Mozambique Political Process Bulletin

2008 Local Election Issue 21 – 9 December 2008

Turnout 46%

The turnout overall was 46% of registered voters, compared to 28% in 2003 municipal elections and 43% in 2004 national elections. As always, turnout was highest in Mocimboa da Praia, a highly politicised town, at 71%. Next was one of the new municipalities, UlonguIé, at 68%. Lowest were Mocuba, at 31%, and Cuamba and Alto Molócuè, at 33%.

In the hard-fought cities of Beira and Nacala, turnout was 57% and 56%. In two large cities with little contest, turnout was below average: Nampula (38%) and Matola (41%). Maputo was just above average, at 47%.

Frelimo dominance increased …

Frelimo will dominate the municipalities, with 41 mayors (with Nacala still to be decided) and majorities in 42 of the 43 assemblies. Renamo was pushed back. There are now 9 municipalities with no Renamo assembly member and 8 in which it has only one. In Gaza, Renamo has a single seat in Xai-Xai and none in the other four municipalities.

In the ballot for mayor (presidente), Frelimo had more than two-thirds of the votes in 33 of the 43 cities.

… but there is still an opposition

But Frelimo lost in Beira, and races were close in Nacala, Gurué, Quelimane and Marromeu – indicating the presence of an active opposition, largely without involvement of the Renamo central office.

Small groups lost ground

Four small parties and three citizens’ lists won assembly seats, but in general the small players lost ground everywhere except in Beira.

The two parties with some national presence, PIMO which had single seats in three assemblies and PDD which (as IPADE) had single seats in two assemblies, were reduced to a single seat each in Beira. Except in Beira and Angoche, PIMO failed to win more than 1% of the vote in any municipality. Where it stood, PDD generally gained between 1% and 2%, and reached 4% in Milange and 5.4% in Gorongosa.

Two other small parties won single seats –Monamo (Cuamba) and Unamo (Milange).

Three citizen’s lists won seats. A new group, GDB, won 7 seats in Beira. Juntos Pela Cidade (JPC) is the only group to have won seats in all three municipal elections, but it has been reduced to a marginal position – from 15 seats in 1999 to 5 seats in 2003 to only 2 this year. Naturma, which won 5 seats in Manhiça in 1999 and did not stand in 2003, won a single seat this year.

Four citizen’s lists won seats in 2003, but of those, two did not stand again and one did not win a seat.

Protest in Maputo?

Normally the number of voters in a city for mayor and assembly is virtually identical. It is possible to vote for one and not both, but this is rarely seen. Therefore the difference in the number of votes in Maputo city is odd – 308,323 for mayor but 314,758 for assembly – which means that more than 6000 people did not vote for mayor.

Frelimo’s decision not to select the present and highly successful mayor, Eneas Comiche, and instead choose someone seen as more closely aligned to the party, was not popular in some parts of the city. Did some people decide not to vote for mayor as a protest?

The importance of nulos

This Bulletin may seem obsessed with invalid votes (nulos), but this year they determined the result in Gúruè and in 2003 they determined the outcome in Marromeu. And we also know that nulos are often involved in fraudulent activity, and that those actions can show up as unusual numbers in the final statistics.

Nationally, 2.7% of ballot papers for mayor (presidente) were left blank and another 3.2% were considered by polling stations to be invalid, usually because they appeared to have marks for more than one candidate. In all, 41,071 invalid ballot papers for mayor (and a similar number for assembly) were sent to the CNE in Maputo for reconsideration. Of those, 18% (roughly one in six) were considered valid by the CNE – “requalified” – and included in the final totals.

Of course, the quality of the polling station staff varies. Catandica, for example had an average set of nulos but only 1% of those was considered valid by the CNE, whereas in Massinga 30% of mayoral nulos were considered valid by the CNE – suggesting that the Catandica staff were much better trained than those in Massinga. But not all discrepancies are so easily explained. Articles below are based on a close study of the final results.

Nulos push Aniceto over the line in Gúruè

A second round in Gúruè was avoided by just 6 requalified nulos. As in Nacala, the race between Frelimo and Renamo was very close, and the few votes for two minor candidates meant that the Frelimo candidate seemed unlikely to gain more than half the votes – requiring a second round between the two candidates with most votes.

In Gúruè, Frelimo candidate José Aniceto had 4934 votes (48.84%) and Renamo candidate Latino Ligonha had 4686 (47.34%). But when the nulos were reconsidered by the CNE, one-third were considered valid, and nearly all of those went to Aniceto – 66 compared to 28 for Ligonha. The final tally gave Aniceto exactly 5000 (50.03%) and Ligonha 4714 (47.17%).

But it was very close. If Aniceto had received only 60 instead of 66 requalified nulos, a second round would have been necessary in Gúruè as well as Nacala.

This is the second time requalified nulos have made the difference. In 2003 the Frelimo candidate in Marromeu had a tiny majority, 1942, compared to 1938 for his Renamo opponent. But when the nulos were requalified, the CNE gave only 14 to the Frelimo candidate and 19 to the Renamo candidate, meaning he won by a single vote.

Half of nulos accepted in Beira and Nacala

In Beira for assembly and Nacala for mayor, more than half of nulos were considered valid. These are high percentages which hint a more than poorly trained polling station staff.

In Nacala, for mayor, there were 2613 nulos – at 5.3% this is double the national average. Of those, 1410 (54%) were considered valid, and of those, roughly two-thirds went to Frelimo. The Bulletin watched the requalification process at the CNE and what was odd about those nulos was that no normal polling station staff would have considered the ballot papers invalid – they were clearly and correctly marked for Frelimo. How did they get into the nulos pile?

By contrast, for Nacala municipal assembly, there were only 2318 nulos (4.7%), of which 36% were considered valid – more for Frelimo but not as many.

The number of votes for the Frelimo candidate and assembly list and the Renamo candidate and assembly list were all very similar and close to 21,500 before requalification. Yet there seem to have been 400 extra Frelimo votes in the mayor nulos pile. Where did they come from? One hint is that there were 450 fewer blank ballot papers for mayor than for assembly. Did someone during the count secretly take blank votes, mark them for Frelimo, and throw them in the nulos pile knowing they would be validated by the CNE?

Beira shows a similar disturbing pattern for municipal assembly. For mayor, 26% of nulos were accepted as valid – within the range of possibility. Of those nulos, 750 went to Daviz Simango, 134 to the Frelimo candidate Lourenço Bulha, 50 to the Renamo candidate Manuel Pereira, and 25 to others. But for assembly, an incredible 59% of nulos were accepted as valid, and of those 1400 went to Frelimo, only 88 to Renamo, and 84 to others. That 1400 ballot papers were accepted by the CNE seems very strange indeed.

Four muncipalities with high numbers of nulos

Nulos exceeded 5.5% with low levels of requalification in only 4 municipalities: Ilha de Moçambique (8.6% mayor and 7.7% assembly), Monapo (8.8% and 8.7%) and Angoche (6.0% and 6.5%), all in Nampula, and Ulongue (5.6% and 6.5%) in Tete. We have already raised questions about the high level of nulos in the Ilha and in the requalification of the nulos by the CNE the Bulletin saw clear examples of ballots for Renamo which had been improperly invalidated by a second ink mark. It seems likely that something similar happened in Monapo, at least, and perhaps in the other two.

Lost bags?

Finally, we note that Quelimane had 14% of its mayor nulos requalified but none of its assembly nulos. Similarly, Mueda had 8% of its assembly nulos accepted as valid but none of its presidential nulos. We wonder if this means two plastic bags of nulos were lost in transit.

Table corrections

The final results tables distributed Sunday contained a few minor typing errors. Send us a note if you spot any more. (j.hanlon@open.ac.uk) These, and any others which we find, will be corrected in the version of the tables posted on our websites:
In English: http://www.cip.org.mz/pub2008/index_en.asp
Em Portugues: http://www.cip.org.mz/pub2008/