Ghosts can register and vote without being seen and Gaza ghosts will have 6 seats in parliament and Inhambane ghosts 1 seat. But candidates for parliament must be real people who can be seen, so these 7 seats will be filled by real Frelimo members of parliament. But there has been some restraint in Gaza, which has 10 ghost seats in the current parliament and only 6 in the next one.
Because parliament has a fixed 250 seats, ghost seats must be taken from real people in other provinces. Sofala loses 2 seats and Maputo City, Tete, Zambezia, Cabo Delgado and Niassa lose 1 each.
The issue is not just ghosts; there is also a problem with real adults not being registered. We reported in March that Niassa and Sofala had only half the number of registration brigades as Gaza and Inhambane, so in Niassa and Sofala it is likely that many real voters in opposition areas were not registered. Brigades in Niassa and Sofala had to register about 3,300 adults per brigade, while in Gaza and Inhambane each brigade only had to register 1,700 voters. (https://bit.ly/Moz-El-220)
In Nampula we think there is both under-registration because of each brigade having to register 4,200 adults, and over-registration because the entire province has registered 103% of adults. Nampula's number of seats jumps from 45 currently to 49 in the new parliament.
Detailed data for Nampula is only available for the first 30 days of registration, plus last year. Of 8 districts with municipalities, 6 had already registered more people than they had voting age adults - that is, ghost voters. But in districts without municipalities, registration then was only 70%, suggesting many rural voters were not being reached. The Nampula data on https://bit.ly/Moz-El-248 and a table is on https://bit.ly/Moz-El-Nam-CIP
Calculating seats
The table below calculates the number of seats for each province, shows the differences from the two previous national elections.
There are 250 seats in the AR and 2 are assigned to Europe and Africa. Since the first election the law has said that the number of seats within Mozambique is calculated by first dividing the registration by 248 (67,135 for those registered so far), and then dividing the number registered in each province by that quotient. The method seems simple but it is wrong. As we see this gives 249 seats instead of 248. Every year the CNE, totally and secretly and never saying how it is done, changes the number of seats for one province to make the total come out to 248. This is never reported, but of course this bulletin does report. For all other seat assignments the Mozambican law uses the d'Hondt system, which is the most common correct way to assign seats. See https://bit.ly/Moz-El-dHondt
We then compare the number of provisional seats for this year to the past two national elections. The biggest changes are due to Gaza ghost voters - an increase of 8 seats in 2019 when the population did not increase, and now a cut of 4 this year because ghost voters were publicised. Nampula went from 47 seats in 2019 to 45 in 2019 due to seats stolen by Gaza, but now with its own ghost voters it has jumped to 49.
Estimating ghosts
The next table shows how we estimate ghost voters and the effect of excluding them.
The National Statistics Institute (INE) predicts the number of voting age adults, and it should be impossible to register more people than that. Yet in 7 of 11 provinces, there are more people registered than there are voting age adults. These extra people must be ghosts, because we cannot see them and census takers never found them. In Gaza 1 voter in 3 is a ghost.
Sign up for free AllAfrica Newsletters
Get the latest in African news delivered straight to your inbox
Mozambique has a high registration rate, in part because people want a voters card as an ID card, but it is rarely over 95%. So we estimate that no more than 95% of adults register, and everyone above that is a ghost. For two provinces, Maputo Cidade and Niassa, registration is below 95%, so we accept the registration number. For the other 9 provinces, we calculate 95% of adults and say the rest are ghosts. As the table shows, there are a maximum possible of 15,276, 785 real voters. And we exclude an incredible 941,031 ghosts - 6% of the total voters roll.
Eliminating the ghosts would totally change the composition of parliament. Gaza would drop from 18 to 12 seats, and lose the 6 seats of ghost voters. Inhambane would lose 1 seat. In exchange Sofala would gain 2 AR seats.
