The Question
This question was posted by a member of the ACE Practitioners’ Network

As a result of recent violence surrounding recent elections – such as that in Kenya, Mongolia and Pakistan – International IDEA has begun work on identifying existing methodologies that may be used to forecast the potential for violence during the electoral cycle.  Although we are aware of ongoing efforts to monitor potential violent outbreaks in a more general sense, we have but a few examples of tools that are being used to predict/prevent violence that is triggered by the electoral cycle.

Do you know of any tools used by electoral bodies or other institutional or civil society stakeholders to map risks of violence during the electoral cycle? If so what are these tools and their methodologies?

Summary of responses
Several replies stress the importance of gathering information on electoral violence from people on the ground, coupled with a methodology that then enables this local information to feed into a tool and form the basis for analysis.

The Election Violence Education and Resolution (EVER) methodology developed by IFES is cited by two experts as a valuable tool and resource for mapping the risk of electoral violence. Lebanon (2009) and Burundi (2010) are two cases where this methodology has been employed. The methodology is based on a systematic and detailed analysis of factors that are related to electoral violence and the level of risk that they pose. Incidents are continually tracked in an effort to establish trends.

The project in Burundi (Amatora Mu Mahoro or Elections in Peace) incorporates the crowd sourcing software Ushahidi. This allows people to report information on instances of electoral violence via SMS, phone and email, and in real time. All this data is then mapped by the software.

The Institute for Democracy in Africa (Idasa) has also developed an Information Communication System that maps and analyses election-related violence and which relies on input from citizens.

Non-technological based initiatives for charting and preventing electoral violence include the examples of Sierra Leone, where district-level committees made up of stakeholders acted as monitors and reported violence/conflict, and Kenya, where Political Parties Disputes Tribunals were designed to hear a range of political disputes in an attempt to forestall electoral violence.

Examples of related ACE Articles and Resources
Encyclopaedia:
• Focus on Election and Security

Support and Advice:
• Consolidated reply: UNDP guidelines on prevention of election violence  

External Resources
• BRIDGE

• IFES, Election Violence Education and Resolution (EVER) project

• IFES, Lebanese Election Violence Risk Assessment (LEVRA) project

•  USAID: Electoral Security Framework handbook 
•  Amatora Mu Mahoro (Elections in Peace) 
•  IDASA

Names of contributors
1. Koki Muli
2. Magnus Ohman
3. Richard Chambers
4. Gregory Minjack
5. Jeff Fischer
6. Gabrielle Bardall
7. Dr. Noor Mohammad
8. Derrick Marco

There is also the body of information that has been developed by the BRIDGE Project in its newer Modules. However, development of early warning tools for electoral violence is still at nascent stage. Yet, electoral violence in some cases can be predicted based on past experience, tensions and an assessment of the electoral environment and prevailing circumstances in a country. For example, in the case of Kenya, electoral violence was not surprise, what was the surprise was the magnitude and the level of brutality and impunity especially from security apparatus of the State, not to sound to excuse criminal gangs. Since the re-introduction of multi-party elections during the 1992 general elections, violence has been attendant to any general elections and in some cases even by-elections. Cases of electoral related albeit in small scale has been experienced concurrent to the general elections in 1997, tensions though not high in 2002, because of the nature of those transitional elections; and very serious hate speech, violence and balcanisation of people in tribal and regional blocks - in the Referendum of 2005. Indeed, many can claim, the electoral violence of 2007, could have been avoided if the conflicts and political problems associated with the 2005 Referendum on the proposed Constitution had been dealt with.

Since then Kenya has been attempting to put in place electoral conflict resolution mechanisms. The former Electoral Commission of Kenya (ECK) had Constituency Peace Committees which attempted to resolve electoral and political disputes shortly before and during elections to avoid electoral violence. The Committees were temporary and were constituted only during electiosn or by-elections - they were also not archored in law.

Now the Political Parties Act, 2007 which became operational in 2008, establishes Political Parties Disputes Tribunal whose members are to be appointed by the Chief Justice with approval by Parliament. The Tribunal shall determine disputes between members of a political party; between members of political parties forming a Coalition (as is the case in Kenya) and shall also hear appeals from the Registrar of Political parties who can also hear certain disputes. The purpose of this Tribunal was to forestall electoral violence since much of it emanates from political parties, their contestants and supporters. To this end you can see it as a tool of resolving disputes at the earliest opportunity possible though in itself not an early warning tool. However, the Chief Justice hasnt appointed the members of this Tribunal yet and the Kenyan Coalition disputes cannot be addressed by this Tribunal.

Experts in elections are beginning to put together a list of conditions, when they exist in a country, should be used to serve as early warning mechanism for electoral violence. As I mentioned - political tensions, discontent, past experiences, electoral environment and certain benchmarks may be an indicator. This is work in progress. 

If you want to get early warnings about potential violence, it is difficult to beat having local people on the ground monitoring events. In Sierra Leone, the Political Parties Registration Commission (PPRC), one of the country's two EMBs, created district level committees in early 2007, which were active until the autumn of 2008. They seem to have played a similar role to Constituency Peace Committees that Koki Muli refer to in Kenya. 

These committees consisted of representatives from each of the political parties; civil society; religious groups; the police; and the National Electoral Commission. Their role was to monitor compliance with the political parties Code of Conduct, which the parties had voluntarily agreed on. The work included preventive activities to avoid violence as well as detection of potential conflicts. When such indications occurred, the committees reported to the PPRC, and together they could quickly intervene to reduce tension. 

While the 2007 and 2008 elections were not entirely without violence, the work of the committees was widely hailed by both international and domestic actors as playing a crucial role in detecting and defusing instances of conflict during the electoral cycle.

The committees had no legal mandate nor any power to enforce sanctions, but I would argue that this was of little importance. Much more relevant was that they consisted of people with local expertise who could quickly intervene before conflicts escalated to initiate a dialogue between the involved stakeholders.

IFES has worked extensively on developing a methodology for analysis of issues of election-related violence through its Election Violence etc (see www.ever.org). 

Ahead of the 7 June 2009 parliamentary elections in Lebanon, IFES has developed The Lebanese Election Violence Risk Assessment (LEVRA) project to identify the risk indicators for election-related violence.  The project uses a methodology that is based on systematic analysis of factors (both global and country-specific) that are common to or symptomatic of incidents of political, confessional and/or electoral violence.  This includes identification of the different types of relevant possible violence (Lebanon has experience of many different kind of conflict) as well as aggravating and mitigating factors that may increase or reduce the risk of violence countrywide or in specific districts.

Analysis is undertaken for each of Lebanon’s 26 electoral districts and countrywide: this includes a review of socio-political data (including previous results and disaggregated voter registration data by confessional group) and the historical and current political environment (including the level of political competition and previous incidences of violence).  IFES also engages in a broad and cross-party consultation with electoral stakeholders before reaching and releasing its baseline assessment on the level of risk indicators for electoral violence in each district. 

The LEVRA project also includes an ongoing tracking and mapping of relevant incidents (location, severity, perpetrators, injured parties) to identify possible trends of violence during a six-month pre-election period and the campaign; information of these incidents is received from different sources but does not involve field research or attempts at independent verification of reports.  Based on this tracking, IFES reviews its baseline assessment of risk indicators on a weekly basis.

The beneficiaries of the project include the security agencies who are able to use the risk assessment as a tool for the strategic security planning.  The LEVRA risk assessment model has been utilised by the Lebanese Ministry of Interior in its security planning for the election period.  The project has also involved training sessions on risk assessment models, study visits on security planning and expert advice and review on the development of risk assessments as tools within security planning.

The project’s development has sought to respect the sensitivity of undertaking risk assessments during an election process; with this in mind, the assessment findings are not made public.  However, the LEVRA program has sought to maximise the availability and use of electoral and security information and risk analysis to all relevant stakeholders (including election observers and different kinds of civil society groups involved in elections and conflict prevention/resolution). The information is updated on a weekly basis and is available to subscribers on a website.

Following the election, the project will publish a comprehensive report, detailing its methodology and lessons learned from its application.  Information can be obtained from [email protected] and [email protected].

 USAID has just commissioned Creative Associates to create an Election Violence Framework for embassy officers.  I am not certain when it will be available, but it should be a good addition to this sector's tool box.  Check with Washington, DC-based Creative Associates.

Attached is a PDF of the Electoral Security Framework handbook developed by Creative Associates for USAID referenced above.  It can be placed in the tool box. 

In addition to Richard Chamber's example, I suggest you look at IFES' current work in Burundi during the 2010 elections. Much effort was made to ensure that Burundi does not fall back into political strife and violence as the country held the five elections (communal, presidential, national assembly, senate and local) between May 21 and early September. The centerpiece of IFES’ work in Burundi is the first-ever nationwide election violence prevention system named Amatora mu Mahoro (Elections in Peace), which results from the collaboration of over a dozen Burundian and international organizations. The Amatora Mu Mahoro system draws on IFES’ ongoing EVER Project (Elections Violence Education and Resolution) which has seen successful results in monitoring and preventing election violence in over a dozen countries since 2003. The system also introduces the Ushahidi web platform in Burundi (www.burundi.ushahidi.com). Ushahidi was developed by Kenyans in response to the electoral violence that shook the country in 2007/2008, and allows the rapid sharing of information on peace and violence incidents, via SMS.

One of IFES' largest EVER initiatives, Amatoro Mu Mahoro draws on a rigorous research methodology that reports verified occurrences of peace activities and election-violence incidents gathered from monitors in all of Burundi’s 129 communes. By offering comprehensive, reliable data for violence mitigation efforts, the program seeks to consolidate Burundi’s achievements in peace-building in the last decade. In essence, the EVER methodology gathers information on incidents and indicators of violence and peace activities, and includes real-time reporting on incidents. The data is transmitted through the Ushahidi platform, which allows monitors to submit reports on the incidents via SMS, email and telephone. The accrued information is compiled on an online map and shared with responders and local teams trained in election-violence mitigation so that they can respond to the incidents. The data is also used to compile in-depth pre-election reports on trends and risk areas, to enable election planners to adequately prepare. In total, Amatora mu Mahoro integrates monitoring data from over 450 trained monitors in all 129 of Burundi’s communes.

In addition to violence monitoring, Amatora mu Mahoro actively supported civic education and peace and reconciliation activities around the country by tracking their outreach and encouraging programming in areas of heightened need. By incorporating “peace promotion” events alongside incidents of violence, Amatora Mu Mahoro sought to affirm the positive, peace-building impact of civic and electoral education, as well as offer a practical tool for civic educators to reach the areas of greatest need.

Amatora mu Mahoro is at the center of an array of election violence-prevention activities being organized by IFES around the 2010 elections. These projects include national civic education campaigns, theatre skits, community festivals, a powerful documentary film (Iyo Menya ("If Only I Had Known," 16 min., Kirundi subtitled in French)).

Data from other EVER programs worldwide is also available through IFES' Applied Research Center.

I would like to share an Indian experince. In 1996 when I was working as Chief Electoral Officer in the state of Uttar Pradesh which had nearly 125,000 polling stations and 100 million electors; the intelligence agencies predicted widespread voilence in the Provincial Assembly elections 1996 and the Election Commission of India had a tough task ahead.

Initiatives that helped prevent voilence were five fold. One, a list of all pending cases of electoral offences in the past elections were expedited. Two, an analysis of the past events of electoral voilence in the area was carried out and location of such events was plotted on a map to look for any trends and suitable remedial action. Three, the undertrials on bail were put in jails after cancellation of their bails. Four, a list of persons who could create problem on the poll day was prepared and all such identified persons were asked to give a commitment/ bond for good behaviour on the poll day - in total nearly 1,400,000 persons filed such bonds. And five, all the licenced arms were deposited and a campaign to sanitise the place was launched.  The action was taken in a transparent manner so that no party could claim discrimination. 

Result was a peaceful poll. I hope some of these steps can be replicated in some other parts of the world with suitable modifications.  

The I-C HUB was used in Nigeria elections (2003 & 2007), in South Africa (2009) and in Sudan (2010). The information part of the I-C HUB firstly undertakes a national mapping exercise of triggers and flashpoints. A tracking reporting mechanism that includes coordinates of incidents as well as identification of perpetrators as well as victims is designed. This is a very simple system that can be linked to a GIS tracking system to plot incidents on a geographical map. The information part also has an analytical part to it that unpacks the triggers, actors, victims etc. The communication part of the HUB is what we call the response component. Responses can be very focused and targeted or it can be name and shame type national forms of communication. It utilises local community networks, women and religious groups, feeds into EMB formal systems, can engage security agencies and political parties and where necessary use commercial media to disseminate the information.

The I-C HUB is what we like to refer to as an information gathering - response system. In this sense it can be likened to an early warning system but unlike an early warning system this system's response time should not be more than 72 hours (at least where it is possible). Beyond that the threat can become an explosive incident that often results in protracted conflicts and violence at times placing a heavy burden on the peacefulness and therefore the legitimacy of the electoral process itself. Idasa also works with other international agencies to role out this program. In 2010 it worked with ERIS to role out the tracking component of the HUB in Sudan. It is also in conversation with Intl Ideas and others to find a more coherent model that can be used as a common instrument to resolve conflict.