Zimbabwe's dodgy voters' roll

Report by RW Johnson on preventing electoral fraud in the next elections

RW Johnson, Preventing Electoral Fraud in Zimbabwe: A Report on the Voters' Roll in Zimbabwe, South African Institute of Race Relations, June 6 2011

This Report on the Zimbabwean voters' roll as at October 2010 is based on the work of many dedicated Zimbabweans over many years, who have laboured mightily over the voters' roll in trying to acquire it, turning it into readily usable form, and then analysing it.  I would like to register my debt to all of them. 

Given past history, it will be understood that it was not easy to acquire a copy of the current roll.  All I would wish to say is that this was indeed achieved, and I am deeply grateful to those who made it possible. However, for the usual Zimbabwean reasons, I am unable to name the people involved without jeopardising their security.

The roll acquired was in digital ASCII (and thus computer-readable) format. It could thus be loaded into a database which also includes all the fruits of previous attempts to ascertain the truth about the voters' roll. The full details of the database are given in Appendix 1.

Comparing the voters' roll as of October 2010 with the roll used in the 2008 harmonised elections, we find that in 2010 there are 366,550 new voters who have not appeared on any previous roll.

This is extremely surprising considering that the overall population of Zimbabwe has been falling due both to a very high mortality rate and large-scale emigration. It might have been expected, nonetheless, that there would be some new voters in the youngest age group of roughly 18 to 25.

There is also a very unlikely total of 49,239 new voters over the age of 50 - and this when average life expectancy in Zimbabwe has fallen to 44.8 years.  Even more surprising is the fact that 16,033 of these new voters are over the age of 70 years, while 1,488 of them are over the age of 100. The complete list of this 100-plus age group of new voters is appended.[1]

Then again, a number of these new voters have no valid address, despite the stipulation requiring this.  In the Gwanda Central constituency alone, there are three people without a valid address, for example.

The age distribution of new voters is shown on the following table:

Table 1: The age distribution of new voters on the October 2010 roll

All told there are 228 new registered voters who are under age. Some of these under-age "voters" are actually small children,[2] as illustrated by the table below.

Table 2: Examples of under-age voters on the October 2010 roll

Further, it is important to point out that the 366,550 new voters who have been added to the roll as of October 2010 are by no means equally distributed around the country, as one might expect. In the extreme cases, one finds that the constituency of Cheredzi South has only 33 new voters added to its roll while the constituency of Gokwe Nembudziya has no less than 13,896 new voters added to its roll. There can be no satisfactory reason for such extraordinary variations. The overall picture of constituency variation in the distribution of new voters is shown in the map that follows.

Figure 1: Voters' roll increase per constituency, October 2010

One of the most striking anomalies is the number of exceptionally old people among the new voters.  There are, indeed, no less than 4,368 new voters over the age of 90 years on the voters' roll as at 1st October 2010.  If one amalgamates the list of new voters with the old list, one finds an extraordinary total of 132,540 persons over the age of 90 on the roll. Given the average life expectancy of less than half that figure, this is completely incredible.

In addition, we find that once again these nonagenarians are not evenly distributed among constituencies. Instead, they are again bunched into the same minority of constituencies which have had exceptionally high numbers of new voters added to them. This is shown on the table which follows.

Table 3: Distribution of new nonagenarian voters, October 2010

Figure 2:  Distribution of new voters over the age of 90, October 2010

As may be seen, there is a particular concentration of nonagenarians in the three Mashonaland provinces - the Zanu-PF heartland - and in a few selected key seats elsewhere.   In Chipinge East there are no less than 662 of these nonagenarians.  All told, there are 4,368 nonagenarians among the new voters - an almost impossible number. 

It is worth pointing out that there are also 1,425 voters on the roll whose gender is, oddly, declared to be undetermined - denoted with a "U".  Among these new voters of uncertain gender, there are no less than 171 voters under the age of 18 - who, accordingly, should not be on the roll at all.

There are also 1,235 new voters whose address is given as *** Box -.  Their inclusion defies the requirement that addresses must be given.  Even more spectacularly, of these 1,235 people, no less than 1,160 are more than 100 years old.

There are thus a considerable number of centenarians among the new voters.   But this astonishing fact pales into insignificance when the total voters' roll of October 2010, incorporating both "old" and "new" voters, is examined.  Thus, in the single constituency

of Mount Darwin East, we find 118 registered voters over 100 years old, the majority of whom have their date of birth recorded as 1st January 1901. This means they are all more than 110 years old!  Another nine have birth-dates of 1st January 1905, and 25 are given as having been born on 1st January 1910.[3]    

Overall, the full voters' roll shows no less than 41,119 centenarians.  This is an impossible figure.  The United Kingdom (UK), with a total population of over 60 million and an average life expectancy more than 30 years longer than Zimbabwe's, has only 10,000 centenarians. 

The data on the roll also shows that only 19,626 of these centenarians are women, while 437 are of uncertain gender (!) and 21,056 are men.  This is a further absurdity, for the roll thus reveals a large masculine majority among the centenarians, even though women in Zimbabwe, as elsewhere in the world, have a longer life expectancy than men. A male majority among the centenarians is thus an impossibility.  An additional impossibility is a male majority among the centenarians purely among the new voters.

Finally, there are no less than 16,828 registered voters with the same date of birth, given as 1st January 1901. It might be argued that the enumerator simply gave this birth date to all very senior citizens who were in doubt as to their true age - though that already suggests an impermissible degree of official intervention in the registration process. However, if one's credulity is stretched by this extraordinary number of 110-year-olds, it is stretched way beyond breaking point when one learns that no less than 1,101 of these 110-year-olds are registered in Mr Mugabe's birthplace, Zvimba, presumably to act as a reserve category capable of producing particularly pleasing results for Zanu-PF there.

Equally incredible is the fact that the full roll of both "old" and "new" voters includes 624,794 voters over the age of 70, who make up more than 10.6% of the total.  The fact that the entire roll now includes 702 under-age voters (some of them "old" and some of them "new") seems trivial by comparison, though there can be no good explanation for including those whose given ages clearly disqualify them.  

There are other anomalies too. For instance, 1,744 voters are registered in the Harare North constituency, with their addresses given as Kadungure, Fast Track, Chimurenga or other similar housing co-operatives within the Hatfield area.  All told there are 18,525 registered voters with a physical address of "housing co-operative" of some sort.  Yet this is not an address.  To be a member of a housing co-operative is simply to have an associational membership which may, one day, result in housing.  It does not fulfil the requirement of a proper address.

In addition, it is disturbing to note that there are people who are still on the roll even though they are known to have been killed in the violent struggles of recent times.  The roll also includes the names of people long known to be dead:

Table 4: Examples of voters known to be deceased on the October 2010 roll

To conclude, then, the Zimbabwe voters' roll, as at October 2010, is not only a wholly incredible document but an extremely dangerous one, which lends itself to all manner of electoral manipulation or ballot-stuffing. It is more or less guaranteed to produce disputed results.  It is beyond redemption and cannot even be used as one of the building blocks in the construction of a new and authoritative voters' roll.  It simply has to be scrapped completely, while work on a proper roll must begin again from scratch.

How has this happened?

The 2008 roll recorded the names of 5,727,902 registered voters.  Yet, with some four million or more Zimbabweans having fled the country to avoid the excesses of Mr Mugabe's rule, no one was then sure what the current population was. The only certainty was that it had fallen a great deal.  Probe, the Harare-based subsidiary of the Gallup survey organisation, estimated that the population was then anywhere between eight and ten million people.

This degree of uncertainty was remarkable, especially on the part of a survey organisation whose routine job it was to conduct surveys representative of the population. This is not to criticise Probe: it merely illustrates how distressed the current situation in Zimbabwe was and is, with huge, unmeasured population flows out of the country. These flows have continued unremittingly since 2008.

Let us, however, concentrate on those 2008 figures. If one assumes, simply to err on the side of prudence, the higher figure of a population of ten million, one must then factor in the fact that at any given moment some 60% of Zimbabweans are under the voting age of 18, leaving one with a maximum electorate of four million. In addition, all previous surveys have shown that the proportion of those who have registered to vote has never exceeded 80% of those eligible to do so. Usually, the proportion is much lower than that. Again, however, let us be prudent and use the higher figure of 80%. 

That still leaves us with a maximum registration of some 3.2 million eligible voters, while the real figure might well be less.  This indicates that the 2008 register of 5,727,902 voters had more than 2.5 million too many people on it. It goes without saying that, by thus providing the authorities with a reservoir of over 2.5 million fictitious voters, the 2008 roll made electoral manipulation easy. Such fictitious votes could be added to totals wherever Zanu-PF was vulnerable.

It should also be remembered that the figure of 3.2 million voters was a maximum. As noted, most surveys show a registration rate of well under 80%.  Besides, in many Zanu-PF areas, it was very difficult for opposition voters to register, as the local (pro-Zanu-PF) chiefs or headmen were not willing to vouch for their place of residence, as is required.

Dimly, at least, one can descry how this situation has been produced.  In the 2002 presidential election the roll had 5,229,538 voters on it - at least two million too many. In 2008, after a period in which millions of Zimbabweans had fled the country, the roll increased by half a million - a quite impossible result - to 5,727,902 voters.  On the data available, this was achieved by removing about 676,900 voters from the roll (presumably as deceased) but adding roughly another 1,100,660, which was almost twice as many.

Such figures could only have been achieved if Zimbabwe's population had been growing quite explosively, whereas quite the opposite had in fact been happening.  Similarly, in 2010 the roll has increased again to 5,867,643. Again, the data suggests that roughly 366,550 voters have been added while around 127,000 have been removed. Again, this 3:1 ratio of new to removed voters could only have been achieved in a very rapidly growing population, whereas in fact the population had once more been shrinking.

The current situation would thus seem to be the product of two distinct processes.  First, a number of frankly incredible names - of people who are very old, under-age, address-less or un-gendered - have been added to the roll.  It is difficult to imagine any motive other than ultimate electoral manipulation behind the addition of such "phantom" voters. 

However, to this must be added the simple fact that, year in and year out, voters are not being removed from the roll in numbers large enough to reflect Zimbabwe's extremely high mortality and emigration rates. Over time this has created a vast backlog of equally "phantom" names of voters who must have passed away or emigrated long ago.  This is the most charitable explanation for the vast numbers of very old voters on the roll. However, the fact that new centenarians are among the voters added to the October 2010 roll suggests that some (perhaps many) of these very old voters have never existed at all.

The Zimbabwe Electoral Commission (ZEC)

The problems with the Zimbabwe Electoral Commission (ZEC) start with the fact that, although it is supposedly an impartial and independent body, it has presided over many defective voters' rolls and has never found fault with them. Not only is the October 2010 roll an incredible document but Zimbabwe has been operating for many years with voters' rolls which simply defy rational belief.

Even now, the ZEC's recent glib instruction to the Registrar-General to "fix" the voters' roll shows only too well that the ZEC has not yet understood how calamitous the situation is.  Unless there is a clean and authoritative voters' roll, there cannot be authoritative and democratically accepted elections.

In fact, of course, the problems of the ZEC go much deeper.  Just as the Registrar-General is a self-acknowledged Zanu-PF diehard, so too is the ZEC clearly unbalanced and partisan. It is an instrument of Zanu-PF rule, and a simple reflection of Zanu-PF's domination of all public institutions.  This has been evident in the clear and consistent bias the ZEC has shown towards Zanu-PF throughout the bitterly-fought elections Zimbabwe has witnessed since 2000.  Simply in order not to be otiose, we take as an example merely the most recent poll: the 2008 harmonised elections.

In the presidential run-off election in June that year, violence was endemic. It reached such heights that Mr Tsvangirai withdrew as MDC candidate rather than expose his followers to any further violence. Even Mr Mugabe himself acknowledged the high level of violence (see The Zimbabwe Herald 22 February 2011).

It should be remembered that most reputable electoral monitoring bodies such as the European Union and the Commonwealth were not allowed to send in election observers. Only observers from organisations Mr Mugabe saw as friendly to his cause were allowed into the country.  Despite that, five of the six international monitoring bodies observing the election reported extensive violence.  The Pan African Parliament's Observer Mission expressed the majority opinion of these observers in its statement that violence "had prevented the conduct of free, fair and credible elections".

Yet the ZEC, in its statutory report on these elections, studiously ignored this violence, its sole reference to the matter being: "According to the police, save for some parts of Mashonaland Central and Mashonaland East provinces, where some incidents of inter-party violence were reported, the rest of the country was generally peaceful."

The Zimbabwean Constitution says that the ZEC's mandate is to ensure that "elections and referendums are conducted efficiently, freely, fairly, transparently and in accordance with the law".  In 2008 the ZEC asserted that all these conditions had been met. This was an extraordinary statement in the light of reports by five out of six monitoring bodies to the contrary. It also ignored so great a marring of the electoral process that the leading candidate from the first round of the presidential election was effectively forced out of contention in the second round.

In addition, the ZEC took a very long time to deliver the 2008 election results, even though all observers and independent journalists were reporting the results within a few days. Undoubtedly the reason for this delay was that the ZEC was mirroring the uncertainty within Zanu-PF as to how to deal with this new situation. This means that, to all intents and purposes, the ZEC was thus part of the armature of the Zanu-PF State and was incapable of playing its role as an independent and impartial body. 

No doubt the Southern African Development Community (SADC) had this in mind when the Global Political Agreement (GPA) was signed, for the GPA required an entirely new ZEC to be constituted.  This signalled a de facto acceptance that no ZEC members who had signed the commission's report on the 2008 elections were suitable for any future ZEC. Yet these provisions of the GPA have also been ignored, as the following table on the composition of the ZEC reveals.

Table 5: The composition of the Zimbabwe Electoral Commission




Justice Chiweshe

Justice Simpson Mutambanengwe


Mrs Joyce L.  Kazembe

Mrs Joyce L.  Kazembe

Deputy Chairperson

T.P. Gambe

Theophilus Gambe


Prof.  G P  Kahari

Prof. Geoff Felfoe


V  S  Ncube

John  Chigaru


S L Kachingwe

Dr. Petty  Makoni


Bishop  J  Siyachitema

Mrs.  Sibongile  Ndlovu



Mrs. Bessie F Nhandara



Mr. Mkhululi  Nyathi


L C  Sekeramayi

L C  Sekeramayi

Chief Elections Officer

It is disturbing in the extreme to find that the new ZEC's Chief Elections Officer, along with two other members of the new ZEC, were indeed members of the ZEC which drew up the disgraceful report on the 2008 elections referred to above. This is precisely the carry-over of tainted personnel which the GPA sought to avoid and ought to avoid.

There also needs to be far greater transparency in the way the ZEC conducts its business and its electoral administration in general. During the parliamentary elections of March 2008, the ZEC seemed to be paralysed by the prospect of a possible MDC victory and simply clammed up. Rumours of every kind swept the country because the ZEC had failed in its task to provide authoritative information as speedily as possible. 

Then, when the parliamentary results were announced, they were released at a snail's pace and over a day or more. Further, each MDC victory was accompanied by a partnering Zanu-PF victory, apparently with the intention of creating the impression of a virtually tied race.

Yet the fact - as must have been obvious from early on - was that the MDC had won a decisive victory.  A similar attempt to mask the results of the presidential first round merely undermined the credibility of the ZEC further still.

The Background: Gaining Access to the Voters' Roll

For free and fair elections to take place (at last) in Zimbabwe, it is essential that the ZEC be constituted impartially, that it operate independently of all parties, and that it be effective. Similarly, it is vital that an accurate voters' roll be drawn up at last. This must be free from the partisan control and manipulation of the past, have credibility for all political parties, and be easily accessible to the general public.

This last point is worth stressing. In the past the voters' roll has been shrouded in a habitual secrecy as if it were a grave matter of national security. Attempts to gain access to it have been treated with contempt, and every possible obstruction has been placed in the path of those attempting to exercise what ought to be their normal citizen rights to full and easy access to the roll.

 his Report is unique in that it is based on a full examination of the actual voters' roll as it stood in October 2010. Hitherto, the only information available on the current roll was that provided in January 2011 by the Zimbabwe Election Support Network (ZESN), a coalition of non-governmental organisations (NGOs). But the ZESN report[4] was based on a sample survey of a mere 513 people from a "hard" (printed) copy of the roll.

For this Report, by contrast, we have worked on the full roll in digital format, which has enabled a far more comprehensive and authoritative evaluation.

When the ZESN Report was released, it was admitted on all sides - including by Mr Tobaiwa Mudede, the Registrar-General - that the roll was "in a shambles". The ZEC called upon Mr Mudede to explain, and then gave him three months to "fix" the roll. This expectation was ridiculous, in terms of both the time and the personnel required. Quite clearly, an entirely new roll needed to be drawn up, a task not easily performed in three months. In addition, the last person to whom such a task could be entrusted was the very man seemingly responsible for various defective rolls in the past.

In addition, reports from a Pretoria company, South Africa Waymark, which attended a meeting with the ZEC on the matter, provided insight into how the Registrar-General in fact proposed to prepare a "new" roll. His plan was simply to use the existing roll, dating from 2008, as a base - and then to transfer voters on to a "new" roll on the basis of simple self-identification. All this would require was that all new voters should provide their place of residence, ID number, and date of birth.

However, this method of proceeding is not only illegal but openly flouts the requirement, stipulated by both the SADC and the Zimbabwean Parliament, that a proper new voters' roll must be drawn up. Instead, it accepts the notoriously inaccurate 2008 register as the base document and merely adds new voters to it. This method also means, of course - and in defiance of the GPA - that the whole matter of compiling the roll is handed back to Mr Mudede.

Parliament's views

The question of the voters' roll was debated in Zimbabwe's Parliament on 17th November 2009, following the publication of a report on the matter by an NGO, the Research and Advocacy Unit (RAU).[5]  The Parliamentary debate concluded that a new biometric roll was necessary in Zimbabwe if free and fair elections were to be conducted in terms of the GPA. 

Biometrics is the technical term for methods of uniquely recognising human individuals. It is most commonly used in computer surveillance and access control systems, and typically uses both physiological and behavioural traits. Physiological traits include such things as fingerprints, facial recognition, palm prints, DNA, and iris recognition (which has generally replaced retina recognition). Behavioural traits include such phenomena as a person's gait or voice. Nowadays, the technology required to service such a system can be packed into a satchel and is easily carried around. Zambia has adopted a biometric voters' roll and is receiving aid from the United Nations Development Programme (UNDP) to provide the technology.

In the Zimbabwean case, there was overwhelming agreement from both sides of Parliament, including Zanu-PF, that such a system would be best, so as to end disputes over the voters' roll once and for all. It was felt that at a minimum the roll should record a person's name, address, ID number, fingerprints, photo, and date of birth. Unfortunately, however, while such sentiments enjoyed overwhelming favour on all sides of the legislature, Parliament did not proceed to enact legislation stipulating such requirements.

The Zimbabwe Election Support Network (ZESN) thereafter purchased a "hard" (printed) copy of the voters' roll and began conducting some research into it.  A hard copy was far from ideal, but this was the only form in which the Registrar-General would release a copy of the roll at all. Instead of being able to obtain a digital copy which would have been easily amenable to analysis, the ZESN thus had to spend US $30,000 on the purchase of a printed copy so long and bulky that it weighed one-and-a-half tons.

In these circumstances, analysis of the roll was a labour of Sisyphus, as Mr Mudede doubtless intended. Even so, evaluation of the roll immediately brought various problems to light.  Meanwhile, press statements from the ZEC not only acknowledged the chaotic state of the voters' roll but also revealed the ZEC's intention to task the Registrar-General with "fixing" the current roll. This clearly flouted Parliament's wish for a new biometric voters' roll.

The ZEC's decision was also contentious in other ways. It was disconcerting that Mr Mudede, author of previous manipulated rolls, should remain in office when the GPA had stipulated that all personnel earlier involved in electoral administration should be replaced. Instead, Mr Mudede, who turned 68 on 22nd December 2010, has been kept in office well beyond retirement age. This reflects a clear decision by the Mugabe Government to maintain its apparatchik in this key position.  For Mr Mudede is an outspoken Zanu-PF loyalist who, prior to the 2008 elections, announced that he could never envisage declaring anyone but Mr Mugabe the winner of the presidential poll.  This means that he is the opposite of a properly impartial civil servant.

On top of all this, the Registrar-General has a long-standing record of attempting to prevent public access to the voters' roll. For example:

  • 1995    Ms Margaret Dongo, an Independent MP, successfully sued the Registrar-General for failing to supply her with a copy of the voters' roll.[6]  However, all that Ms Dongo finally obtained was a copy of the register for her own constituency - one out of 120 constituencies.
  • 2000    The MDC applied to see the roll for the constitutional referendum and ensuing general election. Its application was refused by the Registrar-General.
  • January 2002     After the MDC had made many High Court applications to obtain the voters' roll, a court order was eventually granted and the roll, as at January 2002, was supplied in digital format by the Registrar-General's office.
  • March 2002      However, the MDC suspected that this roll was not the same as the one to be used in the pending presidential election of March 2002, and the party thus applied for a digital copy of the updated roll. This application was refused and the election went ahead with the Opposition deliberately disadvantaged by not having a copy of the roll.
  • March 2002     A hard (printed) copy of the roll for four constituencies, as used in the 2002 presidential elections, was purchased from the office of the Registrar-General by an individual who was in fact acting for the MDC. However, the Registrar-General quickly realised what had happened. Hence, when additional copies were requested, all further sales of such material were blocked by him.
  • April 2002     A High Court application by the MDC to have the roll, as used in the 2002 presidential election, released in digital format was rejected. [7]
  • The MDC's application to acquire the presidential voters' roll then went to the Supreme Court, and was ultimately denied in a poorly reasoned judgment. Mr Justice J Friedman, acting as an Independent Observer on behalf of the Forum for Barristers and Advocates of the International Bar Association, found "the judgment of the Supreme Court...fundamentally flawed", adding: "The outcome of the appeal can therefore....only be described as highly unsatisfactory." [8]
  • February 2005    An application by the MDC for the voters' roll in digital format for the imminent 2005 general election was denied.
  • These successive efforts created considerable bad publicity and embarrassment for the Zanu-PF regime. Hence, in early 2008 the law was changed to allow a voters' roll to be supplied in digital format to political parties and interested NGOs prior to an election. This proved of great significance when Zimbabwe in 2008 staged "harmonised" elections in which voters cast their votes at the same time for Parliament, the President, the Senate, and their local municipal council.
  • March 2008   The voters' roll was supplied to political parties, though not in the readable digital format specified.  In addition, the roll for some constituencies was delivered only after the election was over.  The experience of the Zimbabwe Electoral Support Network (ZESN) was typical, for instead of being supplied with computer disks with the data in analysable form, the ZESN received - only well after the elections had taken place - a set of 210 CDs of the roll in "tiff" (tagged image file format). This meant that the CDs contained a photographic reproduction of each page of the roll, a format not easily susceptible to analysis.  It must have been a large and deliberate exercise to photograph every page of the roll in that way. It would certainly have been far easier for the Registrar-General to supply the information in the digital form which his office already used.  There seems little doubt that the law was thus flouted, precisely with the objective of preventing timeous analysis of the voters' roll.
  • 2009    An MDC application to obtain a copy of the voters' roll as used in the 2008 harmonised elections was denied by the ZEC.

It is interesting to note that the definition of "digital", in the "electoral reforms" being negotiated as part of the Road Map process, has been refined to "digital and readable". This is only what one would expect in a democracy. However, as is evident from the above history, it is precisely this which has consistently been withheld by the Registrar-General. The fact, of course, is that there is a great deal about the voters' roll which the authorities have wished to hide.

A History of Roll Reports and Audits

Despite these great difficulties, both the main opposition party, the MDC (Tsvangirai), and members of concerned citizens' groups have struggled with might and main to obtain the roll, turn it into analysable form and publicise its defects in the hope of exerting pressure for reform.

Immediately after the 2002 presidential election, four hard copies of the voters' roll used in the election were purchased from the Registrar-General's office by an intermediary acting on behalf of the MDC, and an analysis was carried out.[9] As noted, the purchase of additional copies of the roll was prohibited by the Registrar-General.  Despite the damning findings that emerged from this analysis, no response was forthcoming from the ZEC.

A request by the MDC for a digital copy of the voters' roll, as used in the 2002 presidential election, was denied. However, after a protracted legal battle, the MDC managed to obtain access to the 2002 election residue (all the paperwork left over at the conclusion of the election, such as voting slips, registers used by polling station clerks, and so on). This was then subjected to an analysis which brought many irregularities to light, including multiple voting. A 500-page report, including a full audit of the residue, was then submitted to the High Court and to the ZEC - again without response.

Prior to the 2005 general election, applications by the MDC to obtain a copy of the voters' roll were denied by the Registrar-General. An application was then made to the High Court to force Mr Mudede to provide a copy. Only after the election did the MDC obtain a court order instructing the Registrar-General to sell a copy of the roll to the MDC. Mr Mudede complied by providing a bulky printed copy (rather than a digital one) which made analysis extremely difficult. However, the MDC's team of 20 people proved equal to the mammoth task by using a complicated OCR (Optical Character Recognition) technique. (OCR is the science of scanning the written word into computer-readable type.) When this process finally translated the data into a digitally analysable form, a full report was drawn up detailing multiple problems with the roll. The report, together with covering letters from Mr Tsvangirai and a Zimbabwe human rights activist, Mr R Whitehead, was sent to the ZEC. However, no response was forthcoming.[10]   

The printed voters' roll supplied to the MDC by the ZEC via the Registrar-General for the 2008 harmonised elections was finally also converted into digital format. The analysis thus made possible revealed massive discrepancies, which were identified in a report published by the Research and Advocacy Unit (RAU), an NGO.[11]  

The Zimbabwe Electoral Support Network has also published a report based on a printed copy of the roll, which it purchased in 2010.[12]  To date, there has been no official response to this report.

On 17th November 2009 the issue of the voters' roll was debated in Parliament. Opposition parties used data from the RAU Report to substantiate claims of gross irregularities.  Interestingly, Zanu-PF's representatives made little or no effort to dispute these claims. Instead, there was general agreement on both sides that a new biometric voters' roll was imperative for a democratic Zimbabwe.[13]  Unfortunately, however, the Zimbabwean Parliament did not use the broad majority in support of a new biometric roll to incorporate such a requirement into legislation. Instead, it simply expressed its wish for such a voters' roll.

This Report is thus but the latest in a very long line of efforts to analyse the voters' roll and its imperfections. All these earlier endeavours to obtain and analyse the voters' rolls used in successive elections have yielded essentially the same outcome. Each evaluation has shown that the voters' roll was in a chaotic shambles, that there were enormous numbers of phantom voters, that free and fair elections were impossible in such circumstances - and that the Registrar-General was largely to blame.

It is not unduly cynical to suggest that the reason there has been no official response to any of these reports is that it has suited the ruling Zanu-PF regime only too well to maintain the voters' roll in this condition, as this has facilitated official manipulation of election results. There have been many universal suffrage polls in Zimbabwe since independence in 1980, but it seems that the country has yet to experience a truly free and fair election.

However, one of the peculiarities of Zimbabwe is that an authoritarian Zanu-PF regime has long co-existed with the general framework of a liberal democracy. Hence, the regime continually attempts to legitimate its rule via elections. In most countries with that type of regime, only a cursory regard, if that, is normally given to elections. But in Zimbabwe, this peculiar mix of political traditions - a reflection of Zimbabwe's unique political history - gives the regime the need to maintain a democratic façade.

However, the regime also needs to guarantee election results and so it resorts to all possible manner of manipulation, violence, and intimidation. Since it nevertheless wants to maintain a democratic façade, this means that the regime is capable of embarrassment when these undemocratic stratagems are too openly revealed.  Zimbabwe also needs to retain the support of its neighbours in the 15-member Southern African Development Community (SADC) and of the wider African Union (AU). Hence, it has repeatedly insisted that its elections are free, fair, and democratic, thus giving the Zanu-PF Government the legitimacy on which it insists. But reports revealing the manipulation of the voters' roll suggest the opposite - and this has put pressure on the regime to concede the need for reform.

Hence, once both sides in Parliament had agreed in November 2009 on the need for a new biometric voters' roll, it might have been expected that the ZEC and the Registrar-General would have hastened to provide one. In addition, a Pretoria-based firm, South Africa Waymark, which has experience in many SADC countries, had already submitted a detailed technical quotation to conduct a wholesale re-registration of voters in Zimbabwe and prepare a biometric voters' roll within a 90-day time frame from the awarding of a contract. The price it quoted was US $21 million.

However, Waymark's quotation was dismissed out of hand by the ZEC (although no one doubted that donor funds would have been made available to pay the bill). Early in 2011, the managing director of Waymark, Mr. Pikie Monaheng, thus visited the ZEC and again suggested that it consider the creation of a biometric roll on the lines of Waymark's earlier quotation. This proposal was rebuffed, for the ZEC had by then decided to give the Registrar-General three months to "fix" the roll. This approach by the ZEC is diametrically contrary both to the GPA and to Parliament's express wish, as made clear on 17th November 2009.  

The Delimitation of Constituencies

In 2008 the number of constituencies was expanded from 120 to 210, thus compensating for the abolition of the President's right to appoint a number of MPs of his own. This expansion of the elected seats required a new delimitation of constituencies. The details of this delimitation were included in a report presented to Mr Mugabe on 21st January 2008.  However, this new delineation of boundaries, a process pregnant with possibilities for gerrymandering, was carried out solely by the State and without any consultation with the opposition parties. 

Moreover, the delimitation report was made available to the public only in printed form. Yet the parties and the Media, along with voter education bodies, needed to have access to the data and accompanying maps in digital form if they were to analyse and comprehend it properly.  In other SADC countries, by contrast, mapping information and other accompanying data of this sort is readily available at minimal cost.  This should be seen as a basic democratic requirement.

The Diaspora 

Zimbabwe faces an exceptional challenge to its democratic legitimacy in the form of its enormous diaspora.  No exact figures exist, but the estimate most often used is that some four million of its citizens have fled abroad since the eruption of the country's political crisis in 2000.  The vast bulk of these exiles are in South Africa, though there are also substantial Zimbabwean exile communities in the UK, Australia, and elsewhere.  It is very difficult to find other examples of countries where such substantial proportions of the national community have been forced to flee.

The advent of democracy in Zimbabwe in 1980 was supposed to bring about the final return of the large exile population that had accumulated in Mozambique, Tanzania and elsewhere during the liberation struggle.  It was also seen as crucial to the new democracy that all these exiles should at last be able to exercise their right to vote in the 1980 poll.  Yet now, thirty years after independence, a far larger exile group has been created: a most unhappy historical achievement for a young country.

The need is now more pressing than ever to find a way of including Zimbabwe's exiles in the country's democratic process.  Naturally, one hopes that this would be part of a transition which would encourage most or all of those exiles to return home.

Probably the best example to follow would be that of South Africa. In the first democratic election there in 1994, the whole motif of the poll was that it should be inclusive and great efforts were made to allow exiles to vote. To this end, the many hundreds of thousands of residents who were not citizens but who enjoyed permanent residence status were allowed to vote. This was a one-election-only privilege and thereafter citizens alone were permitted to vote.  Nonetheless, agitation continued to allow at least some of the South African diaspora to vote. In 2009 their right to do so was finally upheld by the Constitutional Court in Johannesburg.

Many other countries also allow such voting as a matter of course. Russia, for example, allows any of its citizens who are abroad to vote in Russian elections by going to the nearest Russian consulate or embassy to cast their vote.

Zimbabwe should clearly attempt to follow this model. It would, however, be essential to conduct some preceding registration exercise at foreign embassies and consulates. Here, exiles could prove their status by displaying their passports or IDs, and could also attempt to establish their links to particular constituencies.  Ideally, as in South Africa in 1994, this would be a somewhat exceptional case: itself part of a large-scale homewards return so that in succeeding elections the number of such diaspora voters would diminish. There is, after all, no need to enfranchise those who have chosen to reside permanently in other countries.

The founding election of democratic Zimbabwe in 1980 was badly flawed and there have been no truly free and fair elections since then. What is now needed is nothing less than a new founding election, this time one which inaugurates an era of truly free and fair elections. In these exceptional circumstances, provision must be made for members of Zimbabwe's extensive diaspora to cast their vote. It is understood that many within the diaspora insist that they will not return until Zimbabwe is a democracy, but once truly free and fair elections have taken place more restrictive voting regulations for Zimbabweans temporarily abroad would apply.


1) The ZEC must be reconstituted without any members of the preceding, discredited commission.  In addition the greatest care must be taken that only independent and impartial citizens are appointed to the ZEC.

2) A new and independent Registrar-General should be appointed.

3) An impartial and international body should be employed to draw up a new biometric voters' roll.

4) Once drawn up, the new voters' roll must be made accessible to all citizens as quickly and cheaply as possible.

5) Thereafter, each new version of the voters' roll should be audited by a professional body to ensure, by means of spot checks, that deceased voters are being properly removed from the roll and that no barriers exist to the registration of legitimate new voters.

6) The audit report of this professional body should be transmitted to Parliament for debate and discussion.

7) Since the ZEC still has not fulfilled several key elements of its mandate, it must also, at least on its website:

(a) set out the legal framework for the conduct of elections;    

(b) define the roles and powers of election observers;

(c) provide full registration statistics;

(d) issue full reports on each election; and

(e) make provision - as is now the norm in many countries - for citizens to check their registration status by internet and SMS.

8) Provision must be made for members of the diaspora to vote, along the lines indicated above.

Conclusion:  The Difficult Birth of Zimbabwean Democracy

The World Bank and the International Monetary Fund (IMF) have found that when they are called in to help countries which have fallen into acute economic difficulties, sometimes their prescriptions work well and on other occasions they don't at all. When such cases were examined it was found that the crucial difference was that in some countries their recommendations were enthusiastically followed by governments who saw them as the way out of a mess, but in other countries (and Zimbabwe was such a case) the government was essentially hostile to their recommendations and was grudging in carrying them out. In the first case things worked well, in the latter badly or not at all.

It is easy to feel that making recommendations on how to improve democracy is the same - that this will work when governments want them to but not when they seek to undermine the recommended measures. It is an easy step from this to conclude that trying to improve democracy in Zimbabwe is a chicken-and-egg situation; that it cannot really be done until a properly democratic government is in power; and that the Zanu-PF regime will always frustrate the chances of such a government coming to power. If that is a logically watertight proposition then making such recommendations - this Report included - is just a waste of time.

It is true that no one ever lost money by betting on Zanu-PF's determination to retain power but such a perspective misses the important historical nuances in Zimbabwe's situation. The present Government came to power in the 1980 election, and it was important to it to insist that this gave it genuine democratic authority. Thus it has held regular elections ever since and Mr Mugabe's singular attempt to install one-party rule and do away with elections was heavily voted down by his own party. For the fact is that it has become increasingly necessary to national self-respect for African governments to point to their democratic legitimacy. The era of military coups is over and however weak and vacillating the AU may be in other directions, it is now tough and unequivocal in condemning each and every military coup and demanding immediate democratic elections. This is also the attitude of (most of) the Great Powers and all hope of donor aid depends on at least a semblance of democracy.

In Zimbabwe's case, such forces have transformed the neighbourhood. When Zanu-PF came to power, Zambia and Malawi were one-party states, Namibia and South Africa were under white minority control, and there had been no multi-party elections in Angola or Mozambique. Today everything has changed and all Zimbabwe's SADC neighbours are at least nominal democracies.

This has made a vast difference. It was in 1989-90 that Mr Mugabe made his proposal for a one-party state. It was heavily defeated even then and it would be unthinkable to propose it now. Those days have simply long gone.  So far has the tide gone out on such things that it is startling to realise that the man who proposed a one-party state in Zimbabwe is still in power when the other local architects of one-party states - Hastings Banda, Kenneth Kaunda, Samora Machel and Agostinho Neto - have long departed the scene.

Mr Mugabe himself is now 87. It is important to realise that while he may remembered as a man who for long acted as Canute, that is all he can be. The democratic tide is coming in and it is a matter of time before Zimbabwe joins its neighbours in this development.

Moreover, this development seems far more certain in Zimbabwe - with its pre-Mugabe tradition of multi-partyism, its strong opposition and vigorous civil society - than it ever did in most of the surrounding countries.

In addition, even under Mr Mugabe, there has been movement. In the early days opposition parties were hardly tolerated at all and even when the MDC was formed many voices within Zanu-PF - Mr Didymus Mutasa was a leading example - argued that all differences and questions should have remained within Zanu-PF and there should never have been a separate opposition party. Gradually, however, Zanu-PF has had to acknowledge that the Opposition exists, that it has legitimacy, and even that it should provide a Vice President and Prime Minister.

Similarly, at first extreme measures were taken to kill press freedom - including the blowing up of the Daily News press - but even Mr Mugabe has had to learn to live with an influential independent Media. Equally, the President's ability to counter the effect of democratic elections by nominating 30 MPs of his own has at last been abolished.

In the 2008 elections it was for the first time required that each polling station publicly post up the exact results of the count at that station, a procedure which made electoral fraud considerably more difficult. All these are real and important advances. Bit by bit, the Zanu-PF regime has been embarrassed by the revelation of irregularities and has been nudged towards reform. The same is true of the voters' roll. Persistent pressure has led to at least some change in the ZEC and, as this Report shows, in the parliamentary debate on the voters' roll even Zanu-PF MPs agreed with the need for a new, clean, and biometric roll.

So, bit by bit - every yard is hard fought - the struggle for democracy in Zimbabwe has continued to gain ground. Today Mr Mugabe is under acute pressure from his SADC neighbours to give still more ground and the same pressure is felt from the United Nations, major donors and the Great Powers. Quite likely what we will see is an inch-by-inch reform which one day gives way to an overwhelming glasnost and perestroika: that is how history works. In the meantime, nothing is more crucial than that the struggle for full democracy should continue. There are many brave people in Zimbabwe who have dedicated their lives to achieving that full democracy. The author of this Report is deeply conscious of their courage: without it, this Report could not have been written.

South Africa was founded as a country in 1910 but it long remained the sort of country where most people cheered for foreign sporting teams against their own, so alienated did they feel. In that era there was the false start of the Bantustans when we were invited to cheer for the independence of bogus "black homelands", artificial creations which could never bring freedom to the majority.

Then, thanks in the end to the same people who had inflicted apartheid on the country, democracy was introduced and in 1994 we (I speak as a voter in that election) experienced our first fully democratic election, which amounted to nothing less than a re-founding of South Africa. Zimbabwe, our immediate neighbour to the north, is very much the same. It has known false dawns, in 1965 and again in 1980 - but real freedom and proper democracy still lie ahead. This Report is intended as a contribution towards that real dawn, still to come.

Issued by the South African Institute of Race Relations, June 6 2011. The full report, including the preface and appendices can be found here - PDF.

Click here to sign up to receive our free daily headline email newsletter




[1]  For new voters over 100 years old on the voters' roll in October 2010, see Appendix 2.

[2] The full list of registered under-age voters is to be found in Appendix 4.

[3]  For list of centenarians with same date of birth in Mt Darwin East, see Appendix 3.

[4]  ZESN, A Report on a Voters' Roll Observation Conducted in Zimbabwe (Harare), cyclostyled, January 2011

[5] RAU Report (by Derek Matyszak),  2013 Vision, Seeing Double and the Dead: A Preliminary Audit of Zimbabwe's Voters' Roll (Harare, cyclostyled, 2009) 

[6]  Zimbabwe Law Reports, 1995, pp228-241, Dongo v Mwashita and Others. High Court, Harare, Judgment No NH 106-95  

[7]  Zimbabwe Law Reports 2002, Morgan Tsvangirai v Registrar-General and Others, High Court, Harare,  HC 12092/01

[8]  The Hon Justice Mr. J Friedman, Report as Independent Observer, 15 January 2003

[9] Zimbabwe Citizens Support Group,  Report No 1 -  2002 Election Data, Voters' Roll and Evidence

[10] Report and letters to the ZEC from Support Group and Mr Morgan Tsvangirai 

[11] RAU Report, 2013 Vision: Seeing Double and the Dead

[12] ZESN, A Report on a Voters' Roll Audit Conducted in Zimbabwe, October 2010.

[13] The record of this debate may be found in Hansard at: