Three Essays on the US Homeowner's Response to a Changing Economy

We examine the response of homeowners in the United States to house price fluctuations and publicly-provided amenities. The US housing market recently experienced a great upheaval, and its effects have been enormous and far-reaching. Understanding the sometimes quirky behavior of homeowners is crucial for establishing effective public policy to ensure that housing markets remain efficient and stable. Specifically, we investigate the homeowner decisions of where to locate and how long to stay as well as the implications of those decisions for housing values.;In the first essay, we explore the homeowner tenure decision and economic factors that induce or prevent the often burdensome and expensive process of relocating to a new residence. Municipalities use scarce resources to compete for new residents and to retain existing ones, seeking not only their tax dollars but also the benefits of stable neighborhoods. Though it is well established that public goods such as school quality, low crime, and clean air attract residents and affect home prices, it is less clear how the provision of amenities affects the tenure for residents that have already moved to the area. Further, the interplay of this effect with those of equity constraints and homeowner loss-aversion has not been investigated. Not only is the desire to move an important factor but the ability to move is as well.;We apply a Cox proportional hazards of a homeowner's annual decision to move or stay to a dataset of home mortgage transactions and local amenities in the Baltimore/Washington D.C. metro area in the years surrounding the US housing boom and bust cycle of the late 2000s. This allows us to disentangle the effects of homeowner credit constraints from fluctuations in both property values and local crime on the homeowner tenure decision. Because this was a time of extreme volatility, we can test for fundamental changes in homeowner behavior that may have been caused by this sudden, extreme change in the housing market.;In the second essay, we analyze the relationship between homeowners' choice of housing, the user cost of homeownership, and US housing policy. Homeowners face the decision of maximizing utility by choosing the optimal level of housing services to purchase, given income and credit constraints. By modeling this decision using micro-level data of home mortgages, we can predict the effects of proposed US federal housing policy that would increase the cost of owning a home. Further, we can use this framework to examine the extent to which homeowners during the US boom period were affected by credit constraints that were implemented to minimize the extent of overextended homeowners and stem foreclosures that sparked the US housing bust.;The US federal government heavily subsidizes homeownership, in part via two policies that reduce the tax burden of homeowners: the mortgage interest and property tax deduction and the absence of a tax on imputed rent for owner-occupiers. In the United States, homeowners who lease a property must report for tax purposes all rent payments they receive as income. Owner-occupiers essentially pay this rent to themselves, but this "imputed rent" is not reported and so is not taxed in the US. In this way, the US federal government not only subsidizes homeownership through explicit tax deductibility of house payments but also owner-occupiership by leaving imputed rent untaxed. We predict that while both policies would decrease demand for housing, taxing imputed rent has a stronger effect. Also, both policies would be progressive; the decrease in house price is larger for higher-income homeowners both as a percentage change and as a portion of income. Finally, we find that mortgagees in the sample not only very frequently surpass typical credit constraints, but that homeowners---particularly those whose debt-to-income ratios surpass 28%---frequently purchase a level of housing services above predicted optimal levels. This overconsumption of housing became more prevalent during the housing boom.;In the third essay, we analyze how a homeowner's evaluation of crime is affected by the type of crime committed and the spatial distribution of crime around a home. Nearby crime is capitalized into house prices, but it is not well understood how the degree of that capitalization varies with the crime incident's distance from the house. We employ a simple hedonic model of house prices in Baltimore City using a micro-level dataset of location-specific crimes and houses to shed light on this relationship.;Because crime is potentially correlated with a myriad of spatial unobservables, as crime could be attracted to areas of greater property value, we additionally estimate a Hausman-Taylor model with nearby tree-canopy density and distance to the nearest park as instruments for neighborhood crime. (Abstract shortened by UMI.).