Thessaloniki Econometrics Summer School (Th.E.S.S.)

Professor Vasilis Sarafidis (Brunel University London)

Venue:
University of Macedonia, Thessaloniki, GREECE

Brief Course Description:

Understanding spatial and network interactions is becoming a core requirement for researchers working with economic, social, and environmental data. Outcomes in these settings are rarely determined in isolation; rather, they are shaped by complex interdependencies across space and networks, linking economic agents—such as firms, financial institutions, regions or countries—to their peers. For example, firms adjust investment decisions in response to innovations by industry peers, countries engage in trade relationships influenced by shared policies and technological diffusion, and financial systems are connected through credit exposures and derivative contracts. Understanding the structure of these interdependencies is critical, as it determines how shocks, policies, and behaviours propagate through economic systems. Neglecting such interconnections can lead to biased inference and misinformed policy.

This course introduces key concepts and methods for analysing spatial and network systems using panel data analysis. It combines theoretical foundations with practical tools for modelling interactions and spillovers across space. Topics include spatial models with fixed and interactive effects, estimation of homogeneous and heterogeneous coefficients, and the identification of spatial and network effects. A key focus of the course will be on estimating latent network structures through high- dimensional econometric techniques. Participants will learn how to quantify and interpret direct, indirect (spillover), and total effects, how to evaluate key network metrics such as density, homophily, in-degree, and out-degree, and how to visualise network structures using graphical tools that support both interpretation and presentation. Applications will be drawn from economics and finance, with most estimation carried out in Stata.