This dissertation consists of three essays in housing and real estate economics which explore housing dynamics, extend housing filtering theory, housing maintenance, construction, and investigate issues in the economic depreciation of housing structures. The first chapter gives an introduction, with special emphasis on the motivations behind this research.
The second chapter develops a spatial, general competitive equilibrium model for metropolitan housing markets. Then the model is parameterized to quantitatively investigate their properties. Using a variety of data sets, a benchmark model is calibrated to the housing market of Riverside, CA, and the equilibrium is then numerically solved using a self-developed algorithm. The chapter investigates the model's equilibrium and compares the housing features in the benchmark model with two alternative steady states. The simulation results replicate the empirical fact that the ratio of maintenance expenditure to the property value is lower for newly constructed housing than for an aged house, and that it is consistent with the one percent maintenance expenditure rule in real estate economics. Moreover, the numerical results from alternative steady states show that a positive demand shock increases the structural density and decreases the housing space, but it does not increase the rent and property value of a housing unit (although each housing unit’s floor area becomes smaller). A positive income shock will increase the rent and the property value of a single housing unit at all levels of housing quality, and also increase the housing space and structural density of new residential construction. This chapter extends and refines housing filtering theory in several aspects. It distinguishes non-depreciable housing space with depreciable housing interior quality, emphasizing the indivisibility of housing services by incorporating a two-dimensional and nonlinear bid-rent function. It provides a quantifiable measure of housing interior quality and demonstrates its scale-independent nature. To the best of my knowledge, it is the first structural model that incorporates two heterogeneous dimensions of housing characteristics in both the demand side and the supply side of the housing market.
The third chapter, motivated by the observation that the land of a property does not depreciate, along with accounting for the capital gains from housing boom and bust cycles, provides a method to separate housing structure depreciation and capital gains on land into two distinct components that make up overall changes in property value. This chapter provides an approach to estimate the housing maintenance technology, and to directly measure the housing structure depreciation while taking account of the maintenance effect and the capital gains on land. The unobservable housing structure value is derived from the difference between the observable property value and the land value associated with the property. The land values are imputed by taking into account the plattage effect (larger parcels have a lower value per unit area of land) and the lot size information of each property. By noticing that general urban accessibility should be controlled when modeling parcel values with this effect, I choose to focus on the New York metropolitan area, where I can separate the urban center from its peripheral regions by utilizing the fact that there are almost no single-family detached houses in Manhattan used for residential occupancy. Using American Housing Survey data, chapter 3 demonstrates that in both the baseline model and the extended model (by incorporating pre- and post-financial recession data), the estimation results are similar and robust.
The fourth chapter explores economic reasons why the average construction cost curve is U-shaped with respect to building height, as noted in previous housing construction literature. A construction cost function is developed to explain such a relationship, particularly with regards to the U-shaped curve. Fixed construction cost is recognized and established as an element in the housing construction function.
The last chapter concludes and summarizes the contributions of this research.