A0552
Title: Semi-parametric estimation of a matrix exponential spatial dynamic panel specification
Authors: Abhimanyu Gupta - University of Essex (United Kingdom)
Shuting Hou - East China University of Science and Technology (China)
Xi Qu - Shanghai Jiaotong University (China)
Jiajun Zhang - Shanghai University of International Business and Economics (China) [presenting]
Abstract: A semi-parametric estimation of a matrix exponential spatial dynamic panel specification (MESS) is investigated with two-way fixed effects, which can also be directly extended to the spatial autoregressive (SAR) specification. The spatial weight matrices associated with the dependent variable, its time-lag term, and the disturbance term are all treated as unknown functions of exogenous distance measures. Using a sieve-based generalized method of moments estimator (GMME), a feasible optimal GMME (FOGMME) is developed under unknown heteroskedasticity. A more efficient estimator is then derived: The feasible best GMME (FBGMME) with the oracle property. Furthermore, the asymptotic distributions of the proposed estimators are established, and their finite-sample properties are evaluated through Monte Carlo simulations. Finally, the model is applied to investigate spillover effects in two distinct empirical analyses: Witch killings and service sector sizes.
