4 turned up in the past year, but they credit the president with improvements in their own economic conditions. * Second, other people – especially the younger generation and rural dwellers – are afraid to express their true political preferences. Self-censorship is evident among those who think that the survey was sponsored by a government agency. They say they approve of the president when, in fact, they may not. * Third, the most important factor is political propaganda. Since 2000, the government has mounted a comprehensive campaign to revive the nationalist fervor of the liberation war. People who trust the ideological pronouncements of the official government media are very much more likely to give the president a positive rating. * Finally, Zimbabweans are sick and tired of the deadlock between the country’s two main political parties. Two-thirds of all respondents in the 2004 Afrobarometer survey in Zimbabwe consider that “problems in this country can only be solved if MDC and ZANU-PF sit down and talk with one another.” These conclusions are supported by the findings from other studies and reports, and deserve some brief consideration. The first conclusion speaks to the satisfaction of the beneficiaries of state largesse, and, as a recent Afrobarometer study points out, in Africa, “winners” – or those who have a party political identification with the incumbent government – are much more likely to be satisfied with their democracy than the “losers” or the non-partisan2. Additionally, Cho (2004) argues that citizens who live under a balanced two-party system are more satisfied with democratic governance than those who live under both predominant one-party systems and fragmented party systems. He also argues that losers in parliamentary systems show lower levels of dissatisfaction with the way democracy works than do losers in presidential systems. This analysis, which included Zimbabwe, showed that, on average, Zimbabweans were highly dissatisfied with the state of democracy in their country, but that the winners were more satisfied than either the losers or the non-partisan, but also that the non-partisan were slightly more dissatisfied than the losers. This analysis was based on the data from the 1999 Afrobarometer survey, but one can see from the 2004 survey that this conclusion would still hold. The second conclusion relates to the much-reported climate of fear, which is corroborated by the huge output of human rights reports on Zimbabwe. It is clear from these many reports that a number of conclusions can be drawn about the violence that has affected Zimbabwe since February 20003: • All reports show that the violence has been disproportionately one-sided, and against the MDC and other groups not supporting ZanuPF; • All reports show that the violence attributed to ZanuPF is different to the violence attributed the other groups, both in the scale and in the nature; • The violence attributed to ZanuPF shows evidence of systematic torture, abductions, disappearance, summary executions and extra-judicial killings, and this is very rarely the case with violence attributed to other groups such as the MDC; • The systematic torture shows a strong associations with officials of the State – members of parliament, the police, the CIO, and other officials – as well as an 2 See Wonbin Cho (2004), Political Institutions and satisfaction with democracy in Sub-Saharan Africa, Working Paper No. 39, AFROBAROMETER. 3 See Reeler, A.P (2003), The Perpetrators of Gross Human Rights Violations in Zimbabwe from February 2000 to March 2003, Paper presented to “Civil Society and Justice in Zimbabwe: A Symposium”, organised by the Zimbabwe Human Rights NGO Forum, Themba Le Sizwe and the International Bar Association, Holiday Inn, Johannesburg, 11-13 August 2003. IDASA: Zimbabwe at the Cross-Roads: Views of Zimbabwean citizens on democracy and transitional justice.