Somalia: Staving Off Violence Around Somalia’s Elections


Nairobi/Brussels — Somalia’s elections are fast approaching but the proper arrangements for monitoring and dispute resolution are not in place. To give authorities time to make procedural reforms, and thus lower the odds of turmoil, politicians should seek consensus behind a delay of one to three months.

What’s new? Somalia is organising presidential and parliamentary elections after a mid-September agreement between the federal government and regional states unblocked an impasse that threatened to delay the contests past the current government’s constitutional term limit. Tensions among key parties remain high, and electoral preparations are lagging.

Why does it matter? Poor relations among the federal government, regional states and political opposition will test the September agreement, potentially resulting in a breakdown in the electoral framework or violent contestation of results. Al-Shabaab and the Islamic State in Somalia could disrupt the process and capitalise on any resulting instability.

What should be done? Somalia’s political leaders should work to reduce prospects of electoral unrest by following through on aspects of the September agreement that bolster confidence – such as enhancing civil society monitoring and establishing a dispute resolution process. To buy time, the leaders could explore a short, consensual extension of the electoral calendar.

I. Overview

Somalia is in the throes of hurried electoral preparations, which began in earnest in November amid pre-election political tensions arguably as high as the country has ever seen. For much of 2020, Somalia’s politicians jostled over proposals to move from indirect voting, whereby clan leaders represent their constituents, to a one-person, one-ballot system. The leaders agreed to make this move previously, but the administration did little to take the step until late in its tenure, when the opposition claimed it would be too much for the time allowed. Though the parties agreed in September to stick with indirect voting, they are struggling to prepare the polls, which are supposed to begin in December. Federal and state authorities should promptly fulfil as much of the September agreement as possible, focusing on setting up an enhanced dispute resolution mechanism and boosting civil society monitoring. At the same time, Somalia’s international partners should quietly explore with the country’s political figures a short, consensual delay to cement those changes, which would make the vote fairer and more transparent.

Tensions between Somalia’s federal government and its member states have dominated the run-up to the parliamentary and presidential elections, which are slated for December 2020 and February 2021, respectively. This longstanding discord at its core relates to unresolved questions over power and resource sharing between Mogadishu and the federal states. It has worsened since President Mohamed Abdullahi Mohamed “Farmajo” entered office in February 2017, in large part due to his efforts to exert greater political control over the states, including by strong-arming some of them into replacing their leaders with his allies. Friction over these manoeuvres in turn has hurt cooperation on issues that require the central and regional governments to work together, such as security arrangements, completion of a provisional constitution and holding elections.

Over the course of 2020, a major point of contention between Farmajo and his political opposition, which chiefly includes leaders from some of Somalia’s federal member states, like Puntland and Jubaland, has been the framework for conducting the forthcoming elections. The last national elections in 2016-2017 relied on indirect voting, in which clan leaders selected delegates to electoral colleges, which in turn chose parliamentarians to fill lower house seats; federal state assemblies selected upper house parliamentarians; and the two houses together named the president. The UN and Western donors pressed the Farmajo government to move toward a one-person, one-vote system for 2020-2021, in part to address suggestions that indirect voting had opened up the 2016-2017 elections to manipulation. In February, Farmajo signed an electoral law that would establish universal suffrage, but he faced intense resistance from some member states, in addition to political parties and clan groupings, who saw it as a mechanism for delaying the elections and allowing Farmajo to extend his term.

The electoral law kicked off months of heated sparring between the federal government and member states. Finally, under pressure from international partners, the parties reached a breakthrough agreement in September, which fashioned the outlines of a framework for the forthcoming elections. The September agreement kept indirect voting in place but introduced some new features that could enhance the legitimacy of the process. Because the agreement eliminated a potentially debilitating source of delay (ie, shifting to a wholly new electoral model), it also raised hopes that it might be possible to move forward with elections on the contemplated December/February schedule.

Yet that timeline presents major challenges. There is a great deal of preparatory work still to be done, time is running short and a fraught political atmosphere complicates efforts. The federal government and member states have already missed initial deadlines, with the establishment of the federal electoral commission coming two weeks after a 20 October deadline, and the formation of regional electoral commissions still pending at the time of writing.

Pushing forward with elections without adequate preparation could be risky. With tensions already running high, if the Somali public lacks confidence in the electoral process either because they see it as rigged or just poorly managed, the frustration could spill over into violence and spawn a major security crisis. Previous regional elections in South West and Jubaland states have already shown how differences between federal and regional interests can quickly turn ugly and result in bloodshed. If there is election-related violence, Al-Shabaab (which is already signalling its intention to disrupt the vote) and the Islamic State in Somalia are likely to fan the flames.

There is a more prudent option, but it must be carefully undertaken. Somalia’s international partners could swiftly explore with the country’s politicians a possible consensual extension of the electoral calendar, perhaps of one to three months, to allow more time for preparations. Authorities would then use this period to fulfil aspects of the September agreement that will help protect the vote’s integrity – for example, by helping civil society take on a new role in monitoring the selection of electoral college delegates or helping federal and state authorities stand up an enhanced dispute resolution mechanism. While opposition parties leery of delay may view this idea with suspicion at first, proponents can allay their concerns by sharing a detailed timetable for taking these technical steps. Some in the opposition have privately indicated that as long as the process is moving, a short wait would not necessarily generate major discontent. Farmajo’s administration can help by stepping back from major policy decisions in order to make clear that for the duration of the extension period, it will be acting in a caretaker capacity.

While there may be some risk of violence whatever the course the federal government and state leaders choose, it can be mitigated if, together, they focus on the most important steps to boost confidence in the process. Extending the time for preparations would be the best way to send a joint clear message that they are working in tandem to help make the elections as fair and credible as possible, and that they are committed to working with local actors, civil society and foreign partners to keep faith with that pledge.

II. Somalia’s Fragile Electoral Compact

A. The Torturous Road to Consensus

Disagreements between Mogadishu and federal member states over what electoral model national elections in Somalia ought to follow have been at the heart of tensions surrounding the elections for some time. Farmajo’s administration has pushed for a one-person, one-vote system, in which every citizen of voting age is entitled to cast a ballot, to replace the past system in which parliament is chosen by clan leaders and their delegates. Western governments have encouraged Somalia to move in this direction in part out of concern about the integrity of the indirect model, which as discussed below did not have a good track record in Somalia’s 2016-2017 elections.

The opposition (primarily leaders from states like Puntland and Jubaland that have poor relations with Mogadishu, along with some political parties and clan groupings) have long suspected that Farmajo’s support for universal suffrage is motivated less by principle than by political calculation. They believe that Farmajo has been banking on the monumental challenges of rolling out the new system, especially given the obstacles posed by institutional weakness, the COVID-19 pandemic and Al-Shabaab’s presence in large parts of the country. They think his design has been to use implementation-related delays as a pretext for staying in power.

Farmajo’s signing of a new electoral law on 20 February, giving every Somali of age the right to vote, thus triggered howls of protest from the opposition, who argued that as a result, the schedule for national polls would inevitably slip. Halima Ismail Ibrahim, chairwoman of the National Independent Electoral Commission (NIEC), announced to parliament on 27 June that conducting national elections under the new model might require a delay until August 2021, a statement that drew further angry condemnation from Farmajo’s opponents. Tensions also flared when the lower house of parliament, aligning itself with the Farmajo administration, passed a series of resolutions to close gaps and address other issues in the electoral law. In doing so, it sought to bypass opposition from parliament’s upper house, which since February had aligned itself with the federal member states on questions relating to the electoral model and did not wish to strengthen the law before resolving the underlying dispute.

Sensing the potential for a major political crisis, Somalia’s external partners, including the U.S., UK and UN, pushed Mogadishu into re-engaging with the member states to find a mutually acceptable solution. Discussions initially organised by the states kicked off in July in Dhusamareb, the capital of Galmudug state, as they worked to align their positions on what kind of electoral system to support. At a follow-up round on 19-22 July, with both President Farmajo and then-Prime Minister Hassan Ali Khaire in attendance, all parties agreed on the principle of holding the election on time, thereby making it probable that Somalia would stick with some form of indirect elections, as they would be more realistic to arrange within the timeframe. The parties appointed a committee to develop proposals on voting preparations and procedures and agreed to reconvene.

Although participants made progress at the second Dhusamareb meeting, they fell short of a clear agreement on either the specific electoral model or the timeline for the forthcoming elections. The prospect of a follow-up meeting to solidify consensus and resolve outstanding issues was also put in jeopardy by Farmajo’s 25 July sacking of Prime Minister Khaire. Khaire had been a key intermediary between the president and member states, but had set himself apart from Farmajo by openly advocating a timely election, a goal that the president had not embraced. Against this backdrop, the leaders of Puntland and Jubaland, two member states particularly at odds with Farmajo, boycotted the third meeting, while the leaders of South West, Hirshabelle and Galmudug states defied their call to join in and attended. This awkward dynamic damaged the unified front the member states had presented just a month earlier.

Held in August, the third Dhusamareb meeting officially closed the door on prospects for a one-person, one-vote electoral contest, with all present agreeing to adopt an indirect model similar to the one followed in 2016-2017. The outcome marked a significant climb-down by the federal government, signalling its susceptibility to domestic and international pressure, although some diplomats believe that Farmajo used the time chewed up by negotiations to quietly get the jump on his opponents in preparing for indirect elections.

Puntland and Jubaland called for another meeting in Mogadishu in September, rushing back to the table to ensure their positions were heard. While both regional governments supported the indirect election model, they dissented on aspects of the third agreement in Dhusamareb that would lessen the role states would play in administering the forthcoming polls. Further negotiation brought the parties to consensus on 17 September behind an electoral framework known as the “Mogadishu model”.

The model is similar to the framework for the 2016-2017 elections, when electoral colleges chosen by clan elders selected lower house parliament representatives, with a separate college created to fill each seat. In 2020-2021, however, the sizes of those colleges are set to double to 101 members each while the number of voting centres per member state will increase from one to two – measures that will, as discussed below, make the process more inclusive. The processes for filling the upper house (selected by the state assemblies) and electing the president (selected by the two houses once the new members have been seated) remain unchanged. Furthermore, the agreement calls for the formation of a new federal electoral commission and separate state electoral commissions for each member state, in effect overriding the NIEC.

Once this federal government-member state dispute was resolved, tensions over the elections between the upper and lower house leaderships in parliament subsided. Parliament approved the “Mogadishu model”, in effect legalising the ad hoc agreement and superseding the February electoral law as the framework for the 2020-2021 elections. This action allowed all parties to move from the design phase toward implementation of the agreement. The long, drawn-out process of achieving consensus, however, had eaten up much of the time needed to prepare for the elections.

B. Another Indirect Election, but Maybe an Improved One

The “Mogadishu model” represents a major compromise between the federal government and member states and, in the present circumstances, was a positive development. Still, by defaulting to key elements of the 2016-2017 model, it disappointed many of Somalia’s Western partners and the UN, which had pushed the country to move toward universal suffrage. These partners have never been fully comfortable with the system of indirect elections, which they tend to see as a temporary fix allowing Somalia to select political leaders as it moves toward universal suffrage, a model they view as more compatible with democratic norms. Their concerns grew sharper following a welter of allegations that some candidates in the 2016-2017 contest were either bribing elders to select favourable electoral delegates or engaging the delegates directly, in order to secure their place in parliament.

Some partners also continue to argue that without more elements of a direct vote, Somalia’s system of corrupt and unaccountable politics is unlikely to meaningfully improve, given the paucity of transparency and inclusivity in the election process. Nevertheless, most of Somalia’s partners are slowly coming around to the idea of another indirect election, given that it is backed by consensus and can be stood up more quickly than moving to a wholly new model.

Moreover, the “Mogadishu model” responds to at least some of the concerns raised by Western partners in that it slightly widens citizen participation and includes measures to improve vote transparency. Although the selection process for the upper house is unchanged as noted above, the process for the lower house has been expanded. The electoral colleges responsible for selecting members now comprise 27,775 total delegates, up from 14,025. In addition to simply allowing more Somalis to participate directly in the election process, this expansion of the number of delegates may help discourage (though almost surely not eliminate) seat buying because it will require payoffs to more delegates. The agreement also provides for an additional voting location in each state outside the state’s capital, which will make voting more accessible to constituents, and increases process transparency by creating the basis for civil society to play an oversight role in the selection of electoral college delegates.

III. Destabilising Factors

Despite the consensus achieved on the electoral model, various political fault lines remain wide open. Unless the parties find a way to temper their acrimony, disputes over electoral preparations or even a contested result could prompt violent fallout. Opposition groups, including clan-based organisations, are poised to mobilise against any form of perceived electoral manipulation by the Farmajo-led government, while Al-Shabaab and the Islamic State in Somalia will almost certainly take advantage of any political instability to cause more disruption.

A. The Shadow of Centre-periphery Divisions

Notwithstanding the September agreement, many federal member states are still very much at odds with Mogadishu. Divergent visions of federal governance are at the heart of many disputes between the government and member states, which are particularly focused on power and resource sharing. These disputes contribute to persistent centre-periphery competition.

While the federal member states are frequently quarrelling with Mogadishu, in some cases, the newly installed leaders are likely to bend in Farmajo’s direction. For example, Farmajo’s installation of compliant allies in leadership positions in the regional governments of South West state in December 2018 and Galmudug state in February 2020, in addition to his alleged attempts to do the same in Hirshabelle state’s ongoing election process, could improve his prospects for re-election and thus his quest to centralise power. Not only are these leaders already favourably disposed toward Farmajo, but the states they govern mainly cover territory in Somalia’s hinterlands and are economically vulnerable, thus leading them to look to Mogadishu for financial support. Their leadership may try to use what influence they have over the process to usher Farmajo back into office, though their clout is not unchallenged, making the outcome far from guaranteed.

By contrast, other federal members states still smarting over the Farmajo administration’s interference in their own regional elections, such as Puntland and Jubaland, see the national contest as an opportunity to thwart his re-election campaign. The leaders of Puntland and Jubaland states, each of which controls a busy port and is economically independent, are very unlikely to support Farmajo. The tension between Jubaland and Mogadishu is particularly fraught and has already resulted in a military standoff in the Gedo region, where the federal government has deployed security units and appointed local officials loyal to Mogadishu. While Jubaland President Ahmed Mohamed Islam “Madobe” has engaged the federal government in discussions to resolve the standoff and has signed off on the “Mogadishu model”, the regional and national governments have not agreed on how to conduct elections in Gedo, which remains a tinderbox.

This problem could be significant. The “Mogadishu model” calls for the selection of sixteen of Jubaland’s 43 lower house seats to take place in Gedo’s capital of Garbaharey, where the federal government now exerts control. Holding the polls ahead of a political resolution risks inflaming these tensions further. The situation could get worse still if Madobe insists on federal troops leaving and Farmajo demurs – a circumstance that could lead to violence and prevent the vote from going forward in Jubaland. The leadership in Mogadishu and Kismayo should urgently come up with a plan for the elections to take place, including by following through with previous arrangements between Madobe and Farmajo to resolve the Gedo standoff.

Tensions between Mogadishu and Jubaland or other states could also play out in parliament if an election dispute comes before it. Because Somalia has no constitutional court, a disagreement over, for example, implementation of the electoral agreement could come before the two houses, which would struggle both to define their role in resolving the dispute and to reach consensus with each other. While Farmajo has developed working relations with many lower house representatives, who have looked at election issues through the same lens as the national government, the upper house, in part because it is chosen by state-level legislators, has remained largely out of his control. A provisional constitution that is at times unclear in delineating the roles of each house has encouraged them to compete for power (as they did in sparring over how and when to address gaps in the electoral law) and the absence of a court empowered to review such issues means that their differences can go unresolved.

Consequently, the best and (given the lack of institutional structures to adjudicate this relationship) only way for parliament to reach decisions about major election-related issues in a way that all the constituencies it represents can trust will be for both houses to agree. To this end, their respective speakers should maintain open lines of communication throughout the electoral cycle and avoid undertaking unilateral legislative moves that could shatter the delicate consensus agreement reached in September.

B. Clan and Political Party Mobilisation

Powerful clans are coming together to voice concerns about the election and mobilise against Farmajo, a sign that if electoral contestation takes off, it will involve not just disputes between Mogadishu and the states but possibly also widespread mobilisation of grassroots clan militias. A large conference of the powerful Mudulood clan in Mogadishu in August where clan leaders called for a timely election, placing additional pressure on Farmajo to come to a consensus agreement, was perhaps the most striking indicator of this possibility to date. Residents of Mogadishu have expressed worries to Crisis Group about some clans’ preparations to arm themselves in the event of an electoral dispute. While September’s agreement has dampened some of these anxieties, a contested electoral process would surely reinvigorate them.

Leaders of national political parties have also warned that they will not accept an election they perceive as illegitimate. Major parties like the Forum for National Parties, led by two former presidents, and Wadajir have been outspoken in expressing their concerns about the elections – among other things weighing in about the prospect of a delayed timeline prior to agreement on the “Mogadishu model” and expressing concern that Farmajo may try to use the security forces, whom they perceive to be polarised, to influence the electoral process. These parties are defined more by clan affiliation than ideology. If they signal strong objections to the conduct of the vote, clan-based discontent is likely to follow.

C. Jihadists

While neither Al-Shabaab nor the Islamic State in Somalia violently disrupted the actual vote in 2016-2017, both present a potentially greater challenge this time around.

The Islamic State, associated with an Al-Shabaab splinter group that has developed over the years in the northern tip of Puntland, has been infiltrating the north-eastern port city of Bosasso, one of the new voting locations for 2020-2021. Al-Shabaab has also stepped up suicide and mortar attacks in Mogadishu. These insurgents are likely to deploy intimidation tactics that may affect the outcome, reprising their playbook from 2016-2017, when they assassinated elders and electoral college candidates, but later extended an “amnesty” to others if they sought forgiveness for participating in the polls.

Such tactics may not only cause some elders or delegates to opt out of the electoral process, but also hand the militants an opportunity to seize on post-election grievances that could emerge if groups feel disenfranchised, and channel these into anti-government sentiment.

IV. Protecting the Vote

The federal and state governments, with support from international partners, should focus on technical steps to mitigate the risk that large groups of Somalis lose confidence in the integrity of the electoral process. A federal-level electoral commission, in coordination with state level commissions, will be responsible for overseeing the vote, and should be given the space to pursue the implementation of technical procedures freely. These commissions will have their work cut out for them, but should pay particularly close attention to the following areas.

A. Clan Elder Selection

The process for selecting clan elders who in turn choose the electoral colleges for the lower house is scheduled for November and will be a closely watched affair. The federal and state-level electoral commissions are responsible for verifying the elder lists, in consultation with the clans. Disputes have previously emerged when the selection of elders became politicised, with federal and state-level political leaders attempting to manipulate the lists. This matter has become a major source of contention between the federal government and member states in state-level elections in recent years.

Given how crucial this process is in determining the outcome of lower house elections and, ultimately, the presidential election, elder selection could again become a flashpoint between the federal government and member states. The 2020-2021 process should preferably rely on elders whom clans have agreed should represent them in previous electoral cycles, which can help reduce the risk of a politicised replacement process. Opening up the process beyond that would allow for manipulation by either the federal government or member states, likely resulting in additional contestation. Already, several elders with whom Crisis Group has spoken have accused the federal government of identifying favoured elders in some areas. While it is impossible to fully insulate the process from interference by political actors, the election committees should confirm that the elders selecting electoral college delegates conform to previous electoral lists, and to the extent that new names must be added due to death or other circumstance, ensure that these nominations come directly from the clans.

B. Upholding the Role of Civil Society

A major change the “Mogadishu model” makes in the electoral process compared to 2016-2017 is that it grants Somalia’s civil society organisations an expanded role in the voting process. That role gives them an opportunity to work with clan elders in selecting electoral college representatives. Somalia’s civil society is led primarily by umbrella organisations already active in mediating among electoral stakeholders. Their inclusion in the “Mogadishu model” likely reflects efforts by Somalia’s political leaders to respond to international pressure, particularly from Western countries and the UN, for a more inclusive selection process that is less concentrated in the hands of clan elders. The specifics of civil society organisations’ involvement, however, have yet to be formalised and will need to be worked out once the federal and state level electoral commissions – which will be responsible for managing their involvement – have been fully set up. One task for civil society might be helping develop a set of common standards to guide the elders’ selection of electoral college delegates in all the federal member states.

Somalia’s international partners should offer technical assistance to assist civil society organisations in taking on this monitoring role and, where feasible, providing financial and logistical support to help ensure that they are fully engaged in the process. International partners should also urge that each state-level election commission confirm with the civil society organisations that electoral delegates have been chosen in a fair manner consistent with guidelines they have put in place, prior to the commencement of voting.

C. Setting Up a Dispute Resolution Mechanism

The “Mogadishu model” calls for the establishment of a 21-member dispute resolution body, consisting of nine federal government and twelve member state representatives, which will have the authority to review and decide on all election-related cases brought to it by candidates and electoral delegates, subject to payment of a legal fee. Given the impending timelines, the federal and state governments must staff and train this body as quickly as possible, in order for it to be fully operational in time for election proceedings. But they must also be careful not to cut corners, as this body’s credibility will hinge on selecting impartial, vetted and trained personnel. Somalia’s international partners may be able to help with some aspects of setting up this mechanism – for example, by providing training and logistical support.