We examine the synchronization of housing prices across 97 districts (Gu)and cities (Si) by using a dynamic factor model. The model allows fluctuationsin housing prices to be decomposed into two latent factors: (1) a nationalfactor affected by economic conditions and economic policy and (2) a regionalfactor driven by region-specific characteristics. We find that the national factorwas the main driver of housing cycles for 2004-2009. However, since 2015,the importance of regional factors has increased significantly in explainingfluctuations in housing prices. The increase in the relative importance of theregional factors appears in the non-capital region. In a panel regression, theunemployment rate and net population outflow explain housing price varianceaccounted for by the regional factors. The mortgage rate and inflation ratehave little impact on the variance explained by the regional factors. Accordingto our results, it would be required that the effective target of housing policiesbe set up at the small administrative district level.