【摘要】This paper sets up a nested random effects spatial autoregressive panel data model to explain annual house price variation for 2000–2007 across 353 local authority districts in England. The estimation problem posed is how to allow for the endogeneity of the spatial lag variable producing the simultaneous spatial spillover of prices across districts together with the nested random effects in a panel data setting. To achieve this, the paper proposes new estimators based on the instrumental variable approaches of Kelejian and Prucha (1998) and Lee (2003) for the cross-sectional spatial autoregressive model. Monte Carlo results show that our estimators perform well relative to alternative approaches and produces estimates based on real data that are consistent with the theoretical house price model underpinning the reduced form.

【关键词】House prices; Panel data; Spatial lag; Nested random effects; Instrumental variables; Spatial dependence

【文献来源】Baltagi B H;Fingleton B;Pirotte A.?Journal of Urban Economics.2014