Aug 10, 2012
Center for Democracy and Civil Society

Puntland’s Democratic Transition: New Electoral Laws

Regions of Somalia

Big news out of Garowe this week. On Tuesday, Puntland’s parliament finally passed the electoral laws that will govern Puntland’s first popular election in decades. Tentatively set for mid-2013, the poll will elect representatives to serve in district councils within each of Puntland’s administrative districts. An additional law will be passed at a later date to address electoral procedures for parliament. But this first step enables political associations to begin registering with the Transitional Electoral Commission for a place on the 2013 ballot.

Although a full English translation is not yet available, I had a chance to meet with the Electoral Commission a few weeks ago to learn about what at that point was the prospective electoral law. Assuming the law’s substance wasn’t dramatically altered by members of parliament, its provisions are quite interesting in how they seek to reconcile a number of powerful, and sometimes contradictory, concerns.

But first things first. Puntland’s district councils will be elected under a system of closed-list, proportional representation. All political associations that meet the Commission’s prescribed conditions will be allowed to contest the initial poll. These conditions include the following:

— A $15,000 registration fee, paid to the Commission
— Recruitment of 500 members from each of Puntland’s eight regions
— The appointment of a single representative to liase with the Commission leading up to the election

Once registered, these associations will develop internal processes to select their list of candidates to present to the public. All lists must include as many individuals as there are seats within a respective district council. For example, if Garowe’s council has 27 members — as it likely will — then all associations contesting seats in Garowe must list 27 names on election day. However, the first five names on all lists must also include at least one of each gender. There is no requirement for ordered placement, but this condition ensures that at least one woman will always be among an association’s first five candidates.

You may have noticed that I keep referring to “political associations” rather than “parties.” This is an intentional distinction created by the Commission. Puntland’s constitution and political party law, while instituting electoral democracy with universal suffrage, limit formal participation in representative bodies to three political parties, all of which must be registered with the government. How will these parties be selected? Through the results of the first district council elections, which will be open to all political associations that meet the aforementioned conditions. If, for instance, twelve associations register and compete, all that receive enough votes to claim at least one seat may join that particular district council, but the top three vote-winning associations for all of the councils will be singled out and formally registered by the government as political parties. No other formal associations will be allowed to contest elections for a period of ten years, after which the Commission will open the system once again for fresh competition to select the next set of three official parties.

With no census to set geographic distribution of representation from various districts and regions, the size of the councils will be determined by the electoral law. Yet this presents a somewhat bigger problem for the new parliament. Although the law governing that election is still forthcoming, the Commission thinks that one possible solution might be using voter turnout rates. In other words, a particular area’s turnout in district council elections will determine its allotment of seats in parliament as well.

This hasn’t been finalized, of course, but the ingenuity of such a plan could be matched by some pretty striking consequences as well. First, let’s take a step back and consider the law as a whole. The Commission had three primary concerns in designing a system: simplicity, legitimacy, and clan influence.

The first, simplicity, was addressed through proportional representation and the mechanism of translating votes into seats. The final law will spell this out in more detail, but my understanding is that the most likely option is assigning each seat a vote value based upon the total number of votes cast. According to one of the technical advisers to the Commission, this may be one of the easier options to understand, particularly for a population with little or no memory of elections. It means that 27,000 votes cast in a district with 27 seats would give each seat a value of 1,000 votes. The second concern, legitimacy, can be seen in the level of participation. Allowing an unlimited number of registered associations to contest the initial poll should in theory leave the population with the impression that the playing field was level, and that everyone had the ability to express their political preferences. Universal suffrage, of course, furthers this aim as well. Finally, clan influence was addressed in perhaps the most explicit manner — requiring all associations to collect 500 members from each region. Because clans are often located in particular geographic areas, this is a clear attempt to prevent the organization of parties around individual clan lines. Longer term, the hope is that requiring parties to recruit and maintain members from every region will help develop a culture where parties speak to cross-cutting issues rather than clan concerns.

Can these provisions serve their intended purpose? Time will tell, but the toughest of the three concerns, not surprisingly, will likely be clan and traditional sources of influence. Despite 500-member rule, two other provisions may have the unintended effect of boosting the role of clans. The first is the $15,000 requirement for association registration. This is a steep sum in a country where well over 50 percent of the population is nomadic and per capita GDP remains one of the lowest in the world. But even this only spells part of the story. According to an adviser to the Commission, conducting a full campaign for the first election could cost an additional $200,000. Aside from wealthy Somali businessmen and ex-pats living in the West, clans seem the most obvious place to secure that type of funding. And there are no campaign finance laws to regulate contributions.

The second problematic provision hasn’t yet been established, but deals with the census issue. If the Commission decides to base geographic representation on voter turnout, as mentioned above, this naturally provides a large incentive to get out the vote. As we see in the US and elsewhere, this requires organization, money, and intimate knowledge of communities — all of which clans possess in spades. It would not be a surprise at all to see traditional leaders mobilize their communities to vote early and often (quite literally, since there are no ID or registration cards).

Considered on the whole, this law seems to create an initial system that is comprehensible and transparent. It sets clear rules for participation, utilizes proportional representation to respect plural politics, and grants communities a local elected body to serve their interests. We may have to wait for the parliamentary electoral law to make broader judgments about the entire system and its corresponding incentives, opportunities, and constraints. And one much larger question concerns the overall compatibility of this system with the new Somalia constitution that I wrote about earlier this week. For now, it will be interesting to watch as new associations register, and to observe how all participating groups choose to distinguish themselves during the campaign season. Their composition, platforms, and ultimate voter-base may speak to the success of the Commission in addressing their biggest concerns.



  • You have provided some great details about the electoral law, which would otherwise be impossible to find publicly. A few thoughts/questions:

    Recruitment of 500 members seems like an easy enough hurdle for any tribe to overcome. I’m not sure of the requirements for membership but this would seem to be something that could be easily bought for money. Unlike vote-buying, this would be impossible to make illegal, and would be much easier to ensure compliance. Once Puntland moves forward with national-level elections, it would make more sense to require any party to obtain a certain amount of votes in every region to take its share of seats. I think this is an overrated method of building national consensus but it has precedence and would require greater effort for building cross-cutting cleavages than the recruitment requirement.

    I want to make sure I am interpreting the party requirements correctly. Does this mean only three parties per region will be allowed to compete at the national level or the top three parties total? If the latter is the case, then that raises a host of issues. What is the point of the 500 member requirement if there are only three parties? I imagine this is an attempt to move elite bargaining to the pre-election stage, as opposed to the post-election stage, which we would be the normal procedure for a parliament. This would be because coalition making would now be done before the election (with tribes picking which of the three parties to affiliate with) as opposed to after. I would have to think through the implications of that, but I suppose it could produce more stability, assuming there are also laws in place preventing party switching. However, if this is part of a strategy to manage salient ethnic cleavages (I don’t know much about the local context) then this is really more of a Satori/Horowitz strategy than a Lijphart one.

    The idea that institutional design can negate the influence of tribal leaders is admirable, but will probably fail. Tribes are institutions too, and they have a lot more history behind them than a new parliament. I’m not arguing against trying to shape things with electoral system design, just that expectations for the impact of those rules should be low at first.

    Do we know the electoral system for the future parliament? Basing district magnitude on voter turnout from one election has its issues but, as you alluded to, absent an accurate census, this is probably the best they can do. One CLPR district would easily solve this problem, and given Puntland’s overall size, I think that would be a reasonable option.

  • Hey David, thanks for the comments.

    On the party limit issue, only three parties will be allowed to contest elections (district and parliament) for a period of 10 years. These will be determined by the top three vote-getting associations from the first district council elections (the largest vote share from all the councils combined). All registered associations (I think) must submit lists for every district in the state (21 right now, but it might increase down the road).

    And you’re absolutely right, a 500-member rule isn’t much of a hurdle. I asked the commissioners and others many times about various scenarios that might lead to different clan alignments, and I’m inclined to agree with you that this set-up shifts the negotiation phase to the pre-election (and even pre-registration) period, where small clans will be forced to join larger clans if they hope to receive representation. Alternatively, large clans (and there are 3 main ones…somewhat ironically) are in a position where they may be held hostage to the wishes of smaller clans, if small clan participation is necessary (at least initially) to form a party at all. This could lead to a variation of king-maker politics, and give small clans disproportionate power.

    But this sort of parliamentary system with only 3 parties may in fact create a degree of government stability. Presumably, only 2 (perhaps sometimes even 1) will be needed to form a majority coalition, which in turn will form the government. There are provisions in the constitution that speak to party participation. Party-hopping is effectively banned, with members losing their seat if they leave their party. Another provision tried to get parties to adopt national agendas, with a requirement that they have branches in all regions (to go along with the 500-member rule). I also think that this isn’t much of a hurdle, since as you mentioned a minimum vote-receiving quota from each region might be more effective for that. But the constitution also endows the Electoral Commission with the power to dissolve or deny registration to parties that violate principles in the constitution — which includes a ban on agendas based upon clan or military affiliation or interests.

    The system for parliament will be similar, although I don’t know what the district magnitude will be. The current parliament effectively distributes seats regionally, since traditional leaders are responsible for nominating members. In the future my sense is that each district will be given a few seats. I’d have to think it through some more, but CLPR with one state-wide district could actually exacerbate the clan issue. Parties might try to run up the vote in areas with larger clan support, and largely ignore (campaign and agenda-wise) other regions. Forcing them to serve up lists in districts all over the state at least nudges them in the direction of responsiveness to many different communities.

    In the end, the effectiveness of any of these laws in achieving their goal may come down to the simple willingness or ability of political actors to enforce them — which I suppose isn’t a great shock. Many people and leaders go to great lengths to decry clan politics in public, but there’s no denying its power. Traditional leaders have been supportive of the process so far, but I think that’s because they understand that their power rests in something much more important than political institutions. They don’t seem to be threatened — which in one way is positive since their support is needed…but also could be a troubling sign.

  • […] [i] After registering for a “political association”, the top three groups following the district vote will become Puntland’s official political parties, and represent their constituents in the state’s representative bodies for the next ten years. The other registered associations will be eligible for only district level seats. See Josh Linden’s reflections from the Center on Democracy and Civil Society at Georgetown for greater detail:…. […]

Leave a comment

Founded in 2004, Democracy and Society is a biannual print journal published by the Center for Democracy and Civil Society at Georgetown University. The D&S Blog provides web-only content, including special reports and investigative series, on issues relating to democracy and development.

Email Subscription to D&S and Blog

* indicates required

Posts by Region

Posts by Topic

Switch to our mobile site