Measur ing Highway Impacts on House Pr ices Us ing Spat ia l Regress ion

Real estate value theorists and practitioners widely acknowledge the importance of considering external influences when analyzing real estate values. Nearby transportation systems are a classic case of offsite factors that may affect real estate values. While highways, rail, and airports generally provide positive benefits to real estate users at the macro level by increasing accessibility, micro-level effects of planes, trains, and automobiles may be negative due to potential nuisances generated by transportation systems. For residential real estate in particular, a critical issue at the micro-level is the impact of noise from nearby transportation infrastructure on house prices. Numerous studies, some of which are discussed below, use a variety of analysis methods to detect and measure the price effects of transportation-related noise pollution generated by airports, railways, and highways in various housing markets around the world. These studies document consistently significant price discounts for noise pollution related to nearby transportation systems but, not surprisingly, the magnitudes of the reported discounts vary across markets. Previous studies typically use one of two alternative approaches to consider the impacts of highways on house prices, including noise pollution. The first of these approaches uses noise levels measured in decibels and the second approach uses transportation infrastructure adjacency and proximity measures. As noted by Seo, 8 4 u A l l e n , A u s t i n , a n d S w a l e h e e n Golub, and Kuby (2014, p. 54), noise levels ‘‘are quite expensive to measure for individual parcels.’’ This study relies on the second approach and uses locationbased measures to assess highway impacts on house prices. The measures include a highway adjacency indicator, nearby highway traffic volume, straight-line distances from houses to highways, and the presence of sound barrier walls to examine highway impacts on house prices. Adjacency to highways, especially more heavily traveled highways, is expected to reduce house prices, but the negative impact of highway noise on house prices may be offset by sound barrier walls and increased distance from house to highway. This study also controls for potential impacts on house prices related to accessibility to the highway system. Reduced driving distance to nearby highway on-ramps may have a positive impact on house prices. The impetus for considering highway impacts on house prices in this study is the ongoing construction of a new multi-lane, median divided, limited access, toll road near Orlando, Florida. Portions of the new highway will have sound barrier walls along the right-of-way. Plans for the Wekiva Parkway were approved in 2004 and construction began in 2013, with a projected completion date in 2021 at a cost of $1.5 billion. This 25-mile highway is the final section of a 112-mile beltway highway around the Orlando metropolitan area. The Wekiva Parkway is intended to relieve congestion on U.S. Highway 441 and State Road (SR) 46, the primary surface routes through this portion of Florida. The massive scale of this highway project will undoubtedly improve mobility through the area, but the project may also negatively impact the prices of individual houses in the project’s vicinity. The purpose of this study is to document micro-level house price impacts of other existing highways in this market in an effort to provide a basis for anticipating the impact of the new highway on nearby houses, both planned and existing. Using transaction data from this market, hedonic price analysis with spatial regression modeling suggests that the average price discount for houses adjacent to highways is approximately 4.0%, holding other value influencing factors constant. The results of this analysis also document significant price discounts related to traffic volume on nearby highways. The results do not indicate that sound barrier walls or increased straight-line distances from houses to the highway impact house prices, but do indicate that houses with shorter driving distances to highway on-ramps sell at price premiums. The following section provides a brief review of the studies that, like the present study, examine highway impacts on the value of residential real estate using the revealed preferences of market participants with either measured noise levels or location-based measures. In the remaining sections, we discuss the data, the spatial regression models used to analyze the data, and the results of the analysis, respectively. The paper closes with concluding remarks. u S t u d i e s A d d r e s s i n g H i g h w a y I m p a c t s o n u R e s i d e n t i a l P r o p e r t y V a l u e s Bateman, Day, Lake, and Lovett (2001) provide in-depth discussions of some of the key issues, methods, and results from an assortment of studies published prior M e a s u r i n g H i g h w a y I m p a c t s o n H o u s e P r i c e s u 8 5 J O S R E u V o l . 7 u N o . 1 – 2 0 1 5 to 2001 that analyze highway noise impacts on residential property values using measured noise levels. The authors review 17 studies that consider the effect of highway noise using hedonic price analysis and report an average price discount of 0.4% per decibel in those studies, with a standard deviation of 0.23%. These authors also conduct their own hedonic price study of the impact of highway noise using data from the Glasgow, Scotland area and report a house price discount of 0.2% per decibel for their sample. Wilhelmsson (2000) provides a notable study that is not included in the literature review provided by Bateman, Day, Lake, and Lovett (2001). Using data from Stockholm, Sweden, his hedonic price estimates indicate an average highway noise discount in house prices of 0.6% per decibel. To put this finding into perspective, Wilhelmsson demonstrates that the difference in value for a house in a noisy and a quiet location in his study sample is approximately 30%. Numerous other researchers have conducted highway noise pollution impact studies on various property types of residential real estate in markets around the world since the publication of the Bateman, Day, Lake, and Lovett (2001) literature review. Becker and Lavee (2003) report a price discount of 1.2% per decibel for apartments in urban areas in Israel. Rich and Nielson (2004) report price discounts of 0.47% per decibel for apartments and 0.54% per decibel for houses in Copenhagen, Denmark. Theebe (2004) reports an average discount of 5% from traffic noise from planes, trains, and automobiles on houses in the Netherlands. Baranzini and Ramirez (2005) report a discount in apartment rents of 0.63% per decibel in Geneva, Switzerland. Day, Bateman, and Lake (2007) report a price discount of 0.55% per decibel for residential real estate in Birmingham, England. Kim, Park, and Kweon (2007) report a price discount of 1.3% for a 1% increase in traffic noise level in decibels for single-family and row houses in Seoul, Korea. Nelson (2008) discusses prices discounts on property values associated with both aircraft (0.8% per decibel) and road traffic (0.54% per decibel). Andersson, Jonsson, and Ögren (2010) report a discount of approximately 0.7% per decibel for single-family houses in Lerum, Sweden. Blanco and Flindell (2011) report a price discount of 0.45% per decibel for apartments and flats in London, England. Brandt and Maennig (2011) report a discount of 0.23% per decibel level for condominiums in Hamburg, Germany. Li and Saphores (2012) report a negligible discount for general highway traffic (0.006% per decibel), but a more substantial discount for highway truck traffic of 0.65% per decibel for houses in Los Angeles, California. Several additional studies examine the impact of highways on property values using location-based measures rather than noise level measurements. Hughes and Sirmans (1992, 1993) use average daily traffic counts and an indicator variable for high-traffic streets to identify significant price discounts in the Baton Rouge, Louisiana housing market. They report price discounts of 9.2% for city neighborhoods and 4.6% for suburban neighborhoods for high-traffic streets. Kawamura and Mahajan (2005) discuss the use of hedonic price models with traffic count data. Larsen (2012) reports that houses adjacent to high-traffic streets sell for discounts of 8.1%. Larsen and Blair (2014) report a price discount of 7.8% for single-family houses on high-traffic streets and a price premium of 13.8% for 8 6 u A l l e n , A u s t i n , a n d S w a l e h e e n multi-unit rental residential properties in Kettering, Ohio. Kilpatrick, Throupe, Carruthers, and Krause (1997) show that proximity to transit corridors, not just adjacency to a corridor, is negatively related to property values and that access to the transportation infrastructure is positively related to property values in a sample of houses from Seattle, Washington. Although the intention of sound barrier walls along highways is to reduce the aural and visual nuisances of highways on nearby properties, Julien and Lanoie (2008) show that noise barrier walls between houses and highways result in price discounts of 6% in the short run and 11% in the long run in a sample from Montreal, Canada. Our analysis follows the lead of Hughes and Sirmans (1992, 1993), Kawamura and Mahajan (2005), Larsen (2012), and Larsen and Blair (2014) in the use of location-based measures to evaluate the impact of highways on house prices. In particular, an indicator variable is used to identify houses adjacent to highways and traffic count data to identify high-volume highways. In addition, we consider the issues of straight-line distance from houses to highways and the driving distance between houses and highway on-ramps following Kilpatrick, Throupe, Carruthers, and Krause (1997) and the issue of sound barrier walls following Julien and Lanoie (2008). u D a t a D e s c r i p t i o n The primary sour