Cem Mert Dalli

I am a PhD student in the Department of Political Science at Aarhus University and a member of the AI & Global Development Lab. As a quantitative social scientist, I am keen on bringing new computational and AI-assisted methods to address complex problems across the social sciences.

My research focuses on developing causal inference methods for observational settings where interventions are spatio-temporally connected, where treatment intensity and timing evolve across space and effects propagate through spillovers and interference. In my dissertation, I mainly work on combining stochastic intervention frameworks with Bayesian spatial models to estimate these effects. A core strand of my work integrates Earth Observation and computer vision into causal design, using satellite imagery and representation learning to measure local conditions and address high-dimensional spatial confounding.

Substantively, I focus on the political economy of development and the long-run consequences of armed conflict. I also maintain a strong interest in historical political development, adopting a longue durée perspective through my work with the Political Leaders through Time (PLT) project, where I study state formation and political leadership.

Before starting my PhD, I worked as a Data Manager at the Quality of Government (QoG) Institute, University of Gothenburg, where I remain on leave. My work there focused on the curation and maintenance of major cross-national governance datasets.