Nyheim (2009)^[1] argues that the future of early warning lies in adopting and capitalising on innovative information communication technologies for data collection, communication, visualisation and analysis. A comparative study of several IT based tools developed and used by different electoral stakeholders will help to illustrate common approaches in analysis and presentation of the electoral risk data.

Instituto Nacional Electoral (INE) Mexico is generating electoral risk maps by assessing risks associated with different factors, including drug trafficking, homicides, robbery, kidnapping, poverty and electoral complexity. Risk levels are projected on a scale of low, mid or high risk. Relevant data are generated by INE or obtained through cooperation and exchange with other government agencies. The risk analysis, presented through colour-coded geographical maps, is used by INE and other agencies to ensure conflict-sensitive planning of electoral operations and related security arrangements. See figure below:              

La Misión de Observación Electoral (MOE) in Colombia is generating electoral risk maps by analysing the rough data sets, some of which are produced and made available by the government. MOE undertakes statistical tests to establish regions in which data deviates. Data sets might include voter turnout figures. Risk analysis, presented through colour-coded geographical maps, is used to inform the broader public, including academics and news agencies, and the government about electoral risks. The established quality of the analysis has influenced the government’s actions and parliamentary discussions. See figure below:       

>

Centre for Monitoring Electoral Violence (CMEV) in Sri Lanka is generating online maps of election-related incidents collected by civil society networks. Factors observed include incidents relating to voter intimidation, campaigning irregularities, violation of electoral procedures, threats against electoral monitors and attacks on candidates. CMEV is presenting this data by placing the static markers on a Google Map. Markers point to a geographical location where incidents have taken place. Information is shared publicly, through media and websites, with the intention to stimulate government action. See below figure:     

USHAHIDI is an online platform available as a global public facility. The UCHAGUZI version is customised to gather information on election-related conflicts and violence. The platform uses Google Maps to present the data on election-related incidents collected through crowdsourcing. Data can be presented in the form of static markers, pointing to individual incidents, or through the aggregate numbers. The platform is used by a number of civil society organisations to inform public or organisations responsible for conducting elections. See figure below: 

International IDEA’s ‘Electoral Risk Management Tool’, which is being tested in several countries, is a desktop application that allows the user to create a country model and populate it with the relevant factors. Risks associated with the model factors can be presented in colour-coded maps, trend charts, static markers and aggregated figures shown on a map. The tool is designed to enhance the early warning and violence prevention capacity of organisations mandated to organise credible and peaceful elections, such as EMBs and security sector agencies (SSAs), as well as other interested organisations. The tool has been offered as a global public good since 2013. See below figure: 

Figure Example of a colour-coded map with static markers and cumulative factors, accompanied by a trend chart.

This comparative overview portrays the main IT-based methods for presenting electoral early warning analysis and highlights their advantages, disadvantages and synergies:

Geographical colour-coding is used to present risks associated with given factor(s) in different administrative or geographical regions. Data are presented on a map, and risk levels are coded in colour. This type of risk mapping is useful for communicating complex risk concepts in a simple and understandable way, but the disadvantage is that it allows only for a snapshot analysis. If the data are collected in successive rounds, geographical colour-coding cannot provide an insight into changing trends. This method is therefore particularly useful for illustrating risks associated with structural factors.

Trend charting is a widely used analytical methodology that is applied less often in the field of electoral violence early warning. Trend charts are useful in providing insight and communicating information about the dynamic factors whose risk properties may frequently change. The disadvantage is that it requires methodologically consistent and frequent data collection, which entails skills, time and resources. However, the combination of trend charting and geographical colour-coding methodologies offsets the deficiencies.

Geographical static markers are valuable analytical indicators for presenting the dichotomous variables on geographical maps. Markers pinpoint the location of an incident, enabling focused mitigation action and prevention of future incidents. The density of static markers in the region will be indicative of the risk levels and the level of action required. The disadvantage of this method is that it records events that have already happened. Furthermore, the presentation of a large number of static markers on small size maps may be messy.

Aggregated figures presented on geographical maps are used to point to the dimension of a factor, for example the aggregate number of incidents in a given region. This methodology may be used to aggregate, and numerically present, incidents recorded through static markers and thus improve and simplify their presentation.

 Early warning methodologies need ongoing improvement, with guidelines put in place for evaluating the effectiveness of different early warning measures. Infrastructure for peace (I4P, elaborated in Part III 1) adds the critical dimension needed for maximising efficiency of electoral violence early warning, prevention and mitigation. On the other hand, electoral violence early warning methodologies and tools can be designed with flexibility to fit different national and electoral contexts; to appeal to diverse pool of users; and to process different types of data close to real time. Such tools can be the cement which will strengthen these complex peacebuilding structures.