The full bulletin in pdf is on https://bit.ly/Moz-El-209
Marromeu shows electoral law cannot deal with serious fraud and political bias
Observers said “no institution can declare with any credibility or certainty which list won in Marromeu” yet the district court and Constitutional Council (CC) rubber-stamped a Frelimo victory. Marromeu shows that electoral court system does not work. Mais Integridade observers found that in many polling stations the presiding officer never finished the count and simply left the polling station. But the district election commission (CDE) claimed that by not waiting all night in the closed polling station, the party delegates left before the count was finished, and therefore could not protest. The court and CC accepted the CDE view.
The electoral law and its court system is biased in favour of the ruling party, but in 2009 and 2013 the CC used its flexibility to redress some of that bias, for example to deal with fake editais in Gurue in 2013. Fake editias are much more widespread this year, but the CC has been much more rigid and not allowed evidence of fraud.
Across the country in the municipal elections, Frelimo used its control of STAE and election commissions to ensure the polling station presiding officers and secretaries were senior people in the Frelimo party. In some places where the opposition was likely to do well, although the count would only take two to three hours, these people were told not to sign or post the editais, and simply wait or disappear, until observers and party delegates gave up and went away. Then they would write alternative editais giving Frelimo victory.
It became a waiting game of who could hold out longest. In the Renamo complaint about Marromeu the party said that in 18 polling stations the editais were never written and given to party delegates. The Mais Integridade observer consortium confirmed that in three schools – 25 de Junho, Samora Machel, and 3 de Fevereiro – presiding officers simply disappeared without posting editais. Observers and delegates gave up at 2 am or 3 am.
But the district election commission in response to the Renamo protest said that party delegates had left before the count was finished, which was shown by the fact that they did not wait all night. The Constitutional Council said that that the district court ruled that “it was proven that the appellant’s delegates ‘left the polling stations before the end of the partial tabulation process when the Frelimo Party was ahead in the votes’.” The CC accepted this. The district court and the CC also accepted the district election commission claims that documents were signed by Renamo, even when Renamo denied this.
Electoral courts are not a check on a biased system
By agreement of Frelimo and Renamo, the electoral system is not neutral or independent, and instead is dominated by the party with a majority in parliament. Renamo’s argument is the no one in Mozambique is “neutral”, so a party based system is inevitable. This was underlined when parliament introduced election commissions with some members chosen by parties in proportion to seats in parliament, and the rest nominated by civil society groups. Party nominated members of the CNE soon agreed informally that civil society members would be chosen according to the political tilt of their NGO, in the same proportion as parliament members. So the ruling party controls all electoral commissions and technical secretariats (STAEs) which actually do the work.
Under pressure from Renamo, two checks were introduced on this ruling party bias. One is that other parties with seats in parliaments have representatives in the STAE electoral machine, allowing them to watch inside the process. The other check is a special electoral court system. Marromeu shows these checks did not work.
The electoral court system has been created piecemeal over time and is not a system at all. The Supreme Court (TS) is the highest court and is also responsible for administration of the court system. Initially it was also the highest electoral court. The Constitutional Council (CC) was created in 2003 to rule on the constitutionality of laws and the legality of actions of state bodies. The role of highest electoral court was passed from TS to CC.
In an attempt to resolve as many disputes as possible locally, district tribunals became the lowest level electoral court, and that system has been changed several times. This year, for the first time, there was an angry public and still unresolved dispute between the TS and CC about which one directed district courts acting as electoral courts. The CC said only it could order a rerun of an election, while the TS said the district court had this power.
In an attempt to speed up procedures, and make better use of district courts, all protests had to be made first at the lowest possible level – polling station or district election commission – and their appeals procedure followed. Only the rejections of those protests could be appealed to district courts and a rejection by the district court directly to the CC. Appeals to the district court had to be within 48 hours of when the polling station or commission decision was made.
There are three gaps in the laws:
1) election commissions are not required to respond promptly, which created a new waiting game – commissions did not respond until after the 48 hour deadline, so no appeal was possible;
2) it is impossible to appeal about being excluded from a meeting and impossible to appeal about a decision that had to be protested at a meeting which they could not attend; and
3) no evidence can be added in subsequent appeals, and all evidence must be presented in the first protest within 48 hours, which is unrealistic for major frauds requiring more investigation.
The role of the electoral courts became even more confusing when Renamo claimed that electoral authorities were submitting fake editais. Courts simply did not know how to deal with this. By law, at the end of the count in a polling station, a signed copy of the final edital (results sheet) is posted on the door of the polling station and given to party representatives; this copy has legal standing. The law does not say how to resolve a case when a party and election commission submit different copies of an edital, both with the same legal standing. The law allows a recount, which is the obvious answer, but is never ordered.
In the Maputo municipal district Kamubukwana court on 18 October 2023 Renamo and STAE submitted editais which were different for the same polling station. The court record shows that Renamo demonstrated that the STAE editais were false – written in blue instead of the official black ink available in the polling station, signed by different people who were not there, and with different stamps. The court simply accepted the official STAE version rather than ordering a recount.
In its Marromeu complaint, Renamo said that 18 editais were fake. The district court argued that falsifying editais was a criminal, not an electoral, offense and so did not consider it. The CC said this was wrong, but as the evidence was not submitted at district level, it could not consider the issue.
CC once did its own investigation, but now does the minimum
The CC sets its own rules, and has not always been so rigid. In its ruling on the 2009 presidential and legislative elections, it said: “It should be emphasised that, in assessing the validity of the electoral results, the Constitutional Council is not limited, under the terms of the law, to examining the documentation submitted by the National Electoral Commission, since such an act of judgement presupposes the verification of the regularity of the acts carried out at all stages of the electoral process, in order to objectively and conscientiously form a judgement as to the freedom, fairness and transparency of the elections.
To this end, the Constitutional Council also takes into account information from other legally recognised sources, namely: (i) electoral litigation appeals, even if rejected on the grounds of prior issues; (ii) reports from national and foreign observers; and (iii) information provided by the media.”
And it did so in December 2013 and January 2014 where it demanded documents relating to the Gurué municipal election from the national, provincial and district elections commissions. It found such “flagrant violations of the law” by polling station staff and the Zambézia provincial elections commission that it cancelled the election in Gurué and ordered a rerun. The elections commissions had given victory to Frelimo, but many of the editais were fake. MDM won the new election.
The new CC website does not have historic rulings, but there is an archive of the old website on http://188.8.131.52/. The two relevant rulings are on http://184.108.40.206/Jurisprudencia/04-CC-2014and http://220.127.116.11/Jurisprudencia/30-CC-2009
Looking more closely at Marromeu
As observers made clear the fraud is so deep that it is impossible to be sure of the results. But we can interrogate the data to make some guesses. Mais Integridade could only obtain results from 39 of 41 polling stations and they report 19 of those with “serious irregularities”. The table below shows the results as reported by Mais Integridade and the district elections commission (CDE).
The CDE reports a higher turnout – 56% compared to 53% reported by Mais Integridade. From the higher turnout the CDE gives all 1000 votes to Frelimo and from the two extra polling stations it also gives 1000 to Frelimo. More than 400 are taken from Renamo. It seems unlikely that the extra 2000 ballots “found” by the CDE were all for Frelimo.
At first sight, the Mais Intergidade parallel count gives a 219 vote margin to Frelimo, but a closer look at the individual polling stations shows that Frelimo gained more than that margin through fraud. There are two common frauds. One is ballot box stuffing, which is done either by putting extra ballot papers in the ballot box, or simply changing the numbers during the count. The best check is to look at polling stations with a very high turnout. The CDE reports the turnout at 56% of registered voters. There are 6 polling stations with a turnout of over 65%; the highest were 89% and 90% both at Julius Nyerere primary school (EPC). Turnouts that high sometimes occur, but they always mean long queues, even at 18.00 at the end of polling, forcing the polling stations to stay open later. This was never reported in Marromeu, so we think these ballot boxes were stuffed – physically or with changed numbers. There are 461 ballot papers over 65% in those ballot boxes. If, as we assume, these are all for Frelimo and we take them out, Renamo wins.
The second fraud is to spoil opposition ballot papers, by putting a second mark for another party on them, or just putting them in the wrong pile. In urban elections, spoiled ballot papers are rarely above 3%. Yet there are 6 over 10% and at EPC 4 de Outobro there are polling stations with 16% and 12%. We take 4% as our cut off and say that all invalid ballot papers over 4% were spoiled intentionally and taken away from Renamo. We find 306 Renamo ballot papers in 20 polling stations falsely declared invalid. That is fraud in half of all polling stations. And this is not small; in one polling station in EPC 25 de Junho 44 Renamo ballot papers were invalidated – not a tiny pile, and one-third of Renamo’s votes in that polling station.
If we take the stuffed ballot papers away from Frelimo, and give the fraudulently spoiled ballot papers back to Renamo – Frelimo’s victory disappears and Renamo wins by 548 votes, as this table shows.
We have to stress that as Mais Integridade says, the fraud was so great that no one knows who won. But we think this estimate suggests that Renamo probably won.
Observation was not easy
Observation in Marromeu was not straight-forward, and Mais Inegridade should be proud of obtaining results in 39 of 41 polling stations. In normal vote-counting in the polling station the presiding officer reads out the vote and shows each ballot paper. The secretary puts tick marks on the classroom blackboard to keep a record, and at the end the presiding officer writes out the results sheets (editais) and minutes (actas), which everyone present signs and gets copies.
In EPC 3 de Fevereiro, in all five polling stations, numbers were not put on the blackboard so observers had to keep score of the count themselves. And no editais were ever posted. At polling station 050330-03 Renamo tried to file a protest and the presiding officer simply walked out with the unsigned editais. In EPC Joaquim Chissano the presiding officer banned the use of the blackboard and did not show the ballot papers as he read them out; the edital of the single polling station gave Frelimo 88% of the vote, the highest in Marromeu.
In EPC 25 de Junho presiding officers banned observers from the count in two polling stations, where editaiswere never posted. These are the two where Mais Integridade could not obtain results.
The count also confirmed one illegal practice. Across the country, Frelimo sent lists of key people who were to be registered first without being present and those comprised most of the first register books – those with the number ending -01. In EPC 3 de Fevereiro and EPC 4 de Outubro Frelimo won the -01 polling stations but Renamo won the rest. In EPC Josina Machel and EPC Samora Machel the highest Frelimo vote was in the -01 polling stations but Renamo still won. These lists often included teachers who were expected to vote Frelimo but in the -01 polling station of EPC 4 de Outubro there were 27 blank ballot papers, 6.2% of the total votes, which looks like a protest vote by unpaid teachers.