TY - JOUR
T1 - Spatial autoregressions with an extended parameter space and similarity-based weights
AU - Rossi, Francesca
AU - Lieberman, Offer
N1 - Publisher Copyright:
© 2023 Elsevier B.V.
PY - 2023/8
Y1 - 2023/8
N2 - We provide in this paper asymptotic theory for a spatial autoregressive model (SAR, henceforth) in which the spatial coefficient, λ, is allowed to be less than or equal to unity, as well as consistent with a local to unit root (LUR) model and of the moderate integration (MI) from unity type, and the spatial weights are allowed to be similarity-based and data driven. Other special cases of our setting include the random walk, a model in which all the weights are equal, the standard SAR model in which λ<1 and the similarity based autoregression in which λ=1 and data do not display a natural order. As the norming rates for the asymptotic theory are very different in the λ<1 - compared with the λ=1 and LUR cases, we resort to random norming that treats all cases in a uniform manner. It turns out that standard CLT results prevail in a large class of models in which the infinity norm of the inverse of the weighting structure that characterizes the reduced-form process is Onγ, γ∈[0,1), and is non-standard in the case γ=1. We use a shifted profile likelihood to obtain results which are valid for all cases. A small simulation experiment supports our findings and the usefulness of our model is illustrated with an empirical application of the Boston housing data set in which the estimate of λ appeared to be very close to unity.
AB - We provide in this paper asymptotic theory for a spatial autoregressive model (SAR, henceforth) in which the spatial coefficient, λ, is allowed to be less than or equal to unity, as well as consistent with a local to unit root (LUR) model and of the moderate integration (MI) from unity type, and the spatial weights are allowed to be similarity-based and data driven. Other special cases of our setting include the random walk, a model in which all the weights are equal, the standard SAR model in which λ<1 and the similarity based autoregression in which λ=1 and data do not display a natural order. As the norming rates for the asymptotic theory are very different in the λ<1 - compared with the λ=1 and LUR cases, we resort to random norming that treats all cases in a uniform manner. It turns out that standard CLT results prevail in a large class of models in which the infinity norm of the inverse of the weighting structure that characterizes the reduced-form process is Onγ, γ∈[0,1), and is non-standard in the case γ=1. We use a shifted profile likelihood to obtain results which are valid for all cases. A small simulation experiment supports our findings and the usefulness of our model is illustrated with an empirical application of the Boston housing data set in which the estimate of λ appeared to be very close to unity.
KW - Quasi-maximum-likelihood
KW - Similarity function
KW - Spatial autoregression
KW - Weight matrix
UR - http://www.scopus.com/inward/record.url?scp=85148739325&partnerID=8YFLogxK
U2 - 10.1016/j.jeconom.2022.11.010
DO - 10.1016/j.jeconom.2022.11.010
M3 - ???researchoutput.researchoutputtypes.contributiontojournal.article???
AN - SCOPUS:85148739325
SN - 0304-4076
VL - 235
SP - 1770
EP - 1798
JO - Journal of Econometrics
JF - Journal of Econometrics
IS - 2
ER -