The Long-Run Relationship between House Prices and Income: Evidence from Local Housing Markets

TLDR

House‑price and income cointegration is widely debated, yet standard tests have low power, especially in small samples. The study applies panel‑data cointegration tests with a bootstrap correction for cross‑city shocks to 95 metro areas over 23 years. The analysis finds no evidence of cointegration, even with these powerful tests, suggesting that the commonly used error‑correction specification may be inappropriate.

Abstract

Many in the housing literature argue that house prices and income are cointegrated. I show that the data do not support this view. Standard tests using 27 years of national-level data do not find evidence of cointegration. However, standard tests for cointegration have low power, especially in small samples. I use panel-data tests for cointegration that are more powerful than their time-series counterparts to test for cointegration in a panel of 95 metro areas over 23 years. Using a bootstrap approach to allow for cross-correlations in city-level house-price shocks, I show that even these more powerful tests do not reject the hypothesis of no cointegration. Thus the error-correction specification for house prices and income commonly found in the literature may be inappropriate.