GIPPS: Election violence and election observation data

EISA in partnership with the University of Michigan and the University of Witwatersrand (Wits) is working on a project to enhance the utility of data collected by election observation groups through the refinement of election observation methodology. The, project, called "ObSERV" introduces random data sampling and geo coded data to create evidence based data sets that will enable academics and practitioners to analyse the link between elections and violence or peace.

2021 marks the third year of EISA'S efforts towards improved access and utility of EISA EOM data to the studies of African politics. Since its adoption in 2018, the project has led to significant innovations in EISA'S EOM methodology and has further strengthened EISA's level of influence in election related research and studies. GIPPS collaborates with EPP on a yearly basis, providing targeted mutual support in where EISA deploys international observers.

More information about the project, including the introductory video, can be found here.

2021


CCNY Michigan Conflict Prevention programme: Phase 2

The CCNY Phase 2 programme was able to resume during EISA's EOM to The Gambian elections in December 2021. Adding a security specialist to the team of analysts collecting data proved a welcome and valuable addition to the data collection, with a comprehensive, constituency-based security assessment tool developed with a security specialist (retired Navy Captain and Advocate Dean Gillespie). This tool was first trialled during the South African Local Govt elections on 1 November 2021, and provided the South African Policy Service Task Team with better information on the ground than their own "sources" were delivering in real time using the EISA Popola system. With a full roll-out, the system appears capable of acting as intended, i.e as an early warning system for escalated risks at local level. This tool, mapped on top of EISA's existing EOM data collection, could provide a unique and significantly improved form of analysis of elections and conflict. Plans to further develop the tool and approach under the programme will continue in the Kenya 2022 election and will be able to be used by EISA when conducting further EOMs during the remaining period of the grant and going forward beyond this grant.

CCNY Michigan Conflict Prevention programme

Phase 2 of this programme has been heavily impacted by COVID19 and the lack of EISA Election Observation Missions during the first 12 months of the project. Refinements to the tools and testing of various adaptations in methodology took place during the reporting period, as well as plans for testing of these new innovations in the field during the South African Local Government Elections in November and the Gambian national elections in December.

GIPPS assisted a civil society consortium lead by Dr Cheryl Hendricks under the umbrella name "Women's Election Consortium for Peace" (WEMP) to conduct election day peace assessments under this project, including training, assistance with data collection, the use of the EISA Popola instrument and advice on deployment and election observation. As part of EISA's agreement to support this process, the project is entitled to make use of election day data collected to analyse and develop election and security assessments further.

CCNY/ Michigan/ Wits Peacebuilding in Election Phase 1 & 2

In partnership with University of Michigan, GIPPS has started the next phase of the Peacebuilding in Elections project with a new attempt to understand historical electoral data collected by EISA since 2013. This data is being analysed through means of machine learning and includes all EISA data collected on IEOMs since the system was first digitised. The aim of this process will be to identify which (if any) of the indicators collected by observers have remained consistent over time and which have become more or less relevant as elections and election observation have evolved to cover them. The output of this activity will allow EISA to improve its election assessment methodologies and will likely support parallel processes such as the Electoral Integrity project and revision of PEMMO that are currently being undertaken by other EISA programmes.

2020


CCNY/ Michigan/ Wits Peacebuilding Election Phase 1 & 2

GIPPS completed a successful pilot project supported by the Carnegie Corporation of New York (CCNY) in partnership with Michigan and Wits Universities which aims to collect more localised election data to better understand local causes of election violence. Phase 1 was completed within budget and delivering on all outputs in September 2020. A Phase 2 proposal was submitted to CCNY in August 2020, and approved with a start date of 1 October 2020, running for 24 months. The project will build on the initial pilot's successes and continue developing more scientific approaches to election day data collection during EISA EOMs.

Post-graduate workshop

The Carnegie project examining local patterns of election related violence in Africa is in its final stages. In this regard, EISA held a research workshop for postgraduate student in partnership with the University of Witwatersrand and the Eduardo Mondlane University. A total of nine students were trained on concepts and theories of election related violence, existing datasets for election violence and introduced to data collected by EISA EOMs during the 2019 elections in South Africa and Mozambique. The expectation is that students will use the knowledge gained from the workshop to produce publishable policy briefs that primarily use EISA EOM data. The inclusion of students in African universities builds towards the projects objective of developing a bigger body of primary data that will be available to African researchers looking at peacebuilding in Africa.

Fact Finding mission to Ethiopia

21 - 23 Jan 2020

The Carnegie project examining local patterns of election related the violence in Africa is in its final stages. In this regard, EISA held a research workshop for postgraduate student in partnership with the University of Witwatersrand and the Eduardo Mondlane University. A total of nine students were trained on concepts and theories of election related violence, existing datasets for election violence and introduced to data collected by EISA EOMs during the 2019 elections in South Africa and Mozambique. The expectation is that students will use the knowledge gained from the workshop to produce publishable policy briefs that primarily use EISA EOM data. The inclusion of students in African universities builds towards the projects objective of developing a bigger body of primary data that will be available to African researchers looking at peacebuilding in Africa.

2019


Local patterns of election related violence
  • Work is continuing on the project examining local patterns of election related the violence in Africa. An adjustment to the existing EISA election observation methodology was trialed in South Africa during the 2019 elections, with a randomised, cluster sampling approach proving both practically feasible and statistically useful. The system, developed in partnership with Professor Rod Alence from the University of the Witwatersrand's International Relations department, was able to not only randomise the stations observer teams should visit, but even suggest the most time efficient route between these stations based on geo-location technology. The system will be further trialed in Mozambique and preliminary data will be analysed by post-graduate students at the University of the Witwatersrand, under Rod Alence, as well as researchers at Michigan to develop a bigger body of primary data available to African researchers looking at peace-building in Africa.
  • GIPPS, in collaboration with the University of Michigan and the University of the Witwatersrand, implemented an ambitious pilot project during the South African and Mozambique elections to develop more systematic and academically rigorous sampling processes during the deployment of election observers during EISA election observation missions (EOMs). The pilot study, supported by the Carnegie Corporation of New York (CCNY), demonstrated its applicability in field conditions under very different circumstances (SA, Mozambique). It is envisaged that all future EISA EOMs will be able to adopt this systematic sampling approach, which allows for more accurate and representative data from its observer team reports and observations, and for enhanced co-operation with academia in the analysis of election trends.

2018


Local Patterns of Election-Related Violence

Oct 2018

Work began on the Carnegie Corporation-funded project examining local patterns of election related violence in Africa. The project got off to a slow start because of administrative issues with EISA’s partner, the University of Michigan. The GIPPs team, along with Prof Rod Alence from the University of the Witwatersrand, observed the coordination of the Mozambique local government elections in October 2018.

2017


Citzen Inclusion in Transitional Processes in Kenya and South Sudan

In partnership with the National Democratic Institute (NDI) and the Oslo Centre, EISA developed a method of assessing citizens’ sense of inclusion in transitional processes in post-conflict democracies. The project hypothesises that post-conflict peace processes are almost entirely determined by political elites, with only superficial attention paid to the views of ordinary citizens, who have a direct stake in the outcome of peace processes but are often excluded from negotiated agreements. The aim of the research, conducted through focus groups in disparate regions in Kenya and South Sudan, is to develop a comprehensive assessment tool that can be adapted to other post-conflict environments and assess citizens’ perceptions of inclusion in processes in which they have a stake. The project will run until the end of 2018.

Donor

Carnegie Corporation of New York