Data Collection in Fragile States

Quality statistics are critical for development interventions to be effective. However, they are hard to obtain in fragile situations as fragility affects the ability to collect data in many ways. From exposure to violence or to other dangers, such as disease, collecting information isn’t possible using traditional means without putting enumerators at risk. There is also the association of conflict with poor-quality roads, inadequate telecommunication infrastructure and, at times, populations that are hostile to representatives of the central government.

Fragile situations can also complicate data collection in other ways such as people being displaced or the lack of sampling frames. A new report, Data Collection in Fragile States, presents innovations developed to deal with common challenges. The report presents innovations, both methodological as well in data collection, to address the data gaps in fragile situations. Through examples, the book shows that it is possible to collect high-quality data in fragile settings and such data collection need not be expensive but also warns that technology is not a panacea for all data collection issues.

Additional Findings

• The proliferation of mobile phone networks and inexpensive handsets has opened new possibilities for data collection. Although they cannot replace face-to-face household surveys in all contexts, mobile phone surveys offer substantial benefits in specific circumstances and for specific data collection needs. For locations where mobile phone surveys are not an option, locally recruited, resident enumerators who live in the community can be used. Mobile phone surveys were used to collect information during the Ebola crisis in Sierra Leone, the drought in Nigeria, Somalia, South Sudan and Yemen, and were used to track the welfare of people displaced by the crisis in northern Mali. 
• Technological advances in geospatial data have the potential to change how survey data are collected. As geospatial technology has improved and become more widespread, costs have come down and the number of available tools has increased, making Geographic Information Systems (GIS)-based sampling approaches accessible to more users. To deal with the absence of sampling frames in the Democratic Republic of Congo and Somalia, satellite images and sophisticated machine learning algorithms were used to estimate population density and demarcate enumeration areas.

A camp for displaced people in the Central African Republic.

• Sampling in chaotic and fluid locations, where people are displaced or live In Internally Displaced Peoples (IDP) settlements requires a rethinking of traditional sampling approaches. Attention should also be paid to elicit truthful information from respondents. Approaches such as endorsement experiments, list experiments, and behavioral approaches can be used to ask questions about sensitive issues such as loyalty to controversial groups along with how to avoid strategic responses when respondents might expect benefits to be associated with certain answers. For instance, in South Sudan, different sampling approaches were tested in an IDP camp to shed light on their precision. 
• Video testimonials provide a cost-effective way to give outside audiences a perspective on the lives of survey respondents. The collection of household data is usually a passive process where respondents are asked pre-formulated questions. This constrains the respondents in sharing their own narratives and emphasizing what they feel is important. The video testimonials give a voice to the poor and help better understand the concerns of the poor as well as empower them to create a narrative that they own. In South Sudan, in addition to estimating the poverty data, video testimonials revealed what it was like to live in poverty. 
• A light-touch approach to generating periodic feedback can help provide information about the performance of donor projects. A new approach called Iterative Beneficiary Monitoring (IBM) provides a feedback system that focuses on a select set of issues to measure outcomes of projects. By keeping data collection focused, IBM facilitates timely data analysis and the rapid preparation of reports. The approach was introduced in Mali to provide feedback on projects about feeding in schools to distribution of e-vouchers and cash transfers.