Citation Context

...ed e¤ects has been considered in Lee and Yu (2007a) for a spatial dynamic panel model with large T . In a group setting with groupsxed e¤ects, a similar transformation can eliminate the group e¤ects (=-=Lee et al., 2008-=-). 5However, our modi…ed likelihood is not one of those which could be constructed from their formulas. 2 same representation. By the transformation (In0Wn), the data generating process (DGP) of Kap...

Citation Context

...r space is therefore Λ = (−(m1 − 1), 1). Example 2 (Complete Bipartite Model). In a complete bipartite graph the n observational units are partitioned into two groups of sizes p and q, say, with all individuals 6If the assumption that W has at least one negative eigenvalue were not satisfied it would be appropriate to set λmin = −∞. Note that if all eigenvalues of W are real the fact that W has at least one negative eigenvalue is implied by the assumption that tr(W ) = 0. 6 within a group interacting with all in the other group, but with none in their own group (e.g., Bramoulle et al., 2009, Lee et al., 2010). For p = 1 or q = 1 this corresponds to the graph known as a star, a particularly important case in network theory (see Estrada, 2011). The adjacency matrix of a complete bipartite graph is A := [ 0pp ιpι ′ q ιqι ′ p 0qq ] . The corresponding row-standardized weights matrix is W = [ 0pp 1 q ιpι ′ q 1 p ιqι ′ p 0qq ] . (2.2) This is not symmetric unless p = q. Alternatively, A can be rescaled by its spectral radius, yielding the symmetric weights matrix W = (pq)− 1 2 A. (2.3) We refer to the SAR models with weights matrix (2.2) or (2.3), as, respectively, the row-standardized Complete Bipartit...