摘要：We present a study of the effects of macroeconomic shocks on housing prices in the Western United States using quarterly state level data from 1988:1 – 2007:4. The study contributes to the existing literature by explicitly incorporating locational spillovers through a spatial econometric adaptation of vector autoregression (SpVAR). The results suggest these spillovers may Granger cause housing price movements in a large number of cases. SpVAR provides additional insights through impulse response functions that demonstrate the effects of macroeconomic events in different neighboring locations. In addition, we demonstrate that including spatial information leads to significantly lower mean square forecast errors.
