Sababi

At Sababi, we conduct our own research and implement studies led by other researchers that demand academic-grade data collection even in difficult-to-study regions.

Social and Justice Attitudes of Traditional Elders in Somalia and Somaliland

In Somalia, clan elders are the main arbiters of interpersonal and intergroup disputes. They are often called upon to resolve intra-household conflicts, act as judges and advocates in minor civil cases, enact punishment and restitution in murder cases, and everything in between. Their role as mediators in inter-clan conflict has elevated their importance in politics — elders are essentially responsible for selecting parliamentary representatives for their clan in Somalia, and elders serve in the House of Elders (or Guurti) in Somaliland. This project is the first large-scale survey of clan elders. Our goal is to understand how elders view issues of justice and dispute resolution and whether there are avenues for reform of Somali customary justice (xeer) that would enhance access to justice for marginalized groups without undermining the only justice institution that serves most Somalis. In addition, the project captures political and social attitudes of clan elders, maps the structure and hierarchy of the elder system, and provides empirical data on the size of clans in different locations.

Partners: This project is being funded by the Knowledge Platform for Security & Rule of Law (KPSRL) through the Knowledge Management Fund.

Linguistic Bias in African Survey Data

Large-scale, representative surveys undertaken in sub-Saharan Africa face many challenges, including poor road systems, large geographic spaces, and low population densities. One largely unexamined challenge is the linguistic diversity of many African countries, which prompts researchers to either translate surveys into many additional languages or systematically exclude respondents who do not speak one of a country’s predominant languages or the lingua franca. In some countries, upwards of 20-30 percent of the population can be excluded using decision rules for translation that are common among even reputable survey research firms. This project estimates the potential for bias in estimates of population outcomes when applying standard survey practices, using data from Ethiopia, Mozambique, Senegal, Sierra Leone, South Africa, and Zambia.