Crime and the Local Housing Market: A Longitudinal Analysis of Home Sales in Los Angeles, 1993-2004

Author: Lyndsay N. Boggess

Dissertation School: University of California, Irvine

Pages: 31

Publication Date: 06/2007

Availability:
Available from the HUD USER Helpdesk P.O. Box 23268 Washington, DC 20026-3268 Toll Free: 1-800-245-2691 Fax: 1-202-708-9981 Email: oup@oup.org

Access Number: 5022

Abstract:

The relationship between crime and communities has long been a focal point of criminological and sociological investigation. Most neighborhood level research, however, treats crime as an outcome of socioeconomic characteristics of places, but crime is also an important catalyst for community change. Increases in crime have negative consequences for neighborhoods, including increased residential mobility (Morenoff and Sampson, 1997) and decreased neighborhood satisfaction (Skogan, 1992). As homeowners relocate to safer communities, crime also initiates changes in the local housing market, such as reduced house values and increased vacant properties (Taylor, 1995). Additionally, changes in the market have consequences beyond the homeowners. Real estate agents' negative perceptions of high crime areas can steer potential buyers away and the out-migration of financially stable households reduces the local tax base. It is reasonable to assume then, that the level of crime in a neighborhood would lead to lower demand for that community, i.e., crime is a "dis-amenity" just as poorly performing schools or a lack of basic municipal services are dis-amenities.

Indeed, numerous studies have measured the impact of crime on housing. Most studies have shown that crime has a small but negative effect on property value (Hellman and Naroff, 1979; Lynch and Rasmussen, 2001), though more so for violent crimes (Tita, et. al., 2006), or in high-crime neighborhoods (Lynch and Rasmussen, 2001). Indeed, there is also evidence that crime can impact housing value in adjoining communities (Burnell, 1988). Sale price, however, is only one component of the housing market. The number of residential property transactions is also indicative of housing demand and neighborhood desirability. Studies that rely solely on measures of property value, therefore, are likely to underestimate the impact of crime on the housing market.

Unlike previous studies, I utilize the rate of residential property transactions to measure the impact of crime on the local housing market. I argue that analyzing the volume of homes sold in a neighborhood provides additional information on the impact of crime that hedonic models of property value likely miss. More specifically, hedonic models of property value rely on data from completed housing transactions, which ignores those situations when crime motivates residents to sell but there is no willing buyer and thus few houses are transacted. In neighborhoods where crime causes few homes to sell or the supply of houses is relatively elastic, changes in demand caused by crime are likely to have a more perceptible impact on quantity sold rather than price. Thus, the true impact of crime on housing may be understated.

In addition to introducing transactions as a complementary measure to price, this study improves our knowledge of the relationship between crime and the housing market in several important ways. By examining the impact of crime on housing from a disaggregated local perspective (census tracts), I can analyze location-sensitive factors such as crime, demographics, and economics that are related to local housing markets that citywide analysis masks. In addition, another advantage of smaller geographic units is that the homogeneity is increased within the unit while simultaneously the variation across units is increased. Localized analysis is important for diverse cities like Los Angeles, where housing and crime vary widely depending on location within the city. For example, the crime rate in southeast Los Angeles is generally higher than most other areas of the city, and houses near the coast are usually worth more than those by the airport. Such important variation is lost in prior empirical literature because of the high level of aggregation.

Further, because different crimes impose different costs to society (e.g., Cohen, 1990, 2005), and property crime impacts a different aspect of the housing market than violent crimes (Taylor, 1995), I disaggregate the analysis by broad crime categories to examine the differential impact of violent and property crimes n the neighborhood demand for housing. My use of longitudinal data allows me to differentiate between the impacts of changes in crime from historic levels of crime. Previous research has shown that potential homebuyers are aware of neighborhoods' reputations for safety, and incorporate this information when deciding where to buy (Taylor, 1995). Other research has suggested that changes in crime are more likely to induce changes in behavior (i.e., relocation) (Tita, et. al., 2006). That is, large or noticeable changes in crime will draw more (negative) attention to particular neighborhoods thereby discouraging homeowners from investing in a community whose crime rate has been increasing over time.

In an additional extension of prior research, I will analyze the relationship between crime and the housing market by neighborhood type based on "high," "medium," or "low" levels of income. Previous crime and housing research rest upon the notion of a singular housing market in which all home sellers compete for the population of homebuyers. Many argue this is an unreasonable assumption, noting that housing markets are segmented by price such that comparable homes within a market (city) vary in price according to the desirability of the particular location of the house (Goodman and Thibodeau, 1998; Gallet, 2004). By separating my analysis into different submarkets based on socioeconomic class, I can determine whether the sensitivity of housing transactions to crime varies by location. In more affluent communities it is reasonable to expect that large increases in crime will increase residential transactions as homeowners move out. In these higher income neighborhoods, residents have the resources to relocate, and changes in crime may be a sufficient motivator to move despite the possibility of increases in property taxes.

On the other hand, homeowners in more affluent locations also have the resources to insulate themselves from crime (with private security, alarms, gates, greater access to public officials, etc.), so they may have a greater menu of options other than simply fleeing the neighborhood. That is, they may be able to stay in their home while investing in services and technology offering personal and property protection. It is possible, too, that higher income homeowners will react differently to violent crime, which is less common in their communities, than property crime. In lower income neighborhoods, however, high levels of crime or large changes in the crime rate are likely to decrease the number of home sales despite residents desire to move out: high crime reduces the desirability of the area and potential homeowners are reluctant to move into such communities. At the same time, lower income homeowners are trapped into their current property. Though housing is the largest source of wealth accumulation for most Americans, though it represents a larger portion of personal assets for lower income households and minorities (Flippen, 2004). Therefore, while more affluent homeowners can rely on additional sources of wealth to use to purchase a new property, low-income homeowners must rely on the sale of their current home in order to facilitate the purchase of a new one.

The remainder of this paper is organized as follows. The next section will present a discussion of the previous research on the impact of crime on the housing market and the influence of residential mobility on neighborhood stability. Because this study takes a novel approach on the topic and no prior published research specifically addresses the impact of crime on the number of housing transactions, the literature reviewed addresses the related issue of the desirability and the impact of crime on property values rather than transactions. I then discuss methodology and data used in this analysis. I conclude with the regression results and a discussion of my finding along with suggestions fur future studies. The empirical results indicate that the level of crime is significantly related to the rate of home sales, though the relationship is specified by neighborhood type. My results further show that the changes in crime rate over time do no significantly impact the rate of housing transactions.

Back to Search Result of EDSRG Dissertations