Zimbabwe Human Rights NGO Forum Report — Of Stuffed ballots and Empty Stomachs election agents who were inadequately prepared or did not perform their task diligently. Some, inspired by pecuniary concerns withheld statistics, demanding an increase in their stipend. Others felt overwhelmed by the predominance of ZANU(PF) sympathisers, police and government officials and did not assert themselves effectively enough to enable them to carry out their observations properly. However, there is sufficient information to examine what happened in a few select constituencies. The MDC expected to win in both Chimanimani and Manyame Constituencies. They lost both: in Chimanimani by about 4 500 votes and in Manyame by just over 7 000 votes. Accordingly, if the MDC’s allegations of stuffing are true, this is the very least number of votes which would have had to have been fraudulently introduced into the count to affect the result. By way of example, using the case of Manyame, this would mean that 700 votes would have had to have been fraudulently introduced at 10 polling stations, or 350 at 20 polling stations or 175 at 40 polling stations. If stuffing took place at 10 or even 20 polling stations, this would be reflected in an inordinate number of votes cast at those stations, and, relative to other stations, a disproportionate number of votes at these stations for ZANU(PF). It is extremely unlikely that stuffing took place at 40 polling stations without being observed by an election agent and that all the personnel at so many stations would have remained silent about such a process. Similarly, if stuffing took place during the time that counting was suspended a large number of seals would have to have been effectively broken and replaced. These scenarios seem unlikely. This leaves stuffing of a few ballot boxes with a large number, or a judicious allocation, of the postal votes or both. The ZESN report has subjected the figures at some polling stations to analysis. Using incomplete MDC election agents returns, those analysing the figures have taken an average percentage of ZANU(PF) voters for known polling stations in a constituency, projected these or extrapolated missing figures for the unknown stations and then compared them to the official result. Any discrepancy between the projected or extrapolated figures of more than 5% is deemed suspect. Unfortunately, this kind of analysis assumes a homogeneous electorate within in a particular constituency. In fact, particularly in the case of Manyame, due to the gerrymandering of constituencies, there is an incomplete community of interests. Pockets of ZANU(PF) supporters, particularly around military bases and resettled land allocated to ZANU(PF) supporters are islands in predominantly MDC territory. Countrywide ZANU(PF) supporters turned out to vote in greater numbers than MDC supporters.75 Manyame lies in 75 For example, in the ZANU(PF) stronghold of Mashonaland Central Province the voter turn out was 57% compared with 32% and 42% for the MDC strongholds of Bulawayo and Harare respectively (figures taken from the Media Monitoring Project Zimbabwe’s table of figures constructed from ZEC’s figures 4/04/05). 25