Housing Bubble in Japan and the United States

Japan and the United States have experienced the housing bubbles and subsequent collapses of the bubbles in succession. In this paper, these two bubbles are compared and the following findings are obtained. Firstly, upon applying twenty years of past data from Japan to the "repeat-sales method" and the "hedonic pricing method", which are representative methods for calculating house prices, it was found that the timing at which prices bottomed out after the collapses of the bubbles differed depending on the two methods. The timing for bottoming out as estimated by the repeat-sales method delayed when compared to the estimate using the hedonic pricing method, by 13 months for condominiums and by three months for single-family homes. This delay is caused by the depreciation effect of building not being processed appropriately by the repeat-sales method. In the United States, the S&P/Case-Shiller Home Price Indices are representative house prices indices, which use the repeat-sales method. Therefore, it is possible that the timing for bottoming out is estimated to be delayed. As there are increasing interests in the timing for bottoming out of the US housing market, there is a risk that the existence of such a lag in cognition causes the increase of uncertainty and the delay in economic recovery. Secondly, when looking at the relationship between the demand for houses and house prices based on the time-series data, there is a positive correlation between the two elements. However, upon conducting an analysis using the panel data, which is based on data in units of prefectures or states, there is no significant relationship between the demand for houses and house prices in both Japan and the United States. In this sense, it is hard to explain whether there is a bubble and the size of the bubble according to prefecture (state) using demand elements. This suggests that it is possible that the concept of demographics having an impact on the demand for houses, which thus caused the house prices to increase, is not effective in explaining the price fluctuations in neither Japan nor the United States. Thirdly, when looking at the co-movement between the house prices and rent, a phenomenon which the rent almost does not fluctuate at all even when the significant change of house prices change in the process of the formation and collapse of a bubble was confirmed for both Japan and the United States. Its background is that landlords and tenants have formed long-term contractual relationships so that both parties can save on various transactional costs. In addition, the imputed rent of one's home is not assessed using market prices in Japan, which is an aspect to weaken the co-movement. A lack of co-movement causes a phenomenon in Japan and the United States where consumer prices that include this rent as an important element do not increase since rent does not increase even if housing prices increase during a bubble period. Thus, it results in a delay towards a shift to tighten credits. Since rent prices do not move together with the house prices even after house prices decrease after the collapse of the bubble, a phenomenon which consumer prices do not decrease was observed. This served as a factor for the delay in a shift towards monetary relaxation. Rent prices are an important variable that serves as a node between asset prices and prices of goods and services. It is necessary to increase the accuracy with which it is measured.

25. January.2010

Research Center for Price Dynamics, Institute of Economic Research, Hitotsubashi University

2010-01-29

English(eng)

technical report

VoR

46

日本学術振興会・科学研究費補助金（学術創成研究費） = JSPS Grant-in-Aid for Creative Scientific Research

Working Paper Series / Understanding Inflation Dynamics of the Japanese Economy

https://hdl.handle.net/10086/18243