Official results announced by the CNE on 27 October and approved by the CC on 22 December excluded 144,918 presidential votes and 144,934 parliamentary (Assembleia da República, AR) votes which are included in the tables published by the CNE and CC.
To add to the confusion, the voter registration is the same in the official results and table – only the votes have been excluded. The exclusion is nowhere mentioned or justified, but analysis of the tables suggests that the CNE excluded the votes, but not the registration, of Mozambicans in the diaspora.
To compound the confusions, after this Bulletin and AIM reported this error, the CC in secret changed its ruling (acórdão 25) for the Presidential vote, but not for the parliamentary vote. After this Bulletin reported the continued error with AR votes, the CC made a second secret change.
Mozambicans abroad can register and vote for president and parliament, with 1 AR seat for Africa and 1 for Europe. Registration totals include those in the diaspora, and there were 144,918 voters in the diaspora who, it appears, did not have their votes counted.
There seem two possible explanations. The first is simple error. At district and national level, vote tabulation is done in secret by the Technical Secretariat for Electoral Administration (Secretariado Técnico da Administração Eleitoral – STAE) and then accepted by election commissions, usually without much discussion. It had been pointed out to STAE that some of its earlier reports had included diaspora registration but not other diaspora data. Thus a STAE error may simply have been rubber-stamped by the CNE and CC because the high level of secrecy meant that parties, observers and journalists were never able to check the numbers before approval.
The alternative explanation is that registration of the diaspora in Africa was clearly inflated, as was registration in Gaza (see page 15). STAE or the CNE may have decided in secret that excluding the diaspora vote was a way of compensating for this form of ballot box stuffing. This is possible, because CNE President Sheik Abul Carimo told the Elections Bulletin that to control gross ballot box stuffing, district elections commissions had been instructed to exclude any polling stations with turnouts of over 100%, but this instruction was never published and no such exclusions were ever reported. This also occurred in secret in past elections.
STAE and CNE have always claimed that the elections laws allow them to change results in secret, without reporting the changes.
The Elections Bulletin (87, 27 Oct) did headline the excluded vote when it was announced by the CNE, but this seems not to have been noted by the CC, which also failed to note that its official results (page 31 of the original version of Acórdão 25) did not agree with the table (page 111 of Acórdão 25) in the 449 page annex to the ruling.
The original ruling, signed by the 7 constitutional council justices, was distributed at a public ceremony on 22 December and posted on the CC website. By early January a new version of the ruling, still called “Acórdão no, 25/CC/2019, de 22 de Dezembro” and with no indication of a change, had been posted with the tables of presidential results on page 31 corrected but not those on page 32 for the AR. A week later, a new Acórdão 25 appeared on the CC website changing page 32.
The original version of Acórdão 25 is posted on our website on http://bit.ly/CCAc25-Orig, the first secretly corrected version on http://bit.ly/CCAc25-Corr-1 and the second secretly corrected version on http://bit.ly/CCAc25-Corr-2
By including the registration of all voters but then excluding the votes of the diaspora – 144,918 votes – this reduced the turnout from 51.8% to 50.7%. It also took 131,623 votes away from Nyusi. The table on previous page compares the original “official” results to those in the annex of the CC 23 December ruling. All data and tables in this Election Bulletin final report use the numbers from the annex to the CC ruling and incorporate both the diaspora registration and its votes.
