Human Rights and Zimbabwe's Presidential Election: March 2002
A Special Report by the Research Unit
May, 2002

Chapter 8:  Polling and the Assessment of the Results

Polling days

Polling problems 

The results  

Problems with the results  

Where did these discrepancies come from? 

(a)   The Registrar-General vs the ESC  

(b)   Spoilt ballots

(c)   The ‘distractor’ votes

(d)   Urban: rural comparisons    

‘Today is D-Day, as Zimbabweans go to the polls to decide their political and economic future in this watershed Presidential Election… After this election, Zimbabwe will never be the same again. It will either emerge as a truly independent African state in charge of its own political and economic destiny or a seemingly prosperous country where the social and economic divide will be perpetuated’.[1]

‘Zimbabweans carry a very heavy responsibility not only for themselves, but for future generations. There are two simple and clear choices to be made. The first is to vote for backwardness, poverty and starvation, corruption, crime, misgovernance, lawlessness and for a government which is not only ruthlessly repressive towards its people, but one which refuses to be accountable in any way… The second is to vote for hope, peace, for a restoration of dignified living standards for most people, and for the rule of law. It is voting for an environment where, once again, Zimbabweans will be able to live, work, and learn in an atmosphere which allows for self-development and determination, and one in which the government enables, rather than hinders, people to prosper’.[2]

Security was extremely tight, with all light aircraft movement within a radius of 25 kms of Harare International Airport prohibited. This prohibition lasted well beyond polling. Unusual troop movements, from barracks to polling stations, started two days before polling.

Polling Days

Polling was originally set only for 9-10 March 2002.

But on the first day, urban queues started at 4am and stretched for kilometres in Chitungwiza and the southern suburbs of Harare. The Standard reported in its front-page headline, ‘Mugabe in trouble’ – huge trouble. Everyone knew what the result of this turnout would be – a clear, probably overwhelming victory for Morgan Tsvangirai. All of the ruling party’s elaborate legal and other pre-poll arrangements to disenfranchise large categories of voters were effectively nullified by this enormous turnout, which The Herald reported to have been nearly 80% in the rural constituencies but only ‘around 60%’ in the urban centres.[3] It was in fact approximately 54% nationally, 55% in rural areas and 50% in urban constituencies.

Many unable to vote on that day slept overnight in the queues, even though some polling stations closed as late as 3.30am on Sunday 10 March, when the queues started even earlier and stretched even further.

According to the ESC figures, 50,0% of Harare’s ‘voting population’ turned out. The Registrar-General’s figures showed that in June 2000, 49,3% of registered voters in Harare province had cast their votes, but in March 2002 only 47,3% of a voter population nearly 80 000 larger. The will of Harare voters had been grossly suppressed by arrangements that prevented them from casting their votes in the time available. In 101 non-Harare constituencies, on the Registrar-General’s figures, the average daily vote was 13 101 (ESC: 12 769), in Harare/Chitungwiza only 9 441 on the first day, rising to 11 047 on the second.[4] By extrapolation, at least 108 566 failed to vote in these 19 constituencies. And that is confirmed by five Harare constituencies which, against all the odds, managed on the ESC figures to exceed the ‘rest-of-the-country’ average on at least one polling day – Glen View (15 351), Harare Central (15 528), Harare North (15 624), Kuwadzana (18 855) and Mufakose (16 389). To attain these day totals, the average number of votes cast at each of these constituencies’ polling stations was, respectively, 2 193, 1 553, 1 736, 2 694, and 1 821. (Assuming that the seven Kuwadzana stations each worked from 7am to 10pm on Sunday 10 March 2002, when they collectively processed 18 855 people, their throughput must have been three voters every minute.[5] Since observers recorded much slower throughputs – five voters per hour in some cases – more questions are raised than resolved by this calculation.)

To contextualise this problem, we may look at the two rural constituencies which recorded enormous turnouts, both on the first day. According to the ESC, 36 164 votes were cast at Gokwe East’s 62 fixed polling stations, an average of 583 per polling station, or one voter every minute if they all worked non-stop from 7am to 5pm as scheduled. In Uzumba-Maramba-Pfungwe, for which the ESC recorded the next highest turnout of 29 659 at 47 fixed and four mobile booths (a total of 55 venues), the average was 582 votes per station and 539 per venue. A different source gives the highest and lowest number of votes cast at UMP polling stations as 1 363 (over two voters per minute) and 264 (2.3 minutes per voter) respectively.

These large numbers are not credible within the time constraints – or, indeed from other angles explored later. And if we acknowledge that the queues in 2002 were much longer than those in 2000, when Harare’s turnout was 49,3%, there is an even bigger problem with both the ESC and especially the Registrar-General’s figures.

Two ministers (Patrick Chinamasa and Jonathan Moyo), as well as the Registrar-General, had reportedly refused to contemplate the extension of polling. Presumably both ministers were following directions from higher authority. ‘There are no national circumstances warranting an extension of the voting’, Moyo was quoted as saying.[6] Having earlier reportedly told Reuters that government would consider extending polling if necessary, on Sunday just before the polls were due to close (between 5 and 6.30 pm), in response to the MDC’s urgent court application to extend voting, Chinamasa insisted on viewing by helicopter a sample of 15 queues.

At 8.40pm on Sunday 10 March 2002, Justice Hlatswayo in the High Court attempted to compel the Registrar-General ‘to exercise [his public authority] discretion in a manner that addresses the shortcomings revealed in the exercise of [his] initial discretion’. Without specifying working hours or opening times, Hlatswayo ordered Tobaiwa Mudede

‘to exercise his powers in terms of section 94(2) of the Electoral Act [Chapter 2:01] to extend the days for polling of the Presidential election … in all constituencies, so that in addition to 9 and 10 March  2002, the polling for the Presidential election in those constituencies shall take place on Monday March 11, 2002…

… to exercise his powers in terms of section 103L of the Electoral Act to extend the days for polling of councillors and mayors in Harare and Chitungwiza … so that in addition to 9 and 10 March  2002, the polling for councillors and mayors in Harare and Chitungwiza shall take place on Monday March 11, 2002…

… to increase the number of staff at each polling station in each of the Harare constituencies, including Chitungwiza, so as to enable all registered voters the right to vote for President, councillors and mayors within the extended time period…

… to repeatedly and prominently cause to be broadcast the extensions referred to …on all channels of radio and television today, that is March 10, 2002, and also on Monday, March 11, 2002…

This order shall continue to operate notwithstanding the noting of any appeal against it’.[7]

The polling stations were closed around 10.15pm and in some cases those who had not yet voted were forcibly dispersed by the riot police. Even ZTV reported that only half of those wanting to vote had been able to do so by the time polls closed on the second day of voting.

In SI 42D/2002 Mudede half-complied with this order, restricting to the presidential poll the extended time:

‘The Registrar-General hereby alters the Electoral (Presidential Election) Election Notice, 2002, by extending the poll for the purpose of electing a person to the office of President to Monday the 11th March, 2002.’

Mudede’s partial compliance was corrected[8] by the final law-making intervention by Robert Mugabe in the election in which he was a candidate.

        WHEREAS the High Court, on the 10th March, 2002, ordered that the election to the office of President be extended till today for the whole of Zimbabwe;

        NOW, THEREFORE, it is hereby notified that His Excellency the President, in terms of section 158 of the Electoral Act [Chapter 2:01], has made the following notice:-

Title

1.  This notice may be cited as the Electoral Act (Modification) (No. 3) Notice, 2002.

Application of notice

2.  This notice shall have effect for the purposes of—

(a)   the election to the office of President; and

(b)   the mayoral and council elections for Harare; and

(c)   the mayoral elections for Chitungwiza;

held on the 9th and 10th of March, 2002.

Validation of closure of poll and extension of polling

3.  Notwithstanding that polling stations for the elections referred to in section 2 were lawfully closed with effect from 7.00 p.m. on the 10th March, 2002, subject to any extension that may have been authorised by the Registrar-General to enable waiting voters to cast their ballots, the Registrar-General is hereby authorised to extend the poll for the elections referred to in section 2 to 7.00 p.m. on Monday, the 11th March, 2002.

But the polling stations did not open at 7am on Monday 11 March. Only those in Harare and Chitungwiza opened, just before 12 noon, despite one report of ongoing voting in Chinhoyi.[9]

According to SATV at the time,[10] because the presiding officers received conflicting instructions from the Registrar-General, the justice ministry and ZANU-PF, re-opening in accordance with Justice Hlatswayo’s order was delayed, for nearly half the voting day.

Recently-appointed Judge President Paddington Garwe heard the MDC’s complaint against this contempt of a court order, and dismissed it. He accepted that the Registrar-General and his staff had not ‘willfully’ failed to adhere to Justice Hlatswayo’s order. ‘Unforeseen mishaps’, funding and ‘communication difficulties’, the facts that ‘polling stations had closed and started transporting ballot material to counting centres’ and ‘when polling resumed it was discovered thereafter that there were a number of cases of persons trying to vote for the second time’, had together rendered compliance impossible, he said. He did not mention that, outside of Harare province, with one exception no polling station had opened at any time on Monday 11 March 2002. Garwe refused the MDC’s application for a full day’s extension on Tuesday 12 March 2002, accepting the State’s claim that ‘all intending voters had voted’.

‘I am not persuaded on the facts before me that any basis has been shown for this court to interfere with the discretionary power of the Registrar-General not to extend the election. Had it been shown that his decision not to extend the polling was so grossly unreasonable that something else can be inferred from it then this court would have been duty bound to intervene and provide relief. But that is not the position’.[11]

Jonathan Moyo dismissed ‘rumours that the Government had used a decree to nullify the voting that took place in Harare and Chitungwiza yesterday as a result of the High Court order that extended voting by one day’, and explained:

‘We as a party are dismayed by the order that was issued yesterday by Justice Hlatswayo directing that voting be extended in the whole country… It is in the Registrar-General’s power to decide when to do things… What happened yesterday was the court basically usurping the powers of the Registrar-General’.

Moyo’s attitude may (or may not) have related to reports which came in later from voters in Binga and Hwange constituencies, that they had been told voting had been split by candidate, with 9-10 March set aside for voting for Robert Mugabe, and 11-12 March reserved for voting for Morgan Tsvangirai.[12]

According to the ESC, in Harare and Chitungwiza only 50 390 managed to vote on the truncated third day, compared to 179 380 on the first and 209 886 on the second days. On ZTV’s report that only half the Harare voters wanting to do so had managed to vote by the end of the second day, 340 000 (more than triple the extrapolated figure of 108 566) would have been disenfranchised by the ‘unfortunate mishaps’ and ‘administrative oversights’ of election disorganisation.

Polling Problems

Polling was extremely slow at all urban centres, at rates as low as five per hour. There were reports of ‘preferential queues’ for ZANU-PF youth, not registered in that constituency, at the Hugh Beadle Primary School polling station in Sauerstown (Bulawayo North-East).[13] Elliot Manyika was reported to have ordered police in Chitungwiza to harass voters queueing to vote, and to have switched off his cellphone when asked about this allegation.[14]

There was difficulty maintaining order among severely-frustrated voters in many places – Kuwadzana, Warren Park, Haig Park, Glen Norah, Zengeza, etc, as the queues stood still or at best inched forward under new and inexperienced polling station staff drawn from the ministries of defence and home affairs – and some, reputedly, from the semi-literate wives of war veterans. And as the voters vowed not to depart before voting, against assumptions that their frustrations would drive them away, the riot police stood guard, sometimes even closing polling booths as well as injuring voters.

Stations were closed and voters dispersed by riot police at Ward 11 in Zengeza, Chitungwiza; Mount Pleasant and Budiriro in Harare; and at all of those stations at which there were still queues when the stations were finally closed on the unscheduled third day of voting. The unedifying spectacle of the Zimbabwe Republic Police breaking up the queues of disenfranchised voters was screened on South African TV.[15]

Polling stations in the Kariba (Tavoy PS) and Mhondoro (Norton PS) constituencies were reported to have received incomplete voters rolls, with missing surnames.

Twenty of the 52 polling stations in Gokwe Central and three in Masvingo city were reported to have run out of ballot papers.[16] Nonetheless, according to the Registrar-General, Gokwe Central managed to record 28 062 votes, 8 072 more than the ESC noted for that constituency – in a province, admittedly, with unfinalised reporting.

The MDC complained officially that in Mazowe East nine named ZANU-PF youth, who had earlier ‘caused mayhem’ in Chiweshe villages, had been issued with police uniforms, which they wore during polling queue control without displaying ZRP numbers.[17] The police themselves had earlier expressed concern about impersonation:

‘Police have recovered imitations of some police and army identity cards, and their origins have been traced to a studio in Masvingo. But the cards do not have the security features found on the genuine police and army IDs’.[18]

In Kuwadzana Extension, ‘war veterans’ assaulted and dispersed people, preventing them from voting. Their colleagues closed Shamu polling station in the Odzi farming area at 2.30pm on the first day of voting and the ballot boxes were transferred for safe keeping to Odzi police station.[19] Joseph Chinotimba was reportedly permitted to brandish two AK47s within 100 metres of the peri-urban polling station at Bunkers Hill Farm.[20]

There were numerous reports of prohibited electioneering and intimidation within 100 metres of polling booths, mainly by ZANU-PF. Similar MDC behaviour was generally outside (if only just) the prohibited radius.

In those constituencies worst affected by the enforced absence of MDC polling agents, there were huge discrepancies between the verified vote count and the results announced. For example, in Mount Darwin South, 21 000 voters were recorded by the MDC but 33 015 appeared in the announced result, compared to the ESC’s 35 631 with over 17 000 votes cast on day 2.

Information on the turnaway rate was, as usual, very sketchy, although some press reports put the total turned away at over 360 000.[21] (In parliamentary by-elections since June 2000, the turnaway rate has been as high as 20%.) There was no officially-reported information at all for nine of the ten provinces. For only one province, Mashonaland East, was the total given. Here, 32 632 at least were denied ballots (and possibly another 20 976 – from the reports it was not clear whether the first was a running total for days one and two). As a percentage of total votes cast (ESC figures), that is 8.9%, possibly 14,8%. Other fragments of turnaway information included: 2 000 wannavotes disallowed in three areas of Zvimba North; another 1 001 at a few stations in the two Hwange constituencies; 54 at Nembudziya in Gokwe Central; 96 at one Buhera North station on the first day of voting. About 1 000 Harare and Chitungwiza voters were accused of trying to vote twice, were arrest and deprived of their vote.

If we use the apparently complete, official figures for Mashonaland East, 550 polling stations each turned away an average of at least 60 voters, possibly 98. Extrapolating those rates to the national total of 4 712 fixed and mobile polling stations, yields the number of potential voters deprived of their votes in a range between 282 720 and 461 776, or somewhere between 9,2 and 15,1%. In early polling in Harare, it was counted at approximately 10%, so both figures are entirely feasible. They do not include those voters in Harare and Chitungwiza who had not managed to get into polling stations before they were finally closed.

The Results

As the last of the polling results were long delayed, presumably as the Registrar-General informed Robert Mugabe privately of the final result, ZTV began screening footage of liberation war atrocities, the independence celebrations, and ‘third chimurenga’ songs against the background of the National Heroes Acre. It was clearly both celebratory and intimidatory. The Norwegian foreign observers had already, in a preliminary statement, condemned the entire election as not meeting minimum international standards in any of its aspects prior to counting. The Zimbabwe Election Support Network had been similarly condemnatory.

When the final result was announced, at noon on Wednesday 13 March 2002, the entire Election Directorate panel looked extremely grim, without a single smile or note of congratulation. The official results were initially given as, and a month later altered to, the numbers appearing in the second and third columns of the tabular layout overleaf. The fourth column shows the totals generated independently by Microsoft Excel 98 from each of the constituency results as originally announced by the Registrar-General on ZTV on 12-13 March 2002. Finally, the difference between independent calculations and the Registrar-General’s 14 March figures appear in the last column. 

Problems with the results

Independent calculation of the results shows the announced totals to have been incorrect. To begin with, the correct overall total for the figures given was not 2 298 758, but 2 998 759. (For later emphasising this discrepancy of 700 000, The Daily News editor was charged under the Freedom of Information and Protection of Privacy Act with publishing false information.[22])

Then, on the final, verified figures for individual constituencies announced by the Registrar-General himself, the totals for the national vote as well as the two major candidates were, simply, wrong. The constituency results totalled 3 062 303 – 63 544 more than the Registrar-General’s figure. Moreover, the announced constituency results gave Robert Mugabe only 1 676 559 votes, while Morgan Tsvangirai’s total was 1 280 591, reducing the real difference between them by 30 843 from 426 811 to 395 968. Adding these two problems together yields 94 387 votes, one-third of which were definitely in favour of Morgan Tsvangirai. The probability is that the rest were also in his favour.

Even more worryingly, the ‘verified’ and announced results for some constituencies seemed to be highly unstable even while on the television screen. Six constituencies were affected:

1.     In Buhera North, the total votes cast shifted from 30 182 to 30 246, while Robert Mugabe’s total accreted another 1 000 votes, from 15 248 to 16 248. The ESC total was only 19 171 votes, 17 448 (91%) of which were cast on the first day of voting. Thus the Registrar-General’s announced figure was 157% of the ESC total, in a province that had submitted final returns to the ESC. The 2002 turnout was 126,2% of that in June 2000, while the spoilt papers fell from 803 to 419.

2.     In Guruve South, Mugabe’s total jumped 10 000 from 20 651 to 30 651, while the total shifted from 23 925 to 33 691, compared to the ESC total of 23 419. Turnout was very slightly higher than in June 2000, spoilt papers fell from 415 to 204, and 84,6% of the votes were cast on day 1.

3.     Murehwa South is a constituency largely comprising occupied commercial farms. Here, while the total vote remained static, Mugabe’s share moved up 2 246 votes, from 14 667 to 16 913. The Registrar-General’s total of 23 175 fell short of the ESC total of 26 755 votes cast, by 3 900, or 14,6%. In other words, one in every seven votes recorded by the ESC disappeared from the mouth of the Registrar-General. Again, day 1 turnout was 84,3% of the total, turnout (on the ESC figure) was over 20% higher than in June 2000, but the rate of spoilage was much the same as in the Parliamentary election.

4.     On the home turf of Vice-President Simon Muzenda, in long-resettled Gutu North, turnout on day 1 was only 55,6% of the total – 13 958 votes were cast on day 2. The onscreen total moved up 9 226 from 23 311 to 32 537 votes. Robert Mugabe’s votes accreted 7 353 (up from 15 171 to 22 524), while Morgan Tsvangirai’s total gained 2 723 (from 6 211 to 8 934). These adjustments brought the Registrar-General’s numbers into range of the ESC’s total of 31 407 votes cast. And unlike the three constituencies already dealt with, Gutu North saw a turnout only 95% of its June 2000 figure, while cutting its spoilage rate from 806 to 432 papers.

5.     In Nkayi, turnout rose 15% from June 2000, with 77,8% voting on the first day and another large decrease (38,5%) in the rate of spoilage. 10 votes swelled Mugabe’s total to 11 562, while 1 000 disappeared from Tsvangirai’s initial haul of 16 616. The Registrar-General found only 28 441 votes to count, 3 763 less than the ESC’s 32 204.

6.     Mugabe was also credited with an extra 1 000 votes in Gokwe North, boosting his total from 22 663 to 23 663. In this constituency, where the ESC had counted only 13 190 votes (58,6% of the June 2000 turnout), the Registrar-General found another 19 141 in his tally of 32 331 (143,6% of June 2000). Only 9 029 voted on 9 March, and 4 161 the following day, according to the ESC. But the Midlands (like Matabeleland North and Mashonaland Central and West provinces) was noted by the ESC to have had some constituencies (numbers and names unspecified) which had not submitted their second day figures. So maybe those newly-found 19 141 swelled the second day total for Gokwe North to 23 302 votes, which would have required each of the 53 polling stations to process voters non-stop at a rate of one every minute and 24 seconds for 10 hours.

In total, these on-screen changes added 19 886 votes to Robert Mugabe’s total after the announcement of the final, verified numbers.

Where did these discrepancies come from?

(a) The Registrar-General vs the ESC

The Electoral Supervisory Commission had, on Tuesday 12 March 2002, before counting was finished, given its own verified figures for ballots received at the constituency counting centres to the MDC. Adding up the ESC constituency totals gave a national vote of 3 018 955, compared to the Election Directorate’s televised release of 2 998 759 (a discrepancy of 20 196) and the properly-calculated total of 3 062 303. So even the larger ESC figure was actually 43 348 less than the actual total of constituency votes based on the Registrar-General’s own figures. Where did these extra 43 438 votes appear from in the ‘final’ figures?

One possible answer lay in the immense variations[23] in the ESC and R-G figures at the level of individual constituencies. In not one single constituency did the ESC figures coincide with those of the Registrar-General and Election Directorate. Overall, these differences totalled 432 402 votes, or 14,3% of the ESC total – one in every seven votes ostensibly cast.

292 664, or 67,7%, of these discrepant votes were concentrated in only 30 constituencies, each of which had a discrepancy of at least 5 000, the average being 9 755. These discrepancies were concentrated as follows, with those constituencies recording over 10 000 discrepant votes being italicised:

1.     9 of the 16 Midlands constituencies (56,3%) – Chirumanzu, four of the five in Gokwe (Central, East, North and South), Kwekwe, Mberengwa West, Silobela and Zvishavane;

2.     6 of Manicaland’s 14 constituencies (Buhera North, Chimanimani, Chipinge North, Makoni North, Mutare West and Mutasa) – 42,9%.

3.     5 of 14 constituencies (35,7%) in Masvingo province (Bikita East and West, Chivi North, Gutu North and Zaka West).

4.     4 of Harare’s total of 19 constituencies (21%) – all in high-density areas (Glen View, Harare East, Highfield and Kuwadzana).

5.     Most oddly, only 14,7% (5 of 34) of all constituencies were affected in the three Mashonaland provinces: one in Mashonaland East (Murehwa North); two each in Mashonaland Central (Guruve North, Mazowe West) and Mashonaland West (Hurungwe West, Makonde). This pattern very strongly suggests that the massive vote for Robert Mugabe in the majority of constituencies in these three provinces had already been so heavily rigged that these constituencies could not be used for further ‘adjustments’.

6.     Of the two constituencies heavily affected in Matabeleland North province, Binga gave Morgan Tsvangirai his largest win of 26 886 votes; whereas Bubi-Umguza, the peri-urban constituency virtually enclosing Bulawayo, most surprisingly in view of Matabeleland’s political history and the voting patterns of Bulawayo and those constituencies surrounding it, in June 2000 as well as March 2002, apparently gave 56% of its vote to Robert Mugabe.

7.     Finally, discrepancies in Bulawayo and Matabeleland South provinces were below 5 000 per constituency.

(b) Spoilt ballots

Another possibility lay in the spoiled papers, many of which, from diverse rural constituencies, were totally blank, indicating, perhaps, a strategy to keep out of future trouble among those who would have preferred either not to vote or to vote for the MDC. Rural voters were ‘mobilised’ behind headmen sympathetic to ZANU-PF in queues at the polling stations. Having been persuaded that their vote was indeed secret, many were reportedly warned of retribution if the total vote did not equal the vote for Robert Mugabe. In Bindura district, among others, after the election, ZANU-PF youth sought to chase out of their urban as well as rural homes all 11 279 people who voted for Morgan Tsvangirai.

The Herald[24] published the figure of 115 277 for spoiled papers. ZESN estimated them at 95 670. In his official announcement, the Registrar-General had said they made up the difference between 2 298 758 and ‘around 3 million’ votes cast – ie some 700 000.[25] On the Registrar-General’s official results by constituency, they totalled only 47 673, averaging just 397 per constituency, compared to 62 684 (average 522) in June 2000. Of this 2002 total, 42 889 were spoilt in 86 rural constituencies, and there was a large discrepancy between average rural (449) and urban (141) spoilage rates per constituency. On 10 April 2002, the Registrar-General claimed 48 131 papers had been spoiled.[26]

Given the lack of voter education and organisational chaos at polling stations in March 2002, this official drop of 24% in the spoilage rate, even the marginal fall of 4,4% in the 86 rural constituencies, mostly with increased voter turnouts over June 2000, raises questions. The ZESN figure, just over double the official figure, seems much more probable. But where did the missing half disappear to?

(c) The ‘distractor’ votes

The votes for Kumbula, Maya and Siwela were interesting for a number of reasons. Firstly, they totalled 54 655 (490 less than the figure given by the Registrar-General), in roughly the same range as those polled by Ndabaningi Sithole and Abel Muzorewa in the 1996 Presidential election from which they both withdrew before polling. Secondly, while almost every other of the Registrar-General’s figures changed, they were completely stable.

Their very stability revealed some interesting relationships, detailed below. There is a 2:1:1 ratio linking their urban votes, which changes to a rural (and overall) 3:1:1 ratio. Such ratios among candidates completely unknown[27] who did no advertising which would have reached rural voters, raise the interesting possibility of a ‘fixed base vote ratio’ which could have been applied as a rigging base. Gross deviations may then indicate further and different rigging.

(d) Urban : rural comparisons

A marked class divergence between urban and rural Zimbabwe is often alleged, which in June 2000 seemed to have returned the urban and peri-urban (and some rural) constituencies to the MDC and converted ZANU-PF into an exclusively ‘rural party’. Robert Mugabe’s focus on land would appear to have exacerbated this divide, despite having officially won the count in certain peri-urban constituencies. His bitter post-electoral condemnation of urban vatengesi (sell-outs: those to whom his wife publicly referred as katsi nembwa – cats and dogs), was, fittingly, delivered in his home area of Zvimba. It is therefore instructive to compare urban and rural constituencies on a number of variables (summarised in Table 8.1 below), before attempting to count the votes demonstrated to be problematic as reflections of the voters’ will.

It is clear from these figures that the urban vote was not equal to the rural vote in this election. Urban voters voted 3:1 for the opposition at one-fifth of the polling venues available to country voters voting 2:1 for the incumbent president; they were systematically and quite deliberately disadvantaged. The time and space planned for their use was designed to disenfranchise at least half of the urban voting population – which was well-known to be anti-Mugabe. That goal was achieved.

Table 8.1 Zimbabwe’s 2002 Presidential election – comparative urban : rural ratios

But the net result, as shown (overleaf) in the final summary compilation of the numerical problems, was to cast very large question marks over one million votes – one-third of the total allegedly counted. Mess-ups that magnificent have little hope of being accepted as reflecting the will of the voters, wherever they occur.

But the real problem is that it is difficult to tell whether the mess was a result of sheer, gross incompetence; or whether the rigging involved was systematic. It is likely that the huge urban turnout forced an unscheduled ad hoc ‘adjustment’ of certain constituency results, thus creating a messy outcome not easily disentangled into its triple components of incompetence, baseline rigging, an