not voted by the time observers visited the station at 9:30 AM.

IRI observed that, in general, the assignment of a specific duty (e.g. stamping and issuing ballots) constrained polling chairmen to one spot in the station and made it nearly impossible for him to oversee the performance of other duties.  For instance, in many polling stations the process of checking voters for inked fingers was done outside the chairman’s line of sight, so he would not have been able to tell whether or not this was being done correctly.

In many stations, IRI noted a lack of seriousness in the handling of spoiled ballots.  After voters were issued replacements for ballots that had been damaged or otherwise spoiled, the original ballots were often thrown into piles under tables or in corners of the room, out of the sight of polling officials, rather than in a special envelope as required.

In some areas of the country, particularly Awdal region, a number of polling stations ran out of ballots and had to suspend voting well before closing time with people still waiting in queues. In stations number 138 and 152 in Awdal, ballots ran out at 4:15 PM and 2:30 PM respectively. Many voters had to relocate to other stations, some of which were several kilometers apart.  In some cases, ballots were replenished before closing time and in others, the stations simply closed early for lack of materials.  It is possible that some voters could not make it to other stations in time to vote or were discouraged by the inconvenience of having to move to another station; therefore, some of these voters may have been unnecessarily disenfranchised.

Perhaps owing to the excitement of finally voting for their representatives in parliament, people loitered around many polling stations after they had finished voting. This caused congestion around many stations, and, in some cases negatively impacted the voting process. In many stations in Baki district and at Togdheer’s station number 590, the large crowd made it difficult for people, particularly the elderly, to access the station.

Closing and vote tabulation

Closing time

Of the seven stations at which IRI observed closing procedures, five closed at 6:00 PM, the official closing time. Polling officials at some stations IRI visited during the day informed observers that they planned to close late to make up for opening later than 6:00 AM.  Meanwhile, stations number 332 in Hargeisa and 502 in Sahil stayed open so that voters in the queue by 6:00 PM had the chance to cast their ballots.

IRI also noted that four of the seven stations did not follow the legal requirement that all voters in line by 6:00 PM should have the chance to vote; these stations closed promptly at 6:00 and turned away all remaining voters. At some stations, it was difficult for polling officials to ensure that only voters who had arrived at the polling station before 6:00 PM would be allowed to vote, since many people attempted to join the line after that point.

It was also noted that many late-day voters had already voted at other stations and were seeking to engage in multiple voting.  IRI observed that although most such voters were caught and expelled, their presence in the queue wasted a great deal of time and led to the continued presence of long queues even after closing time.  At stations that did not follow the requirement to stay open until all voters in the queue had voted, this meant that a large number of people were not able to exercise their right to vote.

Vote tabulation, recording and collation

In most closing stations, polling staff made sincere efforts to carry out all necessary procedures.  For instance, in all stations observed, the number of registered voters was cross-checked with the total number of used, unused and spoiled ballot papers.

However, IRI observers noted considerable confusion in tabulation, recording and collation procedures.  Many polling officials, including station chairmen, did not appear sufficiently familiar with counting regulations and had to continually refer to the election manual.  In addition, after a long, frenetic day of voting, polling staff were so exhausted that in some cases they rushed through procedures or performed them half-heartedly.  At Dila A station in Baki district, polling staff took two hour-long breaks after closing time to rest, drink tea and chew qat before beginning to count ballots.

Initially, counting was very slow. Some station chairmen were not clear on how to separate ballot papers while counting and whether to count first by party or by candidate. IRI observers also noted that, in order to speed the process of counting, tired polling officials often enlisted the party agents (who were supposed to be uninvolved observers) in counting and checking the validity of votes. At station number 589 in Burao, as the chairman called out the votes, his secretary recorded the number of votes for each party while the party agents were asked to keep track of the votes received by each candidate in their respective parties.  Both counts were later recorded as the official results.

IRI observed disputes over the validity of votes, with officials and observers contesting the acceptability of ballots where marks crossed a line, where the mark was tiny, where two candidates were marked, and where voters wrote the candidate’s name without making any other mark.  In Awdal region, the color of the ballot (pink) made it difficult for scrutinizers to find the mark, since in all stations a red ballpoint pen was used by voters.

In recording vote totals, IRI observers noted that to produce all the necessary copies of the results form, the carbon paper provided was small and difficult to use.  Polling officials, therefore, often needed to complete each form manually, frequently making mistakes which had to be cancelled and rewritten.

In all stations observed, sensitive materials were properly sealed with the assistance of a NEC official and the results were escorted by police officers and party agents to the NEC district office. All IRI delegates summarized the closing and counting processes in their stations as “disorganized but generally calm and correct,” despite the great amount of time that counting processes consumed.  In some stations, closing and tabulation procedures continued until the early morning hours of the next day.  At station number 502 in Sahil, an IRI observer did not depart until 10:00 AM the following day.

At the end of the exercise, there were no unresolved disputes and the party agents in all stations signed the final report.

V.  Post-Election Period and Results

The election results announced by Somaliland’s NEC on October 15 are as follows:

Table 1: Voting results of the 2005 parliamentary election 

Regions  UCID Kulmiye  UDUB TOTAL Not valid  Total
Awdal  31,492








757 133,777
Hargeisa  75,796






253,229 2,750 255,979
Sahil  18,331






52,479 216 52,695
Sanaag  17,907






89,286 537 89,823
Sool  2,436






20,557 199 20,756
Togdheer  34,583






121,751 126 121,877
Total 180,545






670,322 4,585 674,907

As a result of these totals, the seats in the new parliament have been divided among the parties as follows:

Table 2: Distribution of seats in the new parliament  

Region  Parliamentary Members  UDUB UCID KULMIYE
Awdal  13 7 3 3
Hargeisa  20 6 6 8
Sahil  10 4 4 2
Sanaag  12 5 2 5
Sool  12 6 2 4
Togdheer  15 5 4 6
Total  82 33 21 28


Comparing these results to those of the 2003 presidential election, a number of trends are clearly discernible (See Appendices A and B for the results of Somaliland’s previous two elections).

It should be noted that the tendency of voters in many parts of Somaliland to support candidates from their clans has not changed.  In 2003, since there were only three candidates from which to choose, each party’s geographic performance correlated strongly with the regional/clan bases of its candidate and other top personalities.  However, in 2005, voters were faced with the ability to choose among several candidates from their clan representing different parties.  Therefore, the phenomenon of particular parties dominating certain regions was much decreased.  For instance, in the 2003 election four out of six regions were won with more than 55 percent of the vote.  In 2005, only one region, Awdal, gave any one party more than 45 percent of its vote.

The biggest story in this election was the dramatic increase in support for UCID, Somaliland’s second largest opposition party.  In the 2002 municipal elections, UCID barely edged out several other competitors to become one of Somaliland’s three official political parties.  In 2003, the party polled less than 12 percent in every region except Hargeisa, and managed only 15.9 percent overall.  However, in 2005 UCID increased its support in all six regions by between 6.9 and 26 percentage points.  In four regions, the party more than doubled its support base (in Sahil support for UCID nearly quadrupled), and nationally it gained 11 points to take 26.9 percent of all votes and more than onefourth of the parliamentary seats.

UCID’s stunning success can be attributed mainly to two factors.  The first and most important factor was the party’s strategy to position itself as a moderate alternative.  UDUB and KULMIYE spent much of the campaign attacking each other, only to have these negative tactics backfire; many voters were turned off by both parties and were highly receptive to UCID’s comparatively positive message of peace, development and social justice.

The second factor was the caliber and personal resources of the party’s candidates.  A large proportion of UCID’s candidates were returned émigrés who had achieved academic and financial success in western countries.  UCID heavily promoted these candidates as highly-skilled and qualified professionals, and many UCID candidates were fully prepared to outspend their less-wealthy rivals.  Furthermore, UCID capitalized on the failure of a number of popular and well-qualified candidates to gain nomination by UDUB and KULMIYE for reasons such as clan balance.  Many of these individuals (including the owner of the international-class Ambassador Hotel in Hargeisa) found a home in UCID and were able to easily beat candidates from the parties that had rejected them.  Due to these advantages, UCID was able to grow in the span of two years from a marginal political group to a formidable player in Somaliland’s nascent democracy.

UDUB, while failing to maintain its 2003 level of support, managed to exceed the expectations of many observers.  Despite expectations that the unpopularity of many recent government policies and public discontent about corruption would deal a crippling blow to support for UDUB, the party gained a plurality of seats in parliament.  Between 2003 and 2005, the party’s national support decreased by only 3.1 points, and UDUB retained a commanding lead in President Rayale’s populous Awdal region.  UDUB’s strong support was due, in part, to the party’s unfair advantages in using state-owned media as a vehicle for party propaganda and state resources to support its candidates (although to a significantly lesser extent than in past elections).  Furthermore, campaignyear government policies such as the appointment of ministers and commissions and the initiation of major public works projects may have won new supporters.  There have also been allegations that UDUB engaged in serious irregularities to increase its vote count, especially in Awdal.  However, it is also possible that many observers overestimated the extent to which UDUB supporters would desert the party for KULMIYE.  Although its unrelenting attacks on KULMIYE probably turned off voters to both parties, it is apparent that many voters chose stability over change, and credited UDUB with overseeing three successful democratic elections while maintaining peace.

The only party that performed below expectations was KULMIYE.  The support base of the largest opposition party shrank in all six regions, and in three of those regions it shrank by more than 10 points.  Nationwide, KULMIYE lost eight percentage points.  This result confounded the expectations of many Somaliland-watchers, who believed that the poor performance of the Rayale government would lead voters to desert the ruling party.  This was true to a certain extent, but it appears that most of these disaffected voters bypassed KULMIYE and supported UCID, along with a great number of discontented KULMIYE members.

Though KULMIYE’s setback was in part the result of unfair government tactics such as raiding the opposition party’s headquarters and flooding the public airwaves with anti KULMIYE attacks, part of the responsibility also lies with KULMIYE’s lackluster and overconfident campaign.  The party based its national appeal on the charisma of its leader and the fact that many of its most prominent figures were SNM veterans and heroes of the independence struggle.  However, UDUB used KULMIYE’s SNM ties to remind voters of Somaliland’s period of instability and civil war under SNM leadership.  Also, the appeal of KULMIYE’s national figures did not necessarily translate to its candidates at the regional level, many of whom had mediocre credentials.  Additionally, KULMIYE’s support for the ill-fated attempt to impeach Rayale (which UCID refused to support) may have raised concerns among voters about KULMIYE’s willingness to risk insecurity in order to achieve its political aims.

Normal measures of turnout cannot be calculated in the absence of voter registration. However, in terms of absolute numbers, turnout increased considerably between the 2003 and 2005 elections.  In this election, 670,322 valid votes were counted, compared to 488,543 in 2003, an increase of 37 percent.  Turnout rose most dramatically in Sool and Awdal, which saw increases of 112 percent and 102 percent respectively.  Turnout was most stable in Togdheer, which saw only a six percent increase.

Post-Election Delays

Although Somaliland’s Election Law requires that the NEC central office determine the winners of the election within 10 days of the election, the NEC did not do so until October 15, 16 days after Election Day, when it forwarded the preliminary results to the Constitutional Court.  The court’s ratification of the results took an additional two weeks.  On November 1, 2005, the Vice Chairman of the Constitutional Court of the Republic of Somaliland, Mahmud Hirsi Farah, announced that the court had ratified the results of September 29 parliamentary elections.

Given the highly politicized environment and the scarcity of information emerging from the NEC, many politicians, interested observers, and stakeholders voiced fears of behind the-scenes attempts to fix the results.  Within days of the election, political party leaders began to publicize reports of fraud. Somaliland-focused websites (the main source of news about Somaliland among the Diaspora) were also rife with conspiracy theories.  Most of these involved attempts by the party in power (including particular government ministers) to manipulate the results in remote areas in order to temper UDUB’s poor performance in some major towns.

However, most nonpartisan observers believe that the prolonged post-election period was due to technical problems, including difficulties with transportation, communication, and administrative procedures.  The first problem to emerge in the post-election period arose from the way in which voting totals were reported to the NEC by polling station chairmen.  In observing counting procedures on Election Day, IRI observers noted that producing all the necessary copies of the tally sheet was difficult because the carbon paper provided was difficult to use, and polling officials often needed to cancel their mistakes and rewrite figures, which made carbon copies messy and hard to read.

Several days after the election, a consultant from NEC reported to IRI that a very large number of the tally sheet copies received by election offices at all levels across the country were faint and nearly illegible.  The consultant conjectured that the polling station chairmen had given the top three (i.e. clearest) carbon copies to party agents at the polling stations, and had sent the progressively faint copies to election officers.  Due to this simple clerical oversight, NEC officials had to re-open many locked and warehoused boxes from the polling stations to retrieve the original tally sheets, a task that took many days to complete.

Other technical difficulties in the post-election period included the following:

  • Due to rain and bad roads, the central NEC office did not receive materials from some remote areas of the country until nearly a week after Election Day.
  • The NEC conscientiously sought to resolve disputes among political parties about voting results in specific districts and polling stations by conducting numerous time-consuming recounts.
  • The Islamic holy month of Ramadan began shortly after Election Day, significantly limiting the number of hours each day that election officials could conduct their work.
  • Even though the urban center of Hargeisa contains more voters than most regions, it is still only considered a single district. Since the election law limits the number of “core staff” in each district election office to six, the district election office in Hargeisa, which had to process results from roughly 230 polling stations, was severely understaffed.  As a result, Hargeisa was the last region to complete vote tallying, well past the 10-day timeframe stipulated in the election law.

Allegations of Irregularity

In the days following the election, IRI delegation members heard serious allegations that the ruling party had abused its power on Election Day and in the early post-election period.  Many of these allegations were later repeated in Somaliland’s media and on Diaspora websites.  After meeting with numerous stakeholders in Somaliland, IRI saw no conclusive evidence to support most of these allegations.

However, one particular complaint raised informally by the opposition parties, involving efforts to fix results in remote districts of Awdal Region seemed, credible enough to merit investigation.  In particular districts, including Zeila, Lughaya, and Baki, the number of voters was unrealistically high, especially when measured against corresponding numbers from the 2003 presidential election.

Though district-by-district results had not yet been publicly released at the time this report was written, NEC members confirmed to IRI that the Awdal numbers were troubling.  For instance, the reported vote total in sparsely-populated Lughaya district was more than six times higher than that of the 2003 election (25,138 vs. 4,182).  In Zeila and Baki, the vote totals were roughly three times higher than in 2003.  An NEC staff member told IRI that “There was evidently a determined effort to inflate votes in Awdal… It’s likely there was collusion among the party agents and the election officials.”

NEC officials pointed out that whereas in most parts of Somaliland a system of “staff swapping” was implemented to ensure that key polling station staff (namely presiding officers and ink-checkers) did not work in their own home districts, it was not implemented in some districts of Awdal.  The chief safeguard against fraud at the polling station level was the requirement that accredited agents of all three parties register their approval by signing off on the station register before results could be reported to the District Election Office.  However, the opposition parties insist that in certain parts of Awdal, their party agents were detained or kept from their duties and replaced by agents loyal to the government.  NEC officials did not deny to IRI that this had happened, but stated that the opposition parties had not reported this to the NEC in time for something to be done about it.

An investigation into voting results in Awdal region would have been well merited.  The register of voters for each polling station could have been checked for “ghosts.”  However, NEC did not conduct such an inquiry and these questions were not resolved for two reasons.  First, it would have required NEC officials to reopen the locked box of materials from each polling station; NEC does not possess the authority to do so unless complaints have been raised by party agents at the polling station, district election office, or regional election office level.  Since in this case the results were ratified at all levels, the NEC’s hands were tied.

Second, and more surprisingly, neither of the opposition parties raised a formal complaint to the NEC or to the Constitutional Court, as they were within their rights to do.  It is likely that Somaliland’s opposition parties reached the conclusion that, even if the alleged fraud in Awdal improved the showing of UDUB in this election, the overall results had put the opposition in a stronger legislative position than ever before.  This was especially true for UCID, which went from existence at the periphery of Somaliland politics to nearparity with KULMIYE in parliament.  It is therefore possible that the two opposition parties decided to secure the ground they had gained by accepting the results without contest.

Furthermore, a strong will exists among all three parties to complete this process peacefully and without destabilizing the country.  It was this same will that led KULMIYE in 2003 to accept the results of a presidential election that their chairman lost by less than 100 votes.  In an email to IRI just after the announcement of preliminary results on October 15, one opposition leader said that “we the opposition parties have endorsed the result of this election, disregarding the misbehavior of the government for the sake of our young nation.”  In an interview two months after the election, an UCID official told IRI that “People in Somaliland are hostages to peace.  We don’t cry foul because we’re afraid of creating conflict.”

Although the alleged irregularities may have distorted election results to a certain extent in Awdal region, IRI saw no evidence to indicate that national election results had been significantly affected.  Outside of Awdal, there have been no credible allegations of significant fraud.

Next Steps for Somaliland’s Democratization

Some opinion leaders in Somaliland now view Somaliland’s first direct parliamentary elections as the closing note to “Phase One” of the country’s democratization process. There is not yet a general consensus as to what “Phase Two” will entail, but some of the following objectives are now being widely discussed:

    • Developing expertise in legislative policy-making and agenda building among Somaliland’s new parliamentarians, all but a small handful of whom have no legislative experience;
    • Institutionalizing parliament as an effective check on the executive while maintaining political stability;
    • Democratizing and decentralizing political parties’ decision-making processes;
    • Maintaining party unity and discipline in parliament, especially since many MPs owe their election more to their clans than to their parties;
    • Conducting a nationwide census and voter registration exercise to facilitate the next round of elections;
    • Passing a permanent law to govern all types of elections in Somaliland (municipal, presidential, parliamentary and guurti);  
    • Implementing a continuous program of voter education;
    • Further engaging women, minorities, and other marginalized groups in politics despite their poor performance in this election;
    • Harmonizing Somaliland’s modern democratic institutions with traditional clanbased institutions;
    • Liberalizing restrictive laws and executive orders, many of which may be unconstitutional;
    • Removing remaining restrictions on the freedoms of expression, assembly and the media; and,
    • Seeking greater integration with the international community.

  VI. Findings and Recommendations

Upon completing its observation of the 2005 parliamentary elections, IRI respectfully offers the following recommendations to major stakeholders in Somaliland.  Within each target institution, recommendations are listed in order of importance.

National Electoral Commission:

    • Conduct nationwide voter registration: A national register of voters is an essential tool in preventing multiple voting and other fraud, as it allows for voters to be assigned to particular polling stations, where they can be checked off as they vote.  The list of voters collected by NEC on Election Day can form the basis of such a register, but NEC and other stakeholders should also reach out to government and international partners to fund a national project to register all eligible voters.  This task would be much facilitated if done concurrently with a national census and the development of standard identity documents.
    • Streamline nationwide tabulation procedures: By international standards and those of Somaliland’s Electoral Law, the period between Election Day and the announcement of results was excessively long.  IRI recommends that the NEC conduct a review of tabulation procedures to determine how they might be condensed to allow for results to be announced within the 10-day timeframe stipulated by the Electoral Law.
    • Strive for greater uniformity and consistency in the training of polling officials: The NEC is to be commended for selecting university students as polling officials.  They were energetic, enthusiastic, and approached their work with great seriousness and a sense of responsibility.  Developing a database of young, experienced officials will be an excellent contribution to the efficiency of the electoral process.  However, IRI recommends that the training of polling officials begin on an earlier date and be performed on a consistent basis with sufficient follow-up.  The institute also recommends that the NEC study aspects of polling that were not consistently executed by officials in different parts of the country.  In particular, IRI noted confusion among polling officials as to when and where to ink fingers, how and when to record numbers of ballots received, when to record ballot box seal serial numbers, and how to count ballots quickly and efficiently.
    • Reduce the burden and increase the mobility of polling station chairmen: During voting hours, station chairmen were over-burdened and unnecessarily tied to a particular task (stamping and distributing ballots).  As a result, these chairmen were unable to move around, properly manage the station, or observe when other officials were performing their tasks incorrectly.  IRI recommends that the NEC allow chairmen to better execute their duties by not assigning them a specific balloting task.
    • Use ballot boxes rather than canvas bags: The ballot bag used in this election was too small and difficult to prop up, and the slit was too narrow to accommodate the ballot unless it was folded many times.  The NEC should either return to using ballot boxes or provide each station with rope or metal structures to secure the bag in an upright position.
    • Ensure voting secrecy: Due to the setup of polling stations, many voters were unable to vote in private.  Polling officials in many stations were unable to construct a proper voting booth using the curtains provided, and very few stations set up more than one voting booth as required by law.  IRI submits, as one suggestion, that the NEC provide each polling station with a simple, foldable cardboard structure that can be assembled on top of a small table or desk, to ensure that each vote is cast with absolute privacy.

Government of Somaliland:

    • Conduct a nationwide census: Planning for these elections was complicated by the lack of census data. As a result, the NEC was unable to conduct nationwide voter registration, and had insufficient information on which to base the establishment of polling stations in all parts of the country. Many polling stations did not receive enough ballots and other supplies, adversely affecting the process and possibly disenfranchising many voters.
    • Simplify the electoral system: The system of election in this round was overly complicated, which had a negative effect on several stages of the process; an extremely cumbersome ballot made voter education difficult and caused confusion at polling stations; exhausted polling officials needed long hours to complete counting on Election Day; and, the necessity of two separate counts at every level led to an overly extended tallying period.
    • Establish clear borders between electoral regions: In this election, disagreements about the boundaries between several regions caused unnecessary rancor and required the NEC to cancel a significant number of polling stations, thereby reducing voters’ access to the polls.  The government, in consultation with clan elders and civil society leaders, should draw clear maps to establish district boundaries that would be approved by the parliament.
    • Allow the licensing of independent radio stations: Though the environment for print and televised media is now relatively free, the only licensed radio station in Somaliland is the state-owned broadcaster, Radio Hargeisa, and the Ministry of the Interior has consistently refused to license independent outlets that could provide more impartial and localized coverage. Since radio is currently the only medium with the potential to reach the majority of the population, access to information in remote areas remains severely limited. The ministry should license private media operators while restricting the broadcast of hate speech and clan propaganda.
    • Refrain from the use of state resources in party or candidate campaigns: Numerous complaints were aired by opposition political parties that the party in power was using state resources for campaigns, thereby giving an unfair advantage to its candidates. The government should ensure that the parties compete fairly and that no one party has exclusive access to state resources. Real or imagined cases of the employment of state resources in campaigns can damage the integrity of the electoral process.
    • Give the Election Monitoring Board powers to enforce the Electoral Law and Code of Conduct: Compliance with the Electoral Law and Code of Conduct was a serious issue in these elections. The newly-formed EMB performed admirably in underlining abuses by specific parties and candidates. However, it was only empowered to publicize and condemn such violations.  In future elections, IRI recommends that the independent character of the EMB be maintained, and that the board be given powers to levy serious fines and other punishments on violators.
    • Deter multiple voting through a regularized enforcement policy: IRI recommends that security officers be trained to employ penalties like those practiced at some polling stations.  In these stations, attempted multiple voters were rounded up, their shoes and belts were confiscated, and they were required to sit in one area for several hours.  This appeared to be an effective deterrent against multiple voting.
    • Provide more training and sensitization to security personnel at the polling stations: Policemen and members of the Special Protection Unit need to be properly informed about their role in the electoral process. Security officers should be imbued with a sufficient understanding of how to prevent abuses and maintain order without using excessive force.  Security officers should also be trained to ensure that women and men are given equal access to polling stations.
    • Deploy security personnel to each polling station on the basis of expected numbers of voters: IRI observed that in urban areas, where voters gathered in long queues, the number of security personnel deployed was insufficient, and they were eventually overwhelmed by unruly crowds.  In rural stations, where voters were sparse, order was much more easily maintained.  However, at all polling stations there were roughly the same number of officers.  IRI recommends that additional officers be deployed to polling stations in more densely-populated areas.

Political Parties:

    • Develop coherent policy alternatives to form the basis for party platforms: After conducting interviews with voters at a number of polling stations, IRI determined that the majority of voters could not easily tell the difference between party positions on important issues facing Somaliland. The electoral process hinges on presenting the electorate with choices. It is therefore important for parties to take a critical look at the gamut of issues affecting Somaliland and develop cogent proposals that can be presented to the people both as election-time party platforms and between-election legislative agendas.
    • Increase adherence to the law and to the Code of Conduct: During the campaign season, there were numerous cases of parties flouting the Electoral Law and the code of conduct with impunity, despite the fact that the code of conduct was adopted and signed by all three parties. Parties should thoroughly educate their candidates and officials on their roles and responsibilities, and those found to be in breach should be penalized by the parties themselves.
    • Provide more resources and technical support to regional campaign committees: Despite the effort that was put into the establishment and training of party campaign committees at the regional level, during the campaign these committees were neglected, under-funded and largely ignored by candidates.  In order to coordinate their campaign messages and strategies at the grassroots level, it is essential that the national party leaders give these officials the necessary support.
    • Develop fundraising strategies to support the campaigns of all party candidates, especially women and other marginalized groups with limited access to resources: The lack of direct support by political parties to their candidates in this election impaired the relevance of party institutions and platforms, since candidates derived most campaign funding from clans and their own personal finances.  This is likely to limit the parties’ ability to enforce unity among their members of parliament.  It also means that the success of candidacies for parliament has more to do with the candidates’ personal wealth and their connections within clans than with their qualifications to lead.
    • Proactively support women to run for elective positions at all levels of government: Women were instrumental to the success of the just-concluded election, and women’s civil society organizations provided crucial training and voter education. However, women’s clear enthusiasm for and involvement in the electoral process did not translate into the successful nomination and election of many female candidates.  The poor performance of women is a concern that parties can – and indeed should – take the lead in addressing. Greater participation of women as candidates should be nurtured by the parties through the nomination of greater numbers of women and through the provision of financial support for women candidates.
    • Partner with other players in the electoral process to reflect on lessons learned from the elections: Parties should draw lessons from electoral processes in other democracies that have gone through similar challenges. This can take the form of exchange/exposure visits or discussion and training forums.

Civil Society:

    • Produce practical training materials, manuals, and posters to further voter education: Civil society groups need to augment the work that the parties and the NEC have done to educate voters on issues relating to elections and good and responsible citizenship. Civic education should also be a continuous process targeting all sectors in society. This will ensure that people are aware of their rights as well as encourage constructive engagement with government.
    • Create linkages with civil society groups from other countries to learn from them: This would, in turn, strengthen efforts by the parties, government, and other players in making the electoral process fair, transparent, and democratic.
    • Coordinate election training activities among civil society, political parties, NEC, government, and parliament: This is important to draw on the comparative strengths of each of these groups in ensuring effectiveness and quality of training. In the September 2005 elections, NEC faced serious time constraints in training which are likely to have compromised the quality of the trainings.  
    • Become more deeply involved in post-election activities: Civil society needs to continuously engage the government and other players in the period following the election, especially in ensuring the accountability of parties, government, and the elected leaders.


    • Maintain strict impartiality in reporting campaign events: During this campaign season, there was a widespread perception that state-owned media were routinely presenting UDUB propaganda rather than even-handed reports, while independent media were overly biased toward the opposition parties.  As a result, Somalilanders were not able to turn to any media outlet for objective information.  IRI recommends that media organs strive to cover all relevant political events without bias to any party or individual.  IRI further recommends that print media clearly separate and label news articles and opinion pieces, which are now often printed side-by-side in the same format.
    • Seek training from international partners: Somaliland’s media, one of the most free in its region, can benefit greatly from exposure to journalists from other countries, specifically in the area of political reporting.


VII. Conclusion

The recent elections in Somaliland may be described as historic; Somalilanders’ first direct selection of their own representatives in parliament is a clear turning point in the self-declared republic’s road to democracy.  However, more remains to be done to ensure that Somaliland’s electoral systems and institutions can deliver to Somalilanders the freedom and security for which they have long yearned.

IRI recognizes and applauds the efforts by the government of Somaliland, the NEC, the political parties, and the people of Somaliland in making these elections a success. Regardless of all the difficulties that occurred, in the September 2005 parliamentary elections the people of Somaliland were able to cast their ballots in a peaceful manner and without significant intimidation. It was inspiring for IRI observers to see the enthusiastic faces of the citizens of Somaliland on Election Day, braving scorching sun and pouring rain to make sure that their votes were counted – for political stability, peace, and development.

In this report, IRI has identified certain areas that need critical attention to ensure that Somaliland learns from its mistakes and builds on its achievements. These include rampant multiple voting, as well as procedural and administrative problems that weakened the process. IRI hopes that the relevant authorities will address these concerns and continue to improve Somaliland’s institutions of electoral democracy.

Finally, Somaliland’s political leaders must be resolute in ensuring that the aspirations of Somalilanders are not betrayed but are channeled into goodwill for growth and stability. Political leaders must also continue to reach out to the international community for partnership and support. In order to encourage meaningful partnerships with the international community, Somaliland’s political culture must be one that continues to respect basic democratic ideals.

Download here for the full report

This site uses Akismet to reduce spam. Learn how your comment data is processed.