Preventing Violence in Future Elections:Moving Towards an Early Warning System for Zimbabwe

Keywords:

• peacebuilding
• development
• elections
• violence
• early warning systems

Background

Zimbabwe has held elections since its independence in 1980 whenever they were due (1980, 1985, 1990, 1995, 2000, 2005 and 2008); as a result, elections have been a very visible feature of post-independence politics in the country. However, these elections have been held under different historical, political, social and economic contexts. The 1980 elections were held when the country was coming out of a liberation war and were marred by violence and intimidation of the electorate by both the Rhodesian Front (RF) and Zimbabwe African National Union—Patriotic Front (ZANU-PF). The 1985 elections were held during the civil conflict (Gukurahundi) in Matabeleland and Midlands (1982–1987). Gukurahundi (Shona name for the early rain that washes away chaff before the spring rains) refers to the massacres that were aimed at stamping out Zimbabwe African People's Union—Patriotic Front (PF-ZAPU) ‘dissidence’ and resulted in at least 10,000 deaths, as well as the beating, torture, rape and disappearance of many more villagers (Catholic Commission for Justice & Peace in Zimbabwe 2007).

The 1990 elections were held after the signing of a unity accord between ZANU-PF and PF-ZAPU, but in the wake of the newly formed Zimbabwe United Movement (ZUM), which challenged the status quo by vigorously opposing the idea of one-party legislation. ZUM participated in the elections but strongly complained about intimidation and persecution of its leadership and members during elections. The 2000 parliamentary elections and 2002 presidential elections were marred by violence and loss of life due to politically motivated violence exacerbated by land expropriations that gave rise to the large-scale physical displacement of farm workers and their families, ostensibly because they had aligned with white farm owners in support of the Movement for Democratic Change (MDC)—a new opposition party that emerged during these elections (Bratton & Masunungure 2008, 50).

Despite the MDC's threat to boycott elections on 31 March 2005 unless the government implemented the Southern African Development Community (SADC) principles and guidelines governing democratic elections, the elections went ahead with observers invited only from ‘friendly’ countries. Three months later, the government introduced Operation Murambatsvina (Drive out the filthy), which affected up to 700,000 city dwellers and was widely interpreted as ZANU-PF retaliation against urban voters, who had come out overwhelmingly for MDC in the 2005 presidential elections. In 2008, the 29 March elections did not produce a clear presidential winner, necessitating a run-off election on 27 June. The period leading to the first round of elections was generally peaceful while the period leading to the run-off election was marred by violence, intimidation and disappearances of perceived political opponents, resulting in the withdrawal by one of the two candidates and rendering the election illegitimate. This forced ZANU-PF and the two MDC formations (MDC-T led by Morgan Tsvangirai and MDC-M led by Authur Mutambara) to sign the global political agreement (GPA) in September 2008, and to form a unity government in February 2009, under SADC mediation.

It is clear that in Zimbabwe election processes and practices at any given time greatly reflect the political scenarios on the ground. When the political climate has been tense, agitated, polarised, electioneering has generally been marred by violence. Although some literature exists on elections and violence, questions remain, such as why elections create occasions for violence and what is the relationship between violence-producing and violence-reducing propensities of elections (Rappaport & Weinberg 2000, 18). Beyond Zimbabwe, there are many historical examples of pre- and post-electoral violence. Fears that Papandreou's party would repeat its landslide victory generated a military coup in Athens (1967) to cancel the election (Rousseas 1968, 9–12). Nineteenth-century English voters found elections so threatening that some ‘voluntarily attempted to disenfranchise themselves’ (Seymour 1970, 189). Persistent violent attacks kept African-Americans from voting in the US south for nearly a century (Chalmers 1963, 321). Post-election violence usually occurs soon after the announcement of election results. The December 2007 elections in Kenya resulted in extensive violence that claimed more than 1,000 lives, while the 2007 general election in Nigeria was contested by the opposition and the international observer team and ended in court. The Zimbabwe Election Support Network (ZESN) reported that in the post-election violence period in 2008 they recorded 171 cases of murder, 16 cases of rape, 9,148 cases of assault, 827 cases of displaced persons, 352 cases of torture, 9,590 cases of harassment, 2,165 cases of displacement and 137 cases of discrimination (ZESN 2008, 64).

Electoral violence has become one of the cogent challenges affecting the broader issues of peacebuilding and development in Zimbabwe. As the country prepares to vote in a referendum and national elections in 2013, there is need to explore the relationship between elections and violence, especially when elections are considered a mode of legitimate succession, and to find ways to prevent election-related violence. This briefing proposes a prevention mechanism in the form of an early warning system to prevent violence in future elections in Zimbabwe.

Moving towards an Early Warning System for Zimbabwe

Early warning consists of the collection and analysis of information about potential and actual conflict situations, and the provision of policy options to influential actors at the national, regional and international levels which may promote sustainable peace (Forum on Early Warning and Early Response [FEWER] 1999). Early warning systems therefore seek ‘to identify the likelihood of specified events or conditions occurring in certain locations in the future’ (Schmeidl 1998, 472). They are prospective in nature. That is, they identify situations with sufficient advance timing that preventive measures can be adopted which will reduce the likelihood or the severity of future events. Early warning systems are distinguished according to the phenomena under consideration—for example, election violence, civil war, refugee migration, drought, tsunami and so on.

The electoral violence experienced in Zimbabwe and its devastating consequences (such as loss of life, damage to property, disenfranchisement of voters and loss of livelihoods) could have been minimised or avoided if effective preventive measures were in place. The major purpose of an early warning system as proposed in this briefing is to enable preventive rather than reactive measures to election violence through cultivating the capacity to predict that certain patterns of overt political or electoral behavior may potentially lead to violence. Early warning in this case presumes that the prevalence of conflict or the outbreak of violence is related systematically to a set of historical, demographic, social, economic, political and other variables. The system attempts to identify these key variables with predictive power and develop means to assess the potential or escalating violence, by using different techniques to generate relevant information and analysis.

Developing early warning indicators

Indicators are often developed to measure and monitor the progression of conflict factors against appropriate benchmarks or baseline factors. They are useful in providing targets against which to set contingency plans and responses as well as being valuable pointers to monitor change in the prevalence of violence. In Zimbabwe, as in other countries where violence permeates structures throughout society, indicators for an early warning system will be valuable in a range of areas: across peace and security, political, judicial and socio-economic realms. Possible indicators to develop in the Zimbabwe situation include the following:

a.	Security indicators such as deaths, injuries, disappearances, abductions, rape, torture, politically motivated arrests and cruel, unusual or degrading treatment.
b.	Political indicators such as denial of political rights, political persecution, unwarranted banning of political rallies, unrestricted freedoms of expression, movement and assembly, politically-motivated violence, psychological abuse through reminiscing about violence (e.g. liberation war songs and war cries/slogans/jingles), the rise of violent groups like Chipangano in Mbare or Al-Shabaab in Kwekwe, and interference with election processes.
c.	Judicial indicators such as constitutional protection of individual rights, guarantees of due process, human rights legislation, judicial freedom from political interference, equality under the law, and prosecution of criminals regardless of their political affiliation or status in society/government.
d.	Socio-economic indicators such as freedom of thought, belief and religion; level and type of social interaction between and within groups; political-economic policies; level of economic or employment discrimination; politicized land reform policies; and high levels of social and economic corruption.

It is good practice to involve all stakeholders in developing the indicators; not only should this produce better indicators but it is also an important opportunity to build a common understanding of the national and local contexts, to ascertain joint priorities and to agree on benchmarks of progress. At the national level, an all-stakeholders workshop could be conducted to develop national indicators, while at the local level indicators can be developed through workshops and based on local meanings and understandings of both peace and violence-generating factors. Communities should be allowed to identify indicators in their local languages. Indicators identified at the local level can easily complement national indicators in most cases. Although localised information is more likely to be accurate than national-level aggregated data, the effectiveness of the proposed early warning system will depend on the accuracy of information at all levels.

Proposed structure

Through representative structures, early warnings systems monitor the flow of political events, with special attention to actions that are likely to precipitate the onset of conflict or violence in high-risk situations. The central purpose of early warning is thus not only to identify potential problems but also to create the necessary political will for preventive action to be undertaken. Accordingly, electoral violence early warning represents a proactive political process whereby networks of stakeholders conduct analysis together in a collective effort to prevent likely violence from occurring.

At local level, ward monitoring and response committees (WMRC) could be formed in a similar format to the District Code Monitoring Committees (DCMC) in Sierra Leone. In Zimbabwe, wards are local governance structures commonly used by government for administration, development and electoral purposes. The WMRC would be made up of representatives of local groups, such as, traditional leaders, community leaders, local religious leaders; teachers, nurses, women, and youths; and representatives of political parties, local government and law enforcement agents. The involvement of all stakeholders is important in order to gain political will and dispel any suspicion by the state. Local and international resources would be mobilised to support the activities of the WMRC.

At the national level, a violent response and appeals commission (VRAC) could be established with authority to respond to and deal with election violence in a timely, fair and transparent manner. The VRAC would be made up of representatives of all stakeholders, including government and bodies like the National Association of Non-governmental Organisations (NANGO), Joint Monitoring and Implementation Committee (JOMIC), the Organ for National Healing, Reconciliation and Integration, Human Rights Commission, Zimbabwe Electoral Commission and the media. While multiparty liaison committees exist during elections, their main focus is not violence prevention. Their work is confined to managing conflict between political parties contesting any elections and dealing with enforcement of the Code laid out in the Fourth Schedule of the Electoral Act (Chapter 2:13, 160A-160D). The VRAC members would undergo capacity-building training focusing on receiving and processing early warning information, addressing concerns about the electoral process, responding to election-related conflicts, learning how to engage in dialogue and collaboratively with other stakeholders.

Operationalising the Proposed Early Warning System

The true test of any early warning system is its ability to generate not only timely analysis but also effective response options; an inextricable link to early action is a core principle of an effective early warning system. Consequently, there is need to tailor the early warning system to the requirements and specific needs of its end-users and decision-makers. Moreover, and in view of the wide range of stakeholders involved and the information that should be communicated, the early warning system should be operationalised at all levels. At the local level, there is need to convene orientation meetings; capacity-building training workshops in conflict transformation, receiving and processing of early warning information, early warning and response to electoral violence; and workshops to develop indicators and response strategies. At the national level, there is need to convene an orientation meeting, capacity-building training and workshops to develop indicators and response strategies, and round-table meetings to mobilise resources.

The collection, processing and analysis of information can be done in different ways, depending on the context and severity of the situation. However, accurate information will be needed at all levels. In the current context in Zimbabwe, where many people operate for long periods under fear of being arrested or detained, it is a challenge to work out how people can collect and submit information relating to violence. The following considerations are important when mapping out an operationalisation strategy:

•	Information collected should be reliable, timely, clear, meaningful, accurate, adequate, valid and consistent and should be acted upon at local and national levels.
•	People may use the social media to transmit information to the leaders or committees, at either local level or national level, whichever is possible,
•	People should be able to send in information in various forms, including written, verbal, text, pictures, videos, etc.
•	The model may develop a database similar to what Ushaidi.com did after the violent 2007 elections in Kenya. Ushaidi developed and uses software that enables every mobile phone user in Kenya to report incidences of community tension. This strategy is simple and fast.
•	Other means that may be proposed by stakeholders.

The proposed system would require support and commitment from all stakeholders, especially political will from the state and political parties. To achieve this, ownership of the process by all stakeholders is important. The media also need to play an unbiased role in promoting the process and subsequent activities. Above all, the VRAC will need to mobilise local and international resources to support the activities of the proposed electoral violence early warning system.