In this article the authors examine the nature of housing bubbles and how housing prices are affected by the elasticity of housing supply. They begin by citing from literature to show that price movements of assets, including markets, include an irrational exuberance element as well as their fundamental valuations; this leads to the difficulty in explaining the significant variations in housing prices between 1999 and 2005. As a result, instead of developing a test to detect the presence of a housing bubble, the authors propose to analyze the nature of this bubble; this is especially in light of the fact that housing supply and pricing influence construction activities as well as population immigration, thus impacting general welfare. They develop an endogenous model for asset bubbles by making it dependent on supply inelasticity – the model predicts that when a bubble bursts, houses that are constructed during the bubble witness their prices falling to lower than their pre-bubble values. Further, they also perform an empirical analysis of data on prices, construction activities and supply elasticity of houses during two boom-and-bust cycles (during the 1980s and post-1996); they also study the differences in pricing and supply levels during these cycles. It is observed that the premise of the paper is well established, and the proposed models clearly defined; the predictions from the models are identified properly in terms of stated elastic markets (Orlando and Phoenix), and a previous housing bust (1989-1996) is identified.
Methodology
The methodology adopted is building an analytical model of housing stock supply which considers irrational exuberance to be an exogenous factor, and which analyzes the effect of supply on suppressing or encouraging housing bubbles as well as their effects on welfare. A continuous time model is adopted, with demand being generated by new homebuyers based on a utility function, and supply coming from both new and old homes. An open city model is assumed, implying a uniform flow rate of new buyers into a city, with a purchase occurring if the buyer’s perceived utility in staying within a city combined with his expected resale value exceeds the price of the house. Based on these and some other assumptions, five propositions are outlines and proved. The propositions test various aspects of the model, such as: (i) for elastic supply and a finite number of buyers, equilibrium in the market does not exist – this analyzes irrational housing bubbles; (ii) change in housing price becomes slower as a bubble progresses, but stock supply increases – this analyzes the impact of a bubble on housing process, investment and supply; (iii) the effect of the bubble on housing stock supply – this analyzes price fall to levels lower than the pre-bubble values; (iv) the role of inelastic supply in sustaining the bubble; and (iv) the relationship between inelasticity and housing produced by a bubble. All the propositions are proved and examined logically, and the consequences are clearly defined. While the methodology seems adequate, the authors also point out that the model makes two empirical predictions, but is vague on a few other issues.
Empirical analysis and data presentation In addition to the theoretical model, the authors also carry out an empirical survey and analysis of housing market data – once again, they examine the implications of the bubble, rather than testing its existence, in accordance with their stated objectives. Data for this study is obtained from the Office of Federal Housing Enterprise Oversight (OFHEO), and their indices for metropolitan area-level indices from 2000 onwards are used. A detailed analysis is performed to obtain the ratio of housing prices to their Minimum Profitable Production Cost (MPCC) for 79 metropolitan cities from 1982 to 2007; these mean ratios are then plotted graphically to illustrate the fact that prices increased dramatically above their MPCC levels during the housing boom post 2000. Similarly, another analysis is performed for the 1982-1996 cycle, and an extended plot is drawn graphically. The authors carry out a validation of the presented data against the predictions of their model. For example, one of the main contentions of the model is that highly elastic housing stock supply leads to a lesser probability of having a bubble, and in order to test this, the authors compare the effect on prices of the 1982-1989 cycle in both elastic and inelastic cities. They perform an analytically correct exercise to show that there is, indeed, a significantly different (by a factor of 5) rate of growth in real prices for less versus more elastic areas. The authors support these as well as other conclusions from the data with the help of appropriate tables and graphs, which appear to reinforce the strong correlations between the predictions of the model and the actual empirical analysis.
Errata
The errors and ambiguities in predictions of the models are clearly pointed out, and possible reasons are given. For example, while analyzing the housing bust from 1989 to 1996, the authors remark that their regression results are not as realistic as those for a boom period. In order to highlight this, they contrast between two tabular forms, adding a baseline bivariate regression column in the latter; however, refining this by adding demand controls eliminates the statistical significance of the data. The authors discuss this, as well as other ambiguities in the model, such as the impact of a bubble on new construction activities, within the body of the paper. They point out that adjusting for influences of area-level characteristics during the period from 1982 to 1990 leads to the conclusion that more houses were added by more elastic areas. The graphs and tables appear to support the conclusions drawn by the authors in all the cases. In conclusion, it can be observed that this is a well-presented paper with a sound methodology. The predictions from the model are tested through rigorous statistical tests, and ambiguities are pointed out. The central proposition of the paper – that inelastic housing supply contributes to a housing bubble – appears to stand.