Data innovation in conflict zones: Enabling insights without endangering lives

On April 6, AidData and the W&M Whole of Government Center for Excellence will host a panel on evaluating development policy and interventions when on-the-ground data collection is difficult, dangerous, or infeasible.

March 30, 2023
Madeleine Walker
Refugees visit a mine awareness exhibit at a refugee camp. Photo by Pedram Pirnia for CIDA/ACDI via Flickr, licensed under (CC BY-NC-ND 2.0).

Refugees visit a mine awareness exhibit at a refugee camp. Photo by Pedram Pirnia for CIDA/ACDI via Flickr, licensed under (CC BY-NC-ND 2.0).

How can researchers more safely and easily conduct rigorous analysis in conflict zones? Household surveys, focus groups, and other tools typically used in development research pose risks to both those collecting data and the respondents themselves in areas with active violent conflicts. The alternative, avoiding research in these areas, is also bleak. The UN estimates that two billion people, or 25% of the world’s population, now live in conflict-affected areas. Understanding conditions there is necessary to efficiently target aid and assist international and domestic leaders in developing effective and equitable policy to promote peace. 

The 3rd Annual National Security Conference will be held on April 6, 2023 at William & Mary. Attendance is free and open to the public. Please register to attend in-person here.

Fortunately, methods exist that can help close the data gap. Innovations in remote sensing and machine learning can complement or replace survey data in unsafe areas. AidData’s geospatial impact evaluation (GIE) methodology incorporates innovations in these fields to measure the impact and cost-effectiveness of interventions, even in the absence of traditional on-the-ground data collection. This methodology is particularly useful for creating insights for policymakers in conflict-affected areas.

One such example is AidData’s recent GIE on landmine clearance in Afghanistan. To bypass challenges in on-the-ground data collection, AidData researchers used satellite imagery to develop measures of land use and economic activity. Land use classifications such as built-up land, farmland, and grassland were developed through machine learning, and nighttime lights were used as a proxy for economic activity in electrified areas. In another GIE in Mali, measures of crop greenness as seen from satellite imagery were used to evaluate crop yields in a large-scale irrigation intervention. 

While exciting, remote sensing is not the only method for evaluating development in conflict zones. There are other tools—from remote surveys via phone or text messages, to news and social media aggregation, to data collected by corporations—can aid researchers in evaluating interventions in these conflict-affected areas. For example, a recent paper in The Review of Economics and Statistics used geolocated cell-phone data for users of M-Paisa, a mobile money platform widely used in Afghanistan, to estimate how being near violent events impacted how individuals adopted and used mobile money.

As it becomes more difficult to reach individuals in countries such as Afghanistan, innovative research to guide policy-making is of increasing importance, especially for vulnerable populations. For example, women in Afghanistan are quickly becoming one of the most isolated demographics globally due to Taliban rule. Extreme weather events and political conflict will continue to threaten on-the-ground data collection globally. However, continued data innovation and cooperation between researchers and policymakers will play a valuable role in crafting informed humanitarian and diplomatic responses.

GIEs, remote sensing, and other remotely collected and accessed data sources have the potential to be powerful tools for policy-making in conflict zones. To discuss these innovations, their value for policymakers, and their limitations, AidData is excited to co-host a panel discussion on April 6 with policy and private sector experts to identify how our current and future research can best inform policy in Afghanistan specifically and other conflict zones more generally.

The discussion, “Data innovation in conflict zones: Collecting and utilizing data to guide climate resilience and security policy in Afghanistan,” will be part of The 3rd Annual National Security Conference hosted by the William & Mary Whole of Government Center of Excellence. The panel will include Dr. Ariel BenYishay and Dr. Rachel Sayers of AidData, as well as Afghanistan policy experts The Honorable Ronald Neumann, the former U.S. Ambassador to Afghanistan, and Dr. William Byrd, former World Bank country manager for Afghanistan. Tim Carroll, Weather and Climate Portfolio Lead at Microsoft, will provide a private sector perspective.

Madeleine Walker served as a Junior Data Analyst at AidData from 2021 to 2023.