Essays on Spatial Panel Data Models with Common Factors

Download or read book Essays on Spatial Panel Data Models with Common Factors written by Wei Shi and published by . This book was released on 2016 with total page 172 pages. Available in PDF, EPUB and Kindle. Book excerpt: My dissertation research addresses issues in spatial panel data models, which study the interactions of economic units across space and time. Individuals interact with their neighbors and the outcomes are interdependent. The strength of the interaction depends on the distance between the individuals, which can be based on geography or constructed from economic theory. Accounting for spatial interactions allows one to quantify both the direct effect of a variable and its indirect effect through impacting neighbors. However, two issues often arise. First, spatial dependence can be alternatively generated from common unobserved factors (e.g. economy-wide shocks) where neighbors have similar responses. Second, the distance between economic units can be endogenous, and this will in fact be the case if the distance is constructed from variables that correlate with disturbances in the outcomes. The first chapter studies the estimation of a dynamic spatial panel data model with interactive individual and time effects with large n and T. The model has a rich spatial structure including contemporaneous spatial interaction and spatial heterogeneity. Dynamic features include individual time lag and spatial diffusion. In a standard two way fixed effects panel regression model, the unobservables contain an individual specific but time invariant component, and a component that is time variant but common across individuals. We generalize this model by allowing the interaction between time effects and individual effects. This chapter provides a tool for empirical researchers to guard against attributing correlated responses to common time effects as spatial effects. The interactive effects are treated as parameters, so as to allow correlations between the interactive effects and the regressors. We consider a quasi-maximum likelihood estimation and show estimator consistency and characterize its asymptotic distribution. The Monte Carlo experiment shows that the estimator performs well and the proposed bias correction is effective. The second chapter proposes a unified approach to model endogenous spatial dependences while accounting for common factors. The spatial weights matrices are constructed from variables that may correlate with the disturbances in the outcomes. We make minimal assumptions on the distributions of the factors and follow a fixed effects approach. We provide conditions under which the quasi-maximum likelihood estimator is consistent and asymptotically normal, under the asymptotics where both the cross section and time dimensions become large. The limiting distribution is normal but may not be centered for the estimates of the spatial interaction coefficient and the variances. An analytical bias correction is proposed to improve the inference. The Monte Carlo simulations demonstrate good finite sample properties of the bias corrected estimator. We illustrate the empirical relevance of the theory by applying the method to analyze the effect of house price dynamics on reverse mortgage origination rates.