Housing and the US Business Cycle: Price or Quantity?

This paper analyzes the relationship between housing and the business cycle in the US. We start with constructing and estimating a baseline Markov-switching common factor model, and the importance of housing variables is then evaluated by comparing the results for the baseline model with those for the extended models augmented with the housing market variables. Two strands of extended models are considered: one in which housing variables directly affect the individual macroeconomic series and thereby the business cycle, and the other in which they affect the probabilities of transitions between expansion/recession phases. Results for the first extension support the importance of the housing quantity variables as argued by Leamer (2007): the growth in housing permits and housing new housing starts exert significant effects on the macroeconomic indicators and consequently the business cycle, but house price growth does not. For the second extension, however, we find that house price growth contains useful information beyond what is in the housing quantity variables. More specifically, while higher quantity growth affects which phase will prevail in the next period only when the economy is currently in the expansion, higher house price growth is informative whether the economy is currently in recession or expansion.