effects of this finding against the 2005 poll results. The method was straight forward. From the results data base mentioned above, we identified the set of constituencies for which there were data from both 2000 and 2005. This was necessary since delimitation in 2005 had changed a number of constituencies, and thus this left a set of 74 constituencies for analysis11. We calculated the difference between the votes cast for the MDC in 2000 and 2005, and assigned a binary value to the difference observed [1= improved vote; 0= unimproved vote]. This was done so that we could bring the data in line with the NCA data, and could examine simple frequencies. The first analysis was to examine the differences between rural and urban constituencies as reflected in 2005 poll. As can be seen from Table 13 below, in every case there were lower electoral irregularities in the rural constituencies. Table 13. Electoral Irregularities in rural versus urban constituencies. Rural Constituency Interference with freedom of association 85% Interference with freedom of assembly 74% Interference with freedom of movement 78% Interference with freedom of expression 85% Political violence 94% Gender-based violence 50% Intimidation 82% Militia base reported 48% Inter-party violence 57% Intra-party violence 69% Political use of food 80% Urban Constituency 96% 96% 93% 96% 96% 71% 96% 75% 82% 89% 96% Although it is clear that the electoral irregularities were unacceptably high overall, there is a consistent trend for these to be higher in the urban areas. However, it is also apparent from the election results that the MDC did not uniformly increase its support in the rural areas, and there were a number of constituencies [46%] in which this did not happen. This was then examined separately, and, as can be seen from Table 14 below, apart from the ratings on intimidation and gender-based violence, the electoral irregularities are higher in the constituencies in which the MDC’s share of the vote worsened from the 2000 poll. The differences are however very small – about 5% on average. Table 14. Electoral Irregularities compared to vote share for the MDC [2000 & 2005] Interference with Freedom Interference with Freedom Interference with Freedom Interference with Freedom Political violence Gender-based violence of of of of Association Assembly movement Expression Vote share improved. 86% 82% 82% 86% 91% 59% Vote shared worsened. 93% 83% 85% 93% 96% 58% 11 See notes in Appendix 1. National Constitutional Assembly: THE 2005 PARLIAMENTARY ELECTION. FLAWED, UNFREE AND UNFAIR!