ZIMBABWE: PARLIAMENTARY ELECTIONS - 2005 15 The intention is that the public shall be able to check the protocol once it has been sent to the constituency, and it shall not be changed after that. However, observers noted in a number of instances that the protocols were taken to the constituency centre for review and approval before being posted. The review could take a long time. In the polling station Mai Musodzi of the Mbare constituency the polling station protocol was completed by 11 p.m., but posted only at 9:15 the next morning. In the meantime, the staff, agents and observers, including four from the international community, were detained at the polling station. The implementation of the process raises doubts about the intentions of the election officials. If the protocols are held back in a review process not open to observers, agents and the public in general, a dark shadow of suspicion will continue to surround the tabulation process. It is in this situation crucial that the detailed tabulation of polling station results up to constituency level is made public immediately, so that everybody can check and feel certain that their polling stations have been correctly tabulated. Doubts Raised on the Official Figures The ZEC should be commended for publishing figures for turnout and rejection of voters on Election Day and the following night. In doing so there will always be a risk of inaccuracies, since the figures will depend on a countrywide reporting system from more than 8,000 polling stations. Even with a reasonable amount of inaccuracies it is better to publish results than to keep them secret until the entire process is over. On the other hand, it was noticed that some of the preliminary figures deviated greatly from the official results. At around 8 p.m. turnout figures as per 2 p.m. were broadcast on TV by the ZEC from almost all provinces. Observers as well as the ZEC reported that most people had voted at the time and very few voters turned up after 2 p.m. The turnout in Harare and Bulowayo increased by 20 to 40% from the figures for the afternoon to those of the final count. This may be realistic. However, in Mashonaland Central, and Masvingo they increased by from 100 to 190%, which cannot reflect the development of the vote during the day. MDC has made public complaints based upon the turnout figures that were officially released at 2 a.m. on election night (1 April) giving the preliminary turnout at the end of the vote for at least 30 constituencies. The preliminary figures had turned out to be far from the final figures. In Mashonaland East, for example, for the constituencies Seke, Marondera East, Chikomba, Murehwa South, Mutoko South, Mutoko North and Goromonzi the official figures published are from 4,000 to 13,000 higher than the preliminary figures (of a total of 20,000 to 30,000). In its press conference on 5 April the ZEC said that, "the figures were being given without prejudice and only for the purpose of giving an indication as to the turnout