Suspension of the Tunisian constituent assembly

Tunisia’s transition, which once seemed very promising, has entered quite a crisis, especially since the assassination of a prominent opposition leader on 25 July.

From the BBC:

Tens of thousands of protesters have marched in the Tunisian capital, Tunis, to demand the resignation of the Islamist-led government…

Earlier, the constituent assembly was suspended until the government and opposition open negotiations…

The speaker of the National Constituent Assembly (ANC), Mustapha Ben Jaafar, said work on a new constitution and electoral law would resume once “dialogue commences”.

The constituent assembly’s being “suspended”, I take it, is more like a recess and not a precursor to a dissolution. Still, not a good sign. It was elected almost two years ago, and is well overdue to produce a constitution.

Egypt draft constitution on religion

I was pleasantly surprised to see this in the draft constitution for Egypt:

Article 3 The practice of Christianity and Judaism

Persons embracing Christianity and Judaism shall have the right to revert to their respective religious laws in matters relevant to personal affairs, the practice of religious (affairs) or (rituals), and the nomination of spiritual leaders.

On the other hand, there is this (in Art. 2):

Islam is the religion of the State and Arabic is the official language. The principles of Islamic Law are the main source of legislation and Al-Azhar is the final authority in the interpretation thereof.

and this:

Article 4 Al-Azhar

Al-Azhar is an independent Islamic authority, headquartered in Cairo, and responsible for the affairs of the Islamic nation and the world as a whole. […]

(Egypt’s draft, at least the English translation posted, does not yet define the system of government.)

Tunisia’s draft constitution: Contention over parliamentary vs. (semi-)presidential

A draft, in English, of the Tunisian constitution has been posted at Constitution Net. The draft includes the competing provisions that are being debated by various factions in the Constituent Assembly. Most of the remaining contention concerns the executive format.

The Ennahdha party, which won by far the largest share of seats in the Constituent Assembly election of October 2011, but short of a majority, is campaigning for parliamentarism. This is not surprising, for if the fragmentation of the rest of the party system continues, Ennahdha would be almost assured of holding the prime ministership in a parliamentary system. However, it might struggle to win a presidential election, given that Ennahdha was short of a majority of the vote, and that the opposition might more easily come up with an electable individual candidate than coordinate on an alliance for parliamentary elections.

The two main options are parliamentary and semi-presidential. Under the former proposal, the head of state would be a president selected by the parliament, and all major executive authority would be vested in the prime minister, who along with the cabinet depends on the confidence of the parliamentary majority. Under the latter, the head of state would be popularly elected by two-round majority (see Art. 45).

Moreover, in the proposal with the elected presidency, it would be this president, more than the premier, who would have significant executive power. For instance, the semi-presidential proposal grants the president a veto. If the law is one that the constitution classifies as an organic law, it would take two thirds to override the veto. Otherwise, it would take an absolute majority (half plus one). Even the latter is stronger than the veto that the (unelected) president would have under the parliamentary proposal, which is suspensory only, requiring no larger majority than the original passage of the bill. In case of passing the bill over the objections of the president, it is the president of the chamber, rather than the president, who promulgates the law in the parliamentary proposal (see Art. 57).

Both proposals feature a constructive vote of no confidence (Art. 71), by which the cabinet can be voted out by a majority of legislators only if, on the same vote, the majority elects a new prime minister. Both proposals say that it would take one third of members to initiate a no-confidence vote.

However, the semi-presidential proposal places restrictions on no-confidence (or “censure”) votes :

In the event the specified majority is not attained, the motion of censure may not be reintroduced against the government except after the elapse of a six-month period.

In the event the specified majority is not attained, the motion of censure may not be reintroduced against the government except after the elapse of a six-month period.

I do not see any provision clearly granting the president the right to dismiss a prime minister and cabinet unilaterally. In fact, in addition to the confidence vote provision, both versions at Art. 67 say, “The government shall be held accountable before the Chamber of Deputies.” Without a provision empowering the president to dismiss a cabinet, the system does not meet the criterion of president-parliamentary, the subtype of semi-presidentialism in which the president has a more dominant constitutional role (compared to premier-presidentialism, with its exclusive accountability of the premier and cabinet to the legislative majority). Nonetheless, I might be tempted to classify this proposal as president-parliamentary, given the limits on government responsibility in Art. 71. It would seem that somehow the cabinet must remain responsible to someone during the periods when the parliament can’t engage the government’s responsibility, and that someone would be the president (even if de facto).

The semi-presidential version also allows no-confidence votes against individual ministers, which would seem to undermine the core parliamentary principle of collective responsibility. Such provisions are, however, common in president-parliamentary systems (but not, I think, in premier-presidential ones).

The president, on the other hand, would have little discretion in selecting a prime minister, having to grant the leader of the largest “electoral party or coalition” the first right to form a government after an election (Art. 66). This almost seems, in the present Tunisian party-system context, like a provision meant to maximize the chance for cohabitation! ((Of course, if other parties hold out for a coalition including parties allied with the president, then the resulting cabinet would not be a cohabitation cabinet. But my point is that the constitutional provision appears to give more leverage to the head of the largest party than to the president, which is unusual in a semi-presidential system. An exception that comes to mind is Ukraine, between 2005 and 2012, but in a very different party-system context.))

There is a potentially good argument to be made for a popularly elected counterweight to the dominant Islamist party that has emerged in Tunisia. However, the semi-presidential proposal currently under consideration would seem to risk confrontations over powers between the presidency, on the one hand, and the government and assembly, on the other hand.

Tunisia preliminary results

See the Tunisian Election Live Results Table for regularly updated district-level results.

There are still many districts not reported, and some missing data in those districts that are reported. However, the pattern is clear. Out of a total of 217 seats, 145 have been allocated as of my check. The Ennahda has 60 of these, which is 41%. The next largest parties, Congress for the Republic (CPR) and Aridha Chaabia (Popular Petition) have 21 and 20, respectively. Then Ettakatol with 13 and PDP with 6. No other party currently has more than 4. Thirteen parties are currently on one seat apiece.

The picture is thus one of a single dominant, but likely not majority, party, and a fragmented rest of the field.

At the district level, the expected pattern for the simple (Hare) quota and largest remainders (SQLR) system can be observed. This system is well known among electoral-systems specialists for its tendency to benefit small parties. Seats are much “cheaper” in terms of votes required to win, via remainders than via quotas.

The way SQLR works is that a quota is defined as 1/M of the votes, where M is the magnitude of the district (the number of seats available). In the first state of seat allocation, a party wins seats according to how many of these full quotas it has. Whatever number of seats are not yet allocated after seats by quota are determined are then allocated via largest remainders. In this stage of allocation, the parties’ remainders are ordered from largest to smallest, and a seat–at most one per party–is assigned to each in descending order until all are filled. ((It is this incorrect to call systems like this a “largest remainder” system, as is often said, because that term describes only the stage of allocation that takes place after quota seats are assigned. There are several different quotas (for instance, 1/(M+1) instead of 1/M), and the definition of quota that is used will have a significant effect on the remainders that parties have left over once quotas are assigned.))

A party’s seats in each district are thus the sum of its quotas earned, plus a remainder seat, if any. Under SQLR a large party will use up most of its votes on quotas, and then be eligible for at most one remainder seat. Smaller parties, on the other hand, often will have their full vote totals in the district as their remainder, and hence the last seats allocated in a district often require a much smaller number of votes to win than did the first (quota) seats that were assigned.

The process, and its impact on the cost of a seat, can be demonstrated by reference to a couple of the declared district results.

In the district of Gabes, M=7, Ennahda won 4 seats, and three other parties have one each. All three seats allocated to parties other than Ennahda were remainder seats. Total votes were 138,375. Ennahda had 73,416, or 53.5%. So it is only slightly over-represented, with 57.5% of the district’s seats. Its cost per seat is 18,354 (73,416/4). Each of the three other parties that won a seat has fewer votes than this, and the smallest of them paid only 7,351 votes (5.3% of the total) for its one seat, or 40% of what Ennahda paid for each of its four seats, three of which it won via quotas, and one by remainder.

A second, very striking, case shows how SQLR can sometimes under-represent the largest party. In Jendouba, M=8, Ennahda won 2 seats, and six other parties won a single seat apiece. Ennahda’s votes were 33,136 out of 118,376, or 28%. Yet it won only 25% of the district’s seats. The quota is 118.376/8=14,797. Thus both of Ennahda’s seats were won via quota, and it paid 16,568 votes for each of these. The other parties that won a seat each had vote totals ranging from 12,433 to 3,599. So the smallest party won a seat at a cost that was a mere 21.7% of Ennahda’s per-seat cost. Ennahda missed winning a third seat by 57 votes, as its remainder was 33,136-(2*14,797)=3,542. One party winning a seat on 3% of the votes cast, while another wins 2 seats on 28% is an odd result, but one that is inherent to the formula used, SQLR.

Tunisia Live does not seem to have a running total of national-level votes, so it not possible to tell the vote percentage upon which the preliminary 41% of seats for Ennahda is based. As the two district results discussed here show, SQLR can either over-represent or under-represent the largest party, depending on whether the precise vote totals allow the largest party to win one of the remainder seats or not, and of course on the district magnitude.

It is possible that Ennahda has won a bit more than 41% of the votes. It is also possible that it has less than 40%. Even the latter would imply a lesser degree of over-representation than most “proportional representation” systems would provide the largest party.

With one party so dominating the rest of the field, and smaller parties earning seats much more cheaply than the dominant party, we might anticipate that SQLR will not be the electoral system that this Constituent Assembly adopts for the next election.

Tunisia’s election


Tunisians voted today in elections for a constituent assembly. Turnout is reported to have been around 70%.

According to an election guide prepared by the Project on Middle East Democracy [PDF], the electoral system has the following key features:

An assembly of 217 seats

33 districts, 27 domestic and 6 for Tunisians abroad.

Maximum district magnitude of 10.

Closed lists.

Gender quota requiring every other candidate on a list to be a woman.

Simple quota and largest remainders allocation rule, with no legal threshold.

The six districts for overseas Tunisians will elect a combined 18 seats (magnitudes vary from 1 to 5).

These provisions would mean an average district magnitude of 7.8 7.4, not including the seats for overseas Tunisians.

The simple (Hare) quota with largest remainders tends to favor small parties (for a given magnitude), especially given the large number of parties running lists. Thus, despite a laudable gender-balance provision, many party-district contingents will be of just one (male) legislator–a good case of the inter-party dimension affecting the intra-party dimension.

To be clear, this not a “mixed proportional system” as one blog covering the election states. It is a pure list system, with all seats (again, leaving aside those for expatriates) being allocated via PR. ((Many–in fact, most–PR systems use multiple districts; very few allocate all seats in one nationwide district. I point this out because the cited blog appeared to be referring to the presence of districts when calling the system “mixed”. This is simply not correct terminology.))

The big question, of course, is how well the Islamist party, en-Nahda, will do.

I am not sure when we can expect results. Al Jazeera is running a live blog on the election.

Of course, around here we are delighted that this vote was made possible by the actions of a fruit vendor, even if we take no delight in self-immolation, per se.

Tunisian election delayed

The election of a constituent assembly for Tunisia, originally scheduled for 24 July, has been rescheduled for 16 October.

A delay in a first post-authoritarian election is not always a bad idea, but in this case, “larger opposition parties… fear the interim government may renege on its promise to lead Tunisia toward democracy,” according to Reuters. And almost three months is a long delay.

Tunisia election plans

A promising step for Tunisia’s (potential) democratic transition is the decision to have the first election following the fall of the dictator be for a constituent assembly. It has been announced that the election will be 24 July.

It is not clear (to me) what either the shape of the emerging party system or the electoral system might be. But one item, on, offers some hints, in reporting on a conference that took place in Tunis on 3 March.

“After we realised the first miracle, which is the miracle of revolution, we now need to realise a second miracle, which is the miracle of democracy, freedom and peace,” Culture Minister Ezzeddine Bach Chaouch said at the opening of the two-day seminar.

The event was organised by the Kawakibi Democracy Transition Centre (KADEM) in co-operation with the Citizenship Centre for the Promotion of Democracy (Mouwatana).

On the possible party system:

The political map is not clear for the time being, according to Kawakibi Centre President Mohsen Marzouk. “The Left is divided into several factions,” he said.

“In the midst of that, the Islamic current seems to be the only organised entity. In addition, they have several cadres who were living abroad and returned to Tunisia after acquiring several experiences,” Marzouk said. He added that he expects the development of a centre-right faction composed of people opposed to both the radical left and the Islamists.

Now, on the electoral system (not that this is highly enlightening):

They [conference participants] also stressed the need to establish an electoral system that would ensure a fair representation of women, young people and minorities. Contributors added that the new system must ensure that weak parties are not marginalised by a single political party in the next constituent assembly.

In addition, participants called for “cancelling the majority system that contradicts democracy and adopting a list system through individual voting in two rounds or a comprehensive list system of voting”.

I am not quite sure what a list system through individual voting in two rounds might be. Here’s hoping they go with the “comprehensive list system,” by which I would understand PR with a low threshold. It seems the best way for a first-ever democratic election in a setting where the shape of the party system hardly can be known.

(I am filing this, and any future entries on North Africa, under “Euro-Mediterranean” because the spirit of creating a “block” with that title some years ago was to emphasize the geographic and historical continuity of the Mediterranean–all its shores–and Europe. Now, Libya aside, we might even see a democratic community develop around the Great Sea. We can dare to hope, anyway…)