Elsevier

Land Use Policy

Volume 99, December 2020, 104828
Land Use Policy

Impact of transit stations on house prices across entire price spectrum: A quantile regression approach

https://doi.org/10.1016/j.landusepol.2020.104828Get rights and content

Highlights

  • This study provides evidence on the impact of a heavy-rail-based transit station on house prices.

  • Dataset includes characteristics of single-family houses sold during January 2000–April 2018 period.

  • Houses are located within eight kilometers of Warm Springs station of the San Francisco Bay Area Rapid Transit system.

  • The station increased house prices across the entire price spectrum.

  • The price increase began more than ten years before the rail service commenced.

Abstract

This study provides evidence on the impact of a heavy-rail-based transit station on house prices by employing quantile regression method on a dataset that inventories sales transactions of single-family houses during the January 2000–April 2018 period, and within eight kilometers (five miles) of the Warm Springs station of the San Francisco Bay Area Rapid Transit system. The station is located in Fremont, CA. The results show that a) the station increased house prices across the entire price spectrum, and b) the price increase began more than ten years before the rail service commenced. On a larger note, this study should strengthen efforts to a) provide or extend heavy-rail-based rapid transit systems in areas of high need and b) explore the use of value capture tools to fund transit.

Introduction

The last two to three decades have witnessed an increasing interest in public transit across various levels in the United States (US), from regional to national. Furthermore, the federal transportation acts—starting from the Intermodal Surface Transportation Efficiency Act of 1991 to the latest, Fixing America’s Surface Transportation Act of 2015—have called for transportation-land use coordination. Moreover, federal funding programs such as the Livable and Sustainable Communities Program and the New Starts Program, have provided the much needed financing for public transit projects. However, due to significant funds needed to construct and operate public transit systems, the federal funding is inadequate and the fiscally constrained local and state governments are often unable to fill the funding gap. In such a scenario, any new revenue source that helps fund public transit projects is welcome. Value capture (VC) is one such source.

A set of VC tools, such as the special assessment districts (SADs), capture the public-infrastructure-led increases in land value. Normatively, such tools are based upon the “benefits received” principle, which posits that the beneficiaries of a particular infrastructure/service should also pay for it. If the provision of or enhancements to public transit systems generate accessibility-related benefits to neighboring properties, these benefits get positively capitalized into higher land values—a windfall gain for private property owners. Hence, it is argued that these property owners should help fund public transit systems because they benefit from them (Smith and Gihring, 2009). However, the first step is to empirically demonstrate that the public infrastructure has indeed increased the value of neighboring properties, and to ascertain whether the impacts are consistent across the entire property price spectrum. For example, to the extent higher income households who might live in higher priced houses are unlikely to use public transit, proximity of such houses to transit stations might not significantly increase their prices. On the other hand, to the extent lower income households who might live in lower priced houses often use transit, proximity of such houses to transit stations might significantly increase their prices.

While the US-focused extant literature has demonstrated the property value impact of rail transit investments, most recent studies focus on the light rail systems (see, Cervero and Duncan, 2002a; Chatman et al., 2012; Golub et al., 2012; Hess and Almeida, 2007; Ko and Cao, 2010; Ryan, 2005; Smith and Gihring, 2009). Apart from one recent study that estimates property value impacts of heavy-rail-based rapid transit (see, Mathur, 2019a), most studies that focus on heavy-rail-based rapid transit are older (see, Benjamin and Sirmans, 1996; Cervero and Duncan, 2002b; Cervero and Landis, 1997; Gatzlaff and Smith, 1993; Lewis-Workman and Brod, 1997; Nelson, 1992). Moreover, two research gaps remain in existing literature.

First, the studies often lack robust research design. They are often cross-sectional studies that lack control group, and do not check and correct for spatial dependence. Second, most existing studies estimate the property value impact on an average-priced house, not on the entire house price spectrum—from low to high priced houses. Such fine-grained estimation is important because, as mentioned above, a transit system could impact a low priced house differently than a high priced house. The Warm Springs BART Extension (WSX) Project, which includes a 8.7- kilometer (5.4-mile) extension of the San Francisco Bay Area Rapid Transit (BART), a heavy-rail-based rapid transit system, and the opening of the Warm Springs (WS) station in March 2017 (San Francisco Bay Area Rapid Transit District (BART, 2017a) provides an opportunity to fill these two research gaps by studying the impact of WS station on the values of neighboring properties.

Section snippets

Methodological limitations

Existing studies that estimates the property value impacts of heavy-rail-based transit systems often employ methodological approaches that are not very robust. For example, Baum-Snow and Kahn (2000) and Kahn (2007) use highly aggregated census-tract-level data to estimate the impact of rail transit on a census tract’s median house value, thereby assuming that the transit station’s impacts are homogeneous across various owned property types, such as single-family houses, condominiums, and

Research question

This study seeks to fill various research gaps identified in the above section, specifically, a) check for spatial dependence, b) study house price impacts by analyzing data for both pre- and post-transit service commencement periods, and c) estimate the impact on house prices across the entire price spectrum by employing the QRM to estimate the impact of the WS BART station on the surrounding single-family house prices. Specifically, this study seeks to answer the following research question:

Methods

This study estimates owner households’ marginal willingness to pay for houses in close proximity (0–3.2 kilometers, or 0–2 miles) to the WS BART station compared to those further away (3.2–8 kilometers, or 2–5 miles). Under the multiple linear regression approach, the main estimation equation regresses the sale price of a house i (si) on its structural and locational attributes (lci), including the proximity of the house to the WS BART station (see Eq. 1). α0 is the constant, αi is the

Model structure

The QRMs are used to run nine regression models, one for each decile of house prices—from the 1st decile to the 9th decile. Next, a set of OLS fixed effect models are run to compare their results with the QRMs’ results. One OLS model uses White’s Heteroscedasticity Consistent estimator and the second OLS model uses robust standard errors clustered at the elementary school attendance zone level. Both sets of regression models—QRMs and OLS fixed effects models—include a dummy variable to measure

Conclusions and policy implications

Only a few studies use fine-grained data and sophisticated research design and statistical analysis techniques to estimate the impact of transit stations on house prices. Moreover, these studies only examine a transit station’s impact on an average-priced house. This study fills these research gaps by using parcel-level data to run QRMs to estimate the impact of a transit station across the entire house price spectrum.

The findings of the study have the following policy implications:

  • a)

    The largely

CRediT authorship contribution statement

Shishir Mathur: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing - original draft, Writing - review & editing.

Acknowledgements

The research for this paper was partially funded by the US Department of Transportation Office of the Assistant Secretary for Research and Technology University Transportation Centers Program (grant 69A3551747127) through the Mineta Transportation Institute at San Jose State University.

References (36)

  • John Benjamin et al.

    Mass transportation, apartment rent and property values

    J. Real Estate Res.

    (1996)
  • Roger Bivand et al.

    lm.LMtests: Lagrange Multiplier Diagnostics for Spatial Dependence in Linear Models

    (2018)
  • Robert Cervero et al.

    Transit’s value-added effects: light and commuter rail services and commercial land values transportation research record

    J. Transp. Res. Board

    (2002)
  • Daniel Chatman et al.

    Evaluating the economic impacts of light rail by measuring home appreciation: a first look at New Jersey’s river line

    Urban Stud.

    (2012)
  • Wenhua Di et al.

    An analysis of the neighborhood impacts of a mortgage assistance program: a spatial hedonic model

    J. Policy Anal. Manag.

    (2010)
  • David Fin et al.

    Quantile regression: its application in investment analysis

    Finsia J. Appl. Finance

    (2009)
  • Dean Gatzlaff et al.

    The impact of the Miami Metrorail on the value of residences near station locations

    Land Econ.

    (1993)
  • Aaron Golub et al.

    Spatial and temporal capitalization effects of light rail in Phoenix: from conception, planning, and construction to operation

    J. Plan. Educ. Res.

    (2012)
  • Cited by (19)

    • Externality impacts of “Not in My Backyard” facilities on property values: Evidence from the Hangzhou waste sorting and reduction complex projects

      2022, Habitat International
      Citation Excerpt :

      WSRCs significantly impact property values during the planning phase. Existing evidence indicates that residents who purchase property of different values evaluate urban facilities in different ways, as reflected in the heterogeneous sensitivity of high-, medium-, and low-value properties (Kim et al., 2015; Mathur, 2020; Mueller & Loomis, 2014; Wang et al., 2016; Wen et al., 2019, 2021; Yang et al., 2020; Zhang, 2016). In addition to assessing the average externality impact of WSRCs, we are interested in how externality impacts manifest differently across property values.

    • Bringing the vertical dimension into a planar multilevel autoregressive model: A city-level hedonic analysis of homebuyers' utilities and urban river attributes

      2021, Science of the Total Environment
      Citation Excerpt :

      For instance, environmental amenities (e.g., scenic vistas) or disamenities (e.g., air pollution) which manifest themselves and spread at different intensities over 3-D space (Liu and Jakus, 2015; Chen, 2017; Xiao et al., 2019), will exert varying external impacts on apartment prices along the horizontal and vertical dimensions. Hedonic price models (Rosen, 1974) rely on market transactions of heterogeneous housing units to determine the implicit prices of, but also generate essential information for, urban development planning (Nicholls, 2019; Mathur, 2020; Wu et al., 2004) and designing nature interventions (Kim et al., 2019; Pilla et al., 2019; Bockarjova et al., 2020) so as to contribute to sustainable urbanization. A key methodological challenge for 3-D hedonic price models is to appropriately account for all potential spatial effects (including spatial heterogeneity and spatial dependence) in the 3-D spatial context of modern cities (Anselin, 2003; Anselin and Lozano-Gracia, 2009; LeSage and Pace, 2009; Osland, 2010; Higgins, 2019; Li et al., 2020).

    • Transit-induced gentrification and displacement: The state of the debate

      2021, Advances in Transport Policy and Planning
      Citation Excerpt :

      At least two studies suggested that larger capitalization effects were experienced in higher-income neighborhoods (Bowes and Ihlanfeldt, 2001; Hess and Almeida, 2007). However, Mathur (2020) found positive price impacts across the entire spectrum of properties in the case of the San Francisco BART heavy rail system. Redfearn (2009) makes an important point with respect to modeling efforts that attempt to capture price premiums and fail to recognize the significant variations across time and neighborhood types.

    • Accessibility and proximity effects of bus rapid transit on housing prices: Heterogeneity across price quantiles and space

      2020, Journal of Transport Geography
      Citation Excerpt :

      For instance, it helps policymakers coordinate the interests of various stakeholders: With proper space zoning, commercial properties rather than residential ones can be set close to the BRT corridor to reduce the nuisances and enjoy the convenience induced by BRT systems (Yang et al., 2019a). Moreover, it can be reasonably expected that the effects of BRT on housing prices vary according to housing price levels (Mathur, 2020) and spatial location (Mulley, 2014; Mulley et al., 2016). Put it in another way, the capitalization effects of BRT on housing prices may be heterogeneous.

    View all citing articles on Scopus
    View full text