ECON747: Spatial
Econometric Models and Methods, Term I 2024-25
In many social and economic activities, interactions
often occur among geographical units, economic agents, or social actors, which
generate spatial/social effects among these units or agents or actors,
e.g., neighbourhood effects, peer effects, spillovers, network
effects, social norms, externalities, conformity, imitation, contagion,
bandwagons, herd behaviour.
Spatial econometrics consist of a set of
econometric models and methods, proven to be very effective in dealing with
these issues. Applications are seen not only in specialized fields of regional
science, urban economics, real-estate and economic
geography, but also in more traditional fields of economics, finance, and
social sciences in general.
Spatial econometric models
extend the classical linear or panel data regression models by incorporating spatial lag, spatial error, and/or spatial
Durbin terms to capture spatial or social effects through weight or adjacency
matrices. This course introduces basic spatial econometric models including
spatial linear regression models, spatial panel data models, and dynamic
spatial panel data models, and the associated methods of estimation and
inference such as (quasi) maximum likelihood, M-estimation, and GMM. Common
tests for spatial and/or dynamic effects, e.g., LM tests, standardized LM
tests, and bootstrap LM tests are introduced. Empirical illustrations of the methods are presented using Matlab, Python or Stata.
|
Data Sets |
Lecture time & Venue: Friday 12:00-15:15, SR 4.1, SOE/SCIS
Consultation Hours & Venue: Thursday 10:00-12:00, at
Office: #5058 (New), SOE