Agripath sought to evaluate the use of digital tools for sustainable behavior change in five countries.
Scientific evaluation of digital tools is crucial because development organizations increasingly invest in such cost-effective means
to implement the UN’s Sustainable Development Goals.
Our team investigated if digital development can scale the adoption of sustainable land management practices to make farmers resilient to climate change.
Two models for behavior change were considered: an advisory model, grounded in adoption-diffusion theory to guide conventional agriculture extension,
and a deliberative model, grounded in communicative action theory to address structural barriers to diffusion.
We drew on primary data from a digital advisory service deployed in five countries, and two large-scale field experiments in India and Nepal.
Our findings suggest that the widely applied advisory model is necessary but not sufficient to scale sustainable farming.
Instead, decision-makers should address structural and institutional barriers to adoption.
We demonstrated how this could be achieved with online deliberation, which could democrate development
to resolve disagreement around the removal of such barriers.
Enemies Within: Labeling Defectors to Rival Authorities
How do allegiances shift from incumbent rulers to rival authorities?
From rebel groups to state authorities, political actors take extreme measures to ensure the loyalty of their subjects.
Those who are labeled as defectors or 'traitors' may be ostracized, imprisoned, tortured or killed.
But the expected consequences of such punishments remain disputed by an established body of scholarship on state repression,
civil wars, and criminal behavior. Either the labeling of people as defectors deters undesirable behavior,
leading to widespread conformity with rules set by authorities.
Or it intensifies defection from political orders, as the labeled shift support to rival authorities who support their behavior.
This project relies on a mixed-methods approach to investigate shifts between conformity and defection in the former
German Democratic Republic and the Occupied Palestinian Territories, drawing on archival research, an original lab experiment, semi-structured interviews,
existing survey data, and an empirically validated computational model.
Modeling Early Risk Indicators to Anticipate Malnutrition
MERIAM was a four-year project funded by the UK government, which brought together an inter-
disciplinary team of experts across four consortium partners: Action Against Hunger, the Graduate
Institute of International and Development Studies, John Hopkins University, and the University of Maryland.
MERIAM’s primary aim was to develop, test and scale-up models to improve
the prediction and monitoring of undernutrition in countries that experience frequent climate
and conflict related shocks.
Among my tasks were the co-development of an evidence-driven computational model,
implementation of the model in Python, data construction and analysis, the coordination of the research project,
the implementation of expert interviews, and of focus group discussions
and a survey during two weeks of fieldwork in Uganda and Kenya.
The project evaluated the association between development aid and the likelihood,
escalation, severity, spread, duration, and recurrence of violence, spanning the phases before, during,
and after conflict. It was funded through a research grant awarded to the Center for International Development and Conflict Management
at the University of Maryland.
I assisted on the project during its final phase.
My primary task was to construct and visualize geo-coded resilience indices from various conflict datasets.