Zimbabwe's register is inaccurate and biased towards older citizens, a research group has found.
As Zimbabwe prepares for a presidential and parliamentary elections, a respected research organisation in Harare has found that the voters' roll contains a million people who are either dead or have left the country, 116 000 people over the age of 100, and 78 constituencies with more registered voters than adult residents.
The report by the Research and Advocacy Unit, which will be released on July 5, also notes that close to two million young Zimbabweans below the age of 30 are unregistered.
Young people tend to support the opposition parties.
The research unit's revelations follow Prime Minister Morgan Tsvangirai's challenge this week to the Zimbabwe Electoral Commission (ZEC) to investigate allegations that a shadowy Israeli firm, Nikuv International Projects, had been contracted by Zanu-PF to work on the voters' roll and help to rig elections.
Earlier this year the Mail & Guardian reported that there were suspicions that Nikuv was working with the country's intelligence to manipulate the voters' roll.
The M&G has also reported on Zanu-PF tactics to sway election results, including resettling landless people in selected Movement for Democratic Change strongholds in urban areas, such as Harare, to boost the number of Zanu-PF votes.
Zanu-PF also allegedly intends to ensure that soldiers and the police are mobilised in its favour.
In its report, titled Key Statistics Report from the June 2013 Voters' Roll, the research unit compares the roll with the preliminary results of the 2012 census and percentages of the population in different age bands supplied by government statistical agency Zimstat, which it used in its 2012 survey on health and demographics in Zimbabwe.
Slow population growth rate
According to the unit, Zimstat is of the view that these percentages have not changed significantly in the past few years and this is supported by the fact that last year's census shows that Zimbabwe has a slow population growth rate of 1.1%.
The unit's analysis revealed that only 8.9% of Zimbabweans under the age of 30 appear on the voters' roll, meaning that nearly two million people in that age band are not registered to vote.
Similarly, the 2012 population census indicated that there were 1.1-million people in the 20-to-24 age range, but only 22 5000 of them, or 19.6%, appear to be registered.
The report states: "In numerical terms, this means that a total of 1920424 people under the age of 30 ought to be registered as voters but are not. This is almost 29% of the total adult population of 6 647 779."
Coupled with this under-registration of young voters is a marked over-registration in all age groups of 30 years and above, indicating that the number of registered voters exceeds the population of Zimbabwe by a considerable margin.
In the 40-44 age group, for example, the census indicated that there were about 467 000 people, whereas 759 000 are registered to vote - a 62% variation. Some of this discrepancy might be explained by the fact that people were out of the country during the census, the unit said.
It also found that there were 116 000 registered voters over the age of 100, which it described as "unlikely".
The unit found that the number of registered voters exceeded the adult population in no less than 78 of Zimbabwe's 210 constituencies across the country. In Bulawayo, for example, seven constituencies - or more than half of the total - had more registered voters than adult residents.
In Mashonaland East, a mainly rural Zanu-PF stronghold, the unit revealed that 12 constituencies were over-registered.
It said the problems of "over-registration" and "under-registration" affected the delimitation of constituencies, "since it is based upon the number of registered voters in each constituency".
"Delimitation will be inaccurate to the extent that the voters' roll incorrectly reflects the number of voters in each constituency," the report states.
"Constituencies were last delimited ahead of the 2008 elections. There is to be no new delimitation ahead of the 2013 elections despite extensive demographic changes and changes to the voters roll since 2008."