News
Report of The Brenthurst Foundation Observer Mission
The Brenthurst Foundation Observer Mission has published its report following the 2024 Somaliland Elections.
The 2024 Somaliland Presidential and Political Organizations Elections
Report of The Brenthurst Foundation Observer Mission
www.thebrenthurstfoundation.org
14 November 2024
History
Somaliland went to the polls on 13 November 2024 in the ninth competitive election since its re-declaration of independence from Somalia in 1991. Somaliland held a constitutional referendum in 1991 followed by municipal elections (2002, 2012 and 2021), presidential elections (2003, 2010 and 2017), and parliamentary elections (2005). Somaliland has been a stand-out democratic performer in a region characterised by political upheaval, not least in neighbouring Ethiopia and Somaliland’s one-time state partner, Somalia.
Somaliland’s roots lie in the agreements signed between the United Kingdom and clans in the area in the late 19th century which led to the establishment of the Somaliland Protectorate. On 26 June 1960, the protectorate was formally granted independence and became the State of Somaliland. Five days after gaining independence, it voluntarily united with the Trust Territory of Somalia, the former Italian colony to its east and south, thereby forming the Somali Republic.
The union was problematic from the start. When Somalia’s Siad Barre regime enacted harsh policies against Somaliland’s dominant clan family, the Isaaq, a ten-year war of independence ensued. The bitter conflict resulted in 90% of Somaliland’s capital city, Hargeisa, being destroyed, bombed by Siad Barre’s airforce and stripped bare by looters.
Somaliland defeated Mogadishu’s forces and declared independence in 1991, regarding itself as the legitimate successor state to British Somaliland.
In the absence of Mogadishu’s acquiescence, Somaliland’s independence is not internationally recognised. Yet Somaliland has created one of the most inclusive multiparty democratic systems in Africa. Not only was the 2024 event the fourth presidential election by universal suffrage in Somaliland, but previous elections have resulted in transitions of power between parties. In the May 2021 legislative and municipal elections the ruling Kulmiye Peace, Unity and Development Party of President Muuse Biixi Cabdi won fewer seats than the opposition alliance of the Waddani National Party and the Justice Welfare Party in a tightly fought contest.
Background to the 2024 Election
The 2024 Somaliland election was unusual in three respects: first, it was a combined election for both the presidency and political parties. Under the Somaliland three-party system, the largest three parties which additionally have to have at least 20% support across the country’s six regions, contest the political process. The three incumbent parties (Waddani, Kulmiye, and UCID, with their respective presidential candidates Cabdiraxman Cabdillahi Cigaal, Muuse Biixi Cabdi, and Faysal Cali Xuseen respectively) were being challenged by seven associations for recognition as national parties: (Barwaaqo, Rejo, Shacabka, Hilaac, Talo-wadaag, Horseed, and Kaah).
Second, this was arguably the most contested and also contentious election in Somaliland’s history. President Bihi’s first term should have ended in 2022. While opposition parties accused the president of deliberately delaying the election to stay in power, he justified the postponement on questions over the scheduling and sequencing of party and presidential elections. As a result, the House of Elders (or Guurti) extended the president’s mandate by two years and its own by five years.
Protests against the postponement of these elections led to a limited clampdown by security forces. In the region of Sool political unrest led to a significant loss of life, which tainted Somaliland’s hard-earned reputation for political compromise, mediation and stability. Still, support for the new electoral law by all political parties has helped to recover this reputation to some extent.
Related to these events this election suffered from a comparative lack of international assistance which hampered local observation efforts, especially from the European Union, which ceased funding the local election observer organisation, SONSAF, causing it to radically reduce its observation team to under 300 persons compared to 2021 when it fielded over 800 observers. This has limited its national coverage and led to a concentration of its efforts especially in the urban areas. SONSAF also did not conduct a parallel voter tabulation (PVT) beyond its coverage of select polling stations observed.
There were no serious, publically-available polls (other than unreliable social media ones) prior to the election and campaigning was said to have been conducted less on issues distinguishing the parties than clan allegiance, even though issues especially around youth unemployment and migration and economic stagnation were prominent in discussion.
Somaliland’s agreement in early 2024 to grant Ethiopia a 50-year lease to a coastal area near Berbera created some doubt in Somaliland’s international relations, reinforcing competing interests between regional rivals, including Djibouti, UAE, Qatar and Somalia, the latter which denounced the move given its claim of sovereignty over Somaliland. This has in turn served to destabilise the political order in Somaliland itself, not least through reported external political funding flows, especially to opposition parties over the government. Despite legislation to declare such funding sources, these flows were not subject to public scrutiny, and may have placed the ruling party at a significant disadvantage.
The election process comprised a considerable logistical and financial effort for a country with a per capita GDP among its six million people of just $630. Despite this, 1,227,048 individuals were registered to vote across 2,648 stations countrywide. Security was provided by 5,000 officers. This was the first election in which the IRIS biometric system was deployed, capable of offline checking, to 1,068 polling stations, mostly in the urban areas. Over 13,000 polling officials were recruited, in collaboration with public universities in the six regions, ensuring a highly competitive selection from senior students and university staff.
The NEC partnered with 23 television stations, 4 newspapers, 18 websites, 8 YouTubers, and 62 social media influencers to promote accurate reporting and voter awareness.
The Brenthurst observations were made according to the following set NEC procedure:
- Polling stations: These were mandated to open at 0700 and close at 1800
- Voter Identification: Voters were to present their identification card which was to be verified via IRIS biometrics and check on ink marking. In the event of the absence of biometrics, a manual check was to be conducted.
- List Verification: The Secretary was to verify the voter’s name and card number.
- Political agents: The ten parties/associations were eligible to have agents representing them at the polling stations.
- Confirmation: The Secretary was to underline voter information and apply indelible ink to the voter’s hand.
- Ballot Issuance: Each voter was to receive two ballots for marking, and after voting, to deposit these in the correct ballot boxes. Their ID card was then recovered by the voter.
- Counting: Polling stations were to be converted into tallying centres, where ballots were to be counted in the presence of party agents. Results were to be recorded in forms D1-A (presidential) and D1-B (party) and physically transferred to district tallying centers. Data from all polling stations in the district was to be captured on the Result Management System at the district tallying centre, where D2-A and D2- B forms were to be produced. These forms were then to be forwarded to the regional centres where regional results were to be compiled and brought to the national center for NEC’s final tabulation. Provisional results were to be submitted to the Supreme Court for verification.
The NEC confirmed that polling stations were to be limited to 500 voters for efficiency. Local and international observers were explicitly permitted to observe the voting and counting process.
Election Day Observation
The Brenthurst Foundation, which sent an observation team to the 2021 election, deployed 20 international observers across four districts during the 2024 event: Hargeisa, Borama, Burao and Berbera. In all the group observed voting at a total of 635 stations against which 317,500 voters were registered, thus comprising 26% of the electorate and 24% of all polling stations.
The Brenthurst observer mission was deployed between 9 and 15 November 2024. It was thus not within the scope of the mission to monitor and evaluate the impartiality of the pre-election phase. However, the three main political parties pronounced themselves happy with the fairness of the process.
Brenthurst Observations
Observers from all areas reported that voting took place peacefully. Polling stations were well-managed by staff – mostly university-age women – and opened on time with few exceptions. The presence of large numbers of women voters was also notable.
The voting procedures in the stations were adhered to with voter identification, the allocation of ballot papers, voting in secret and the use of indelible ink to mark voters.
Where necessary, party agents were alerted to the choice made by a voter who required their assistance of an official, to verify that their choice had been correctly marked.
The IRIS system was operational in the vast majority of the polling stations where it was installed, and NEC staff routinely double-checked entries against the physical voter’s roll.
A few voters appeared younger than the minimum age of 15, but they possessed proper identification.
Voting predominantly happened in the morning, leading to high pressure in some venues. The movement of voters may have been hampered by the grounding of public transport during voting day.
After the closure of polling stations, counting proceeded in the open with party agents verifying voter choices and final tallies agreed to by party agents.
Voting materials such as unused ballot papers, stubs and other materials were properly accounted for and recorded with the agreement of the agents.
We witnessed two notable security incidents. In Hargeisa, a confrontation erupted between voters, forcing police to fire a live round into the air as a means of dispersing the crowd. In Burao, crowd control at one polling centre disintegrated in the early morning shove towards the gate to the facility, with the police having to use whips and batons to manage the crowd. In both instances, voting at the station/centres continued.
Overall, the conditions in the 2024 election were a significant improvement on the 2021 event, especially in terms of lighting, the biometric system and efficiency of counting.
There were no serious incidents which threatened the integrity of the election on voting day. In our opinion, this election was free, fair and credible despite the constraints of Somaliland’s financial and institutional means.
Brenthurst Recommendations
Eleven recommendations follow from this mission:
- The independence of the National Electoral Commission should be maintained and strengthened.
- The need to treat international and local observers equally.
- The need for donor funding for local monitors. It is regrettable that some foreign donors withdrew support for the backbone of local observation efforts, in so doing playing along with regional politicking.
- A thorough check on voter registration and the roll to eliminate the risk of under-age voting.
- A clear ban on party political politicking on election day.
- Improved sign-posting and geo-location of polling stations and venues, and greater attention to the needs of voters with disabilities.
- The IRIS system should be extended country-wide; the system and related lap-top requires an independent power system capable of lasting the entire day.
- Crowd marshalling and control should be extended to the outside of polling centres, thus avoiding crushes at entrances.
- Transparency would be enhanced by the publication of polling centre results at the station once counting is complete, in line with international best practice.
- The need for transparency and monitoring of financial flows to political parties, and for full disclosure as per the law.
- Further efforts should be made to promote issue- over identity-based politics, not least through encouraging public policy debates among candidates.
The Brenthurst Foundation
Somaliland Election Observation Mission
9-15 November 2024
Head of Mission
1. Moeketsi Majoro (HE Dr), fmr. Prime Minister of the Kingdom of Lesotho
Members
2. Atupele Muluzi (Hon), President: United Democratic Front, Malawi
3. Aziz Amin (Mr), Co-CEO: Alpha Development Solutions, Afghanistan
4. Choolwe Chibomba (Mr), CEO: Timedia, Zambia
5. Diana Games (Ms), CEO: Africa @ Work, South Africa
6. Edwin Sifuna (Hon), Secretary General: Orange Democratic Movement, Kenya
7. Emily Osborne (Ms), Associate: SABI Strategy Group, United Kingdom
8. Erastus Mwencha (Amb), fmr. Deputy Chair of the Commission of the AU, Kenya
9. Greg Mills (Dr), Director: The Brenthurst Foundation, South Africa
10. Ioannis Kollias (Mr), Lawyer and election observer, Greece
11. Karin von Hippel (Dr), Director-General: RUSI, United Kingdom
12. Kate Almquist Knopf (Ms), Special Advisor: The Brenthurst Foundation, US
13. Mark Burke, Member of Parliament, South Africa
14. Matt Bryden (Mr), Political Analyst, Canada
15. Ray Hartley (Mr), Research Director: The Brenthurst Foundation, South Africa
16. Richard Morrow (Mr), Analyst: The Brenthurst Foundation, South Africa
17. Robert Thomson (Maj Gen ret.), fmr. Senior Officer: British Army, UK
18. Tanele Maseko (Mrs), CEO: Thulani Maseko Foundation, eSwatini
19. Tendai Biti (Hon), fmr. Minister of Finance, Zimbabwe
20. Zenzele Mfayela (Mr), Member: Inkatha Freedom Party, South Africa
Appendix 1: Hargeisa
The team of eight deployed to the Hargeisa district within the Maroodi region where a total of 222 polling stations were observed. This provided a sample size of an estimated 111,000 potential voters and comprised 29% of the polling stations in the region.
There were heightened levels of excitement at the entrances of many centres and stations prior to the official opening at 0700. Few stations opened at 0700 owing to administrative delays (i.e. sealing boxes, briefing of party agents), however, these were not viewed to have jeopardised the voting process. One station only opened at 1030 as the IRIS system had not been delivered.
Voting throughout the day remained peaceful, calm, and orderly.
The IRIS biometric system proved to be erratic: for some stations there were no issues, however, at others, its use was sporadic owing to technical difficulties. That being said, polling stations under the guidance of their managers were able to seamlessly transition to the physical voter’s roll, thereby ensuring the voting process remained on track. A handful of stations refused to commence with the voting process until they had successfully set up the IRIS system, thereby delaying opening to as late as 1030. A recommendation would be for clear thresholds to be set by the NEC on when to switch to the physical voters roll as a fallback.
At some centres, ballot boxes lacked correct covers and could not be properly sealed. In all these locations agents and observers agreed to proceed with voting until the problem could be rectified. The presence of political figures on election day jeopardised the orderliness of voting procedures. A Waddani convoy passing through a neighbourhood which featured 9 polling stations (i.e. 4500 registered voters) caused friction between party supporters and opponents.
This ultimately led to a confrontation between voters and a gunshot being fired by police to disperse the crowd. One individual was later seen being escorted by police towards their vehicle. The Waddani presidential candidate was also seen visiting three separate voting stations in the Hargeisa district, posing for pictures with supporters inside stations. The presence of his media team suggests this was a deliberate public relations initiative.
Many voters needed assistance to mark their ballots, which was provided by NEC officials. This was conducted transparently, with the consent and approval of party agents. Voting closed at 1800 at which point the ballot boxes were sealed in the presence of all party agents and observers. counting commenced at 1855. The process progressed in a structured and effective manner, with any disputes (i.e. spoiled ballots) resolved effectively.
Voting in Hargeisa was a success. Despite being the largest voting district in Somaliland, the operational efficiency of all stakeholders warrants praise due to their professionalism.
Moeketski Majoro, Richard Morrow, Kate Almquist Knopf, Mark Burke, Emily Osborne, Erastus Mwencha, Tanele Maseko, and Matt Bryden
14 November 2024
Appendix 2: Borama
Of the 66 centres in Borama, one team visited 21 centres/114 stations and the second went to more rural areas and visited 17 centres/72 stations.
Total: 38 of 66 centres (57%) and 186 of 241 stations (77%).
Overall, the elections observed by our team were conducted peacefully, with no significant issues that compromised the electoral process integrity. Officials, observers, police and military across the stations we visited were generally professional, with attention to detail. The police/military were less professional and didn’t fully understand the process, but the NEC officials and our guides were able to resolve any issues of access to stations. In particular with the NEC staff, we were impressed by their commitment and knowledge about the entire process. They also gave special attention to accommodating vulnerable voters.
Operational Challenges
A few operational issues were observed in both urban and rural polling areas.
In urban areas, most polling stations were equipped with the iris scanning machines, used for voter verification. The IRIS scanners were generally quick and efficient, though in a few stations they didn’t work properly for technical reasons and some elderly people had trouble using it, but the NEC officials were very helpful. The vast majority worked and were quick (unlike tech seen in other countries).
Some rural stations faced delays due to technical issues such as computer connectivity and power availability.
We saw people being turned away in several instances because they went to the wrong centres or stations. We also witnessed three women with voting cards, at their correct polling station, who weren’t in the system/on the voter’s roll.
Despite these challenges, the process was largely orderly, with delays minimal in most instances.
There were significant challenge to access in the rural areas, especially given the condition of roads. The polling stations were not marked so it was a struggle to locate them, even with someone who knew the area well. In addition, we saw political parties campaigning just outside the station, or very close to the station, especially the Waddani party. There were instances of the display of sample ballots already marked to guide voters, Party regalia was viewed inside stations. No one seemed to object. We also saw two stations with the boxes not matching the lids.
Another common problem seen in rural areas related to the political party observers, they were unable to vote in their constituencies which were elsewhere as they had to spend the whole day in the station. They were complaining about it.
In three or four rural areas two or three stations would be crammed into a small space.
In rural (as well as urban areas) we were very impressed by the high numbers of women involved in all aspects of the election – voting, running elections, party observers etc.
Voter Accommodations
Polling stations prioritised access for visually impaired and elderly citizens, with some receiving voting assistance as needed and they were allowed to vote at the opening. Illiterate voters also received assistance, and their votes were fully disclosed to party officials and polling agents for transparency. While one polling station opened later due to the absence of party agents, operations resumed smoothly after their arrival.
One instance of an uncovered ballot box for the presidential election was noted, although it did not appear to impact the voting process significantly. There was also a mismatched cover which was latched with only two fasteners instead of four. We are not sure if that got resolved either.
Incidents
Child Voters: Cases of apparently underage individuals voting were observed. They had the right ID cards enabling them to vote, but appeared under age. There were instances where children went early to hold a place for their elders, but there were also many children who were observed voting.
Military Presence: Some polling stations experienced minor hostility from soldiers, though it did not escalate. Twice the soldiers didn’t want to let us into the room but eventually relented.
Crowd Control: Minor pushing was observed outside polling stations, though it was swiftly managed.
Vote Count
The vote count was slow but professional and they all took their tasks seriously. They shared the final result with us. We saw no issues, and they kept the door locked so that the soldiers would not come in.
Conclusion:
- Long queues in the morning and generally empty stations in the afternoon, in both urban and rural areas.
- We were impressed by the significant female participation across the board – including with voting as well as working at stations and being observers.
- NEC staff were well trained, mostly young, educated, and polite/professional.
- Party observers were in every station in the right numbers (usually 6 people), which was impressive. The observers were generally positive about the process. And they seemed invested in it.
Our overall assessment is that the election in Borama was conducted in a free, fair, and orderly manner. While minor issues were noted, they did not significantly impact the overall transparency or credibility of the election. The only outstanding/concerning issue was the high number of children voting and the potential of coercion or clan intimidation in the rural areas.
A free, fair and credible process but a process that can be improved.
Karin von Hippel, Choolwe Chibomba, Tendai Biti, and Aziz Amin
14 November 2024
Appendix 3: Berbera
A team of four observers visited 82 out of 152 (54%) polling stations in Berbera, Dara Godle, Xamaas. Lafa Ruug, Laaleys and Geel-Loo-Kor.
Voting went off peacefully with voters eager to cast their ballots.
Most polling stations opened at 7h00. The few that didn’t were open by 8h00 due to some technical and technology problems.
Turnout was impressive before polling stations opened but really tailed off after about 12pm and some stations took a lunch break, closing for 30 minutes to eat, which is not technically allowed. However, overall turnout seemed to be low – about 35% in one of the stations observed at closing.
The identification process went smoothly mostly. Party agents seemed to be content with the process. The IRIS identification system worked well.
There was a visible presence of law enforcement, which helped to keep the peace and order.
NEC officials appeared to be in charge and took the details of observers. Standards were good across all voting stations, even those in the remote rural areas that were visited by the team. There was a strong presence of women running the polling stations and some policewomen were on security duty.
Polling booths were inaccessible to PDPs who couldn’t reach the booth. The conduct of some officials was disorganised and unconfident, letting in only one or two people to vote at a time. There was disorganisation in the post-closing process in some stations, which caused long unnecessary delays in the counting process. Staff seemed to lack the confidence to nail the process down.
The ban on cars except for those with NEC plates on election day was a great way to encourage citizens to participate in the elections in urban areas.
Despite funding issues, the election was professionally run and without incident.
In some areas there was coordination of officials to remain open for prayer breaks, but in one rural area (511-4 Berbera) the entire polling station was closed when we arrived at 15h39 for prayers.
As observers we were widely welcomed by all. Polling stations were spacious and well lit and easy and to move in and accommodate people.
Ray Hartley, Edwin Sifuna, Zenzele Mfayela, and Dianna Games
14 November 2024
Appendix 4: Burao
The team of four deployed to the Burao area, but stretched its area of coverage to include Beer and Yirowee in the east, and Qoyta, Galooley and Sheikh in the west, observing 145 polling stations in total across a 65km radius out of Burao.
This provided a sample size of an estimated 73,000 potential voters, and comprised 38% of the polling stations in the region.
Apart from one polling centre, crowd control was orderly and the voting procedure proceeded smoothly, including the operation of the IRIS biometric system which was present in all urban polling stations. There were problems with the IRIS at just under 10% of the polling stations. With six polling stations at the Sooyal voting centre (590) in Burao, a large crowd formed outside the gates, which were opened and closed intermittently to only a few at a time. A combination of an absence of crowd marshals outside led to pushing and shoving and a violent response from the policemen and soldiers behind the gate. This situation soon however settled down once the trickle of those inside had relieved the pressure out. There was unusual pressure on this particularly centre given the numbers.
This problem seemed to be down to the enthusiasm of the voters rather than a more sinister motive. The voters conducted themselves in every other instance in an orderly manner and with good humour given the lengthy queues.
It is however recommended that the security officers should be based outside as well as inside such perimeter gates to such centres, which would have prevented this disorderliness and distress.
At one centre (585-3) one of the three polling stations had not received their voter’s roll at opening. This was received at 09h00.
We noticed several youthful voters, though their exact ages were difficult to discern and communicate. We were reassured on enquiring that if they possessed a voters hard, however, that all was in order and that they were old enough.
A number of voters, presumably illiterate, had to ask for assistance in making their mark. This was handled transparently, and to the approval of all the voting agents present.
We observed several disabled voters. Several of the party agents requested that more to be done in future to accommodate the disabled.
Kulmiye, UCID and Waddani party agents were present at all the stations observed, and on average 3 of the 7 political associations were represented, Kaah, Horseed and Baraaqoo being the most numerous. When prompted, all agents pronounced themselves happy with the process. There were no dissenters.
The staff, the vast majority who comprised women, conducted themselves professionally and were friendly throughout. The university students rose to the occasion and their deployment into the polling stations is an inspired idea, in investing the youth in the future of their country.
The counting process started at 1900, which took awhile to get started but progressed quickly once underway. Much better lighting was provided than in 2021.
Overall, the election in the Burao area was, as best as can be discerned, conducted professionally and methodically within the constraints of Somaliland’s institutional, financial and human resources.
Rob Thomson, Atupele Mulutzi, Ioannis Kollias and Greg Mills
14 November 2024
Signatories: Brenthurst Foundation Observer Mission
14 November 2024
