Abstract:We provide an analysis that might help distinguish rationally justified movements in house prices from potentially non-rational movements, using a two-sector business cycle model, in which investment in housing is subject to collateral constraints. A large portion of the evolution of U.S. house prices during the past 20 years can be reproduced when expectations of future income growth as published in surveys are used as an input into the model. Changes in growth expectations translate into corresponding changes in house prices, since the value of housing must be linked to expected aggregate income. Only since about 2005 do actual and model-implied house prices clearly diverge, calling for explanations not based on economic fundamentals.