An Introduction to Applied Spatial Econometrics (Hands-On with Stata)

A 4-hour intensive workshop

Most regression models assume your observations are independent. Spatial data break that assumption. Whether you study FDI across neighbouring economies, growth spillovers between regions, productivity clusters across cities, or carbon leakage between countries, ignoring spatial dependence biases your estimates and your conclusions.

This workshop takes you from the intuition (Tobler's First Law and why OLS fails) all the way to publication-ready analysis in Stata. You will learn how to build and justify a spatial weight matrix, test for spatial dependence with Moran's I and LM diagnostics, and navigate the full family of spatial models from SAR and SEM to the Spatial Durbin specifications. Most importantly, you will master the single most misunderstood part of the field: why spatial coefficients are not marginal effects, and how to correctly report direct, indirect, and total spillover effects.

By the end of the session, you will be able to run a complete spatial workflow on your own data, choose the right model, defend your weight matrix against reviewers, and interpret spillovers with confidence.

You will learn to:

• Construct and normalise spatial weight matrices (contiguity, distance, economic)
• Detect spatial dependence using Moran's I, LISA, and Lagrange Multiplier tests
• Estimate and choose between SAR, SEM, SDM, SDEM, SAC, and SLX models
• Decompose results into direct, indirect, and total effects in Stata
• Extend the analysis to spatial panel data

Who should attend: Graduate students, researchers, and analysts in economics, business, and the social sciences who work with regional, country, or firm-level data and want to add spatial methods to their toolkit. Familiarity with basic regression is assumed; no prior spatial experience needed.

Format: 4 hours (2-hour sessions with a break), concept plus live Stata demonstration and a guided hands-on lab. Bring a laptop with Stata installed.