Spatial Spillovers Across Housing Market Cycles

We develop a theory of housing price spillovers based on price references, whereby learning about unobservable location attributes gives rise to spatial contagion in house prices. Our model predicts that expected price changes drive the strength of housing price spillovers across market cycles. While the theory predicts a selfreinforcing upward price spiral during the boom, spatial spillovers do not amplify downward movements in prices during the downturn. It also justifies the use of Spatial Autoregressive (SAR) models with row-normalized weight matrices based on geographical proximity. The model’s prediction is supported by U.S. data from recent housing market boom (2018-2023) and bust (2007-2012).