HOW WILL THE CENTERLINE AFFECT PROPERTY VALUES IN ORANGE COUNTY? A REVIEW OF THE LITERATURE AND METHODOLOGICAL APPROACHES FOR FUTURE CONSIDERATION Prepared by Lee Cockerill, M.A. and Denise Stanley, Ph.D. Institute of Economic and Environmental Studies California State University-Fullerton Fullerton, CA 92834 October 28, 2002
Table of Contents 1) Executive Summary 2) Understanding the links between public transit projects and property values: preparing for the CenterLine in Orange County, CA Pg. 1 3) The CenterLine Project Pg. 2 4) Introduction: Theory and Methods. Pg. 3 a) Property Value and Location Choice pg 3 5) Methods to Estimate the Impact of Rail Transit on Property Values. Pg. 5 a) Method 1 Case Comparison Pg. 5 b) Method 2 Matched Pairs Method.. Pg. 6 c) Method 3 Multiple Regression Pg. 7 6) Previous Residential Property Studies. Pg. 8 7) Commercial Property Studies.. Pg. 9 8) Overall Findings in the Literature Pg. 9 9) Extending the Analysis to the Orange County CenterLine. Pg. 12 10) Appendix A. Hedonic Price Regression Profile. Pg. 15 11) Appendix B. Table. Summary Effects of Rail Transit on Property Values.. Pg. 16 12) Appendix C. Details of City Property Value Impacts. Pg. 17 13) Bibliography Pg.
HOW WILL THE CENTERLINE AFFECT PROPERTY VALUES IN ORANGE COUNTY? EXECUTIVE SUMMARY The economic impact of public transit is an important concern for policymakers and the general public. One of the most important impacts is how a new transit system changes travel and residential and business location decisions, and, subsequently property values of nearby parcels. This report surveys the theoretical and empirical literature in this area to provide guidance to planners of the upcoming CenterLine project in Orange County, California. Economic theory suggests that the arrival of public transit can change the amenities associated with a given set of parcels. Namely, residents who use the transit system may enjoy reduced time traveling to employment, retail, and cultural opportunities while businesses near a transit station can face lower costs and agglomeration benefits. Thus it is often assumed that properties located near a station enjoy a premium over those farther from public transit. Some property owners may suffer a penalty from the nuisance effects of a rail system, but the net impact on the relevant residential property market should be positive. A small set of office, retail and industrial properties should enjoy equivocal positive premiums. The report then provides an overview of the common research techniques to empirically test the theoretical link between transit and land values. Of the three common methods case comparisons, the matched pairs method, and multiple regression (hedonic) models the later is preferred since it incorporates how non-transit attributes of a property also affect observed prices. The true premium or penalty associated just with the arrival of rail transit can be isolated, although this methodology requires intensive data collection and price observations well-after the construction of a transit system. The actual literature on how transit affects residential and commercial property values relies most heavily on hedonic price regressions, with varying results across cities and transit system type. Light-rail transit has enhanced residential property values some 2-18% in Portland, Sacramento, San Diego, and Santa Clara, with larger changes in cities with commuter rail systems. But not all residences benefit equally. Properties located too near a station may suffer nuisance effects, and it appears that in California the largest premiums accrue to owners of multifamily residential properties. Of the few commercial property markets studied, it appears that there are premiums of 4-30% for office, retail and industrial buildings located near rail transit in Santa Clara, Dallas, Washington, DC, Atlanta and San Fransisco. The report concludes that a comparable study of the CenterLine would best be undertaken after system construction is finished. Answers to several questions concerning system ridership uses and the mix of current properties near the proposed lines can provide some initial indications of the possible property value changes due to transit. Additionally, it appears that the CenterLine in its current alignment will be focused on three distinct hubs each of which could experience real estate market changes differently.
1 I. Understanding the links between public transit projects and property values: preparing for the CenterLine in Orange County, CA The economic development impact of a light rail system can include many components, ranging from changes in commercial and residential property values, rental space, to new business startups and employment opportunities. Property tax collections, commercial sales and tax revenues may then be enhanced. Some of these impacts can be accurately assessed, while others remain tentative and theoretically difficult to model. Given the importance of property tax revenues to county budgets, and the likely changes in property values, the following report concentrates on the real estate component of the CenterLine s overall economic impact. This report reviews numerous relevant studies of the links between public rail transit and property value changes. It demonstrates how both light and heavy rail transit projects can influence different property types. It provides tables compiling the salient results of interest, with an eye to drawing out the most relevant cases for Orange County. For instance, there have been successful light-rail systems implemented in San Jose, San Diego, Dallas, and Portland, with new projects underway in places like Minneapolis. The literature on commuter rail systems has examined the BART, Washington, DC Metrorail, and Atlanta MARTA cases in-depth. Consultants to these projects have successfully studied changes in tax revenues and property values, and their study methodologies will be analyzed. Additionally, the economics and social science literatures offer numerous refereed-publications about transportation system economic impacts that lend methodological validation. The first section of this report provides an overview of the proposed new light-rail system in Orange County, the CenterLine Project. Next, the theoretical links between public transit development and real estate market trends are explored. That section highlights three common methodologies to empirically determine if public transit provides a premium or penalty to existing property owners. The next section includes the results of a literature review of several dozen studies across fifteen rail systems. It provides a framework of the variables of interest in each of the study sites, so that they may be compared to Orange County It summarizes the results of each study and draws out lessons concerning the extent and likelihood of property value increases in these cases. Generally, heavy rail systems have provided greater premiums to property owners than light-rail system, but other city-specific factors besides system scale and ridership influence the transit-real estate link. Additionally, commercial properties appear to enjoy higher premiums than do residential properties. The last section of this report comments on the relevance of the methodologies used and the specific data requirements that appear necessary to replicate such studies in the future. For instance, it is most likely that property value and demographic time series data in Orange County before and after system construction are needed for any accurate analysis of the CenterLine project. However, some initial observations about the demographics and property distributions of areas near the proposed CenterLine stations are included.
2 II. The CenterLine Project Traffic congestion and increased population densities have strengthened the need for enhanced public transit in Orange County, California. By late 2001 numerous citizen and public organizations had called for a light rail project, and the OCTA (Orange County Transportation Agency) responded. The design of the CenterLine Project was formalized and a specific alignment running from Santa Ana to Irvine was chosen. Currently preliminary engineering and environmental analysis projects are underway with a final approval scheduled for December 2003. Construction would then occur between 2005 and 2010, with revenue service beginning in late 2011. The CenterLine project would be a single-line, 11 mile, 9 station light rail (LRT) system concentrated in the center of Orange County. It will include important county locations such as UC Irvine, the John Wayne Airport, South coast Plaza, and the Santa Ana Civic Center. Connections to other commuter and heavy rail systems (Metrolink and Amtrak) will be available through the Santa Ana station. Most of the system will be elevated with trains operating at a frequency of every ten minutes during peak hours. Its projected ridership is about 8 million passenger trips per year, comparable to several other LRT systems in the United States. CenterLine proponents suggest the model of LRT established in Dallas, Denver, Portland, and San Diego can be replicated in this increasingly congested metropolitan area. The main purpose of the CenterLine will be to reduce intra-county travel time. Riders will complete trips faster than those in automobiles (especially during peak periods), and the general public will benefit as some automobiles are removed from county roads. Other projected benefits include a contribution to a cleaner air environment in the county, more retail and commercial activities around stations, resulting net increases in some property valuations and then higher tax collections for the county government. This study focuses on this last projected benefit the impact of rail transit on property valuations.
3 III. Review of past studies of transit s impact on property values A.Introduction: Theory and Methods The general theory that property values increase with improvements in public transportation is well established. However, it must be understood that the theory addresses net benefits to the economic region, not that each and every parcel impacted by a transportation project will necessarily and immediately gain in value. Public transportation is a public good; residential property and commercial property combined can expect net increases in value but, there are several complicating issues specific to each transit project that will determine which areas receive benefits or bear some of its potential costs. Transportation investments are intertwined with the economic performance of a region. There is some debate among planners and researchers as to whether transportation investments occur in response to growth pressures or whether transportation accessibility improvements stimulate new economic growth. In either case, changes in the demand for commercial and residential properties become capitalized into property values and act as a reliable indicator of regional economic performance. Transportation access improvements should impact property values by providing greater access for commuters and reducing transaction costs for commercial businesses (Swenson 1998). Regional and urban economists agree on the theory that access to transportation services are capitalized into the price of existing commercial and residential properties. Economic theory also predicts that changes and/or improvements in transportation accessibility will be mirrored in the prices of those properties affected. However in practice, accurately measuring the fraction of a property s value attributable to transportation accessibility is a challenging exercise from the practitioner s perspective. The discussion that follows provides a brief review of the theories on commercial and residential property valuation. A short dialogue on the most frequently used methods to measure the capitalization of rail transit in both commercial and residential property will track the theoretical portion. Lastly, this report includes a review and summary of key issues and potential complications that are inherent in estimating property value impacts from light rail investments. 1) Property Value and Location Choice Theory The market price of land and the amenities that are associated with the property have a value equal to the present value of rental income generated by the parcel. n R PV = + t (1 i) where PV = present value R = rental in year n I = interest/ discount rate T = year
4 Economic theory assumes that the supply of property is essentially fixed, exhibiting a nearly perfect inelasticity in the near term time horizon. If demand for property increases due to a transportation amenity, price adjusts upward quickly and only after a period of time will supply of available property shift out to meet the increased demand. Different forces motivate individual residential and commercial business location decisions. Early theories on residential location choice away from a city center were predicated on the concept that household location choice was a simple matter of balancing marginal commuting costs with any marginal benefit of increased distance from the city center (Muth, 1964). The marginal benefit that the household receives is the decrease in land rent multiplied by land consumption. The marginal cost to the household is the increased cost of additional commute distance. Since public transit would reduce commute costs, it should promote movements to properties nearer station and thus bid up these properties values. Of course many other factors are embedded in location choice, and a detailed description is presented in the methodology section of this report. On the commercial side, property rent is an input cost to firms just as capital and labor expenditures are. Businesses are assumed to maximize profits and choose a location that is best from this perspective. Thus firms locate their operations based upon the location and density of its current workers and potential workers; location of its competitors; location and density of complementary firms; demographics characteristics and location of its customers; and the transaction costs of supplying its products to its customers. Manufacturing, retail, and office firms have relatively common production functions within their respective general industry classifications. Public transit can enhance clustering of businesses and agglomeration benefits. Businesses are able to reduce costs by sharing a pool of workers that have similar training. Input costs on materials may be lower when firms cluster because resources are shared among firms with similar production input requirements. We often observe office firms clustering, manufacturers clustering in industrial parks and retail firms clustering at shopping centers. Also, as employees reduce their commuting costs, lower wage premiums would be needed to attract workers. But there are several mitigating characteristics which affect whether a specific transit leads to significant empirical increases in nearby property values and the extent to which a system in one city produces higher benefits than a system in another city. First, Cevero (1984) first suggested that light rail systems should confer smaller benefits than commuter and heavy rail since the level of service was necessarily smaller. Overall higher quality systems (with more stations, frequency of trains, adjacent parking lots, and thus a larger ridership) should confer the greatest capitalization benefits. Second, the capitalization benefits are expected to be most significant several years after a system opens as the property market undergoes an adjustment period. Generally studies completed five years after the arrival of a transit system represent the most accurate results. Numerous studies have been undertaken in anticipation of a system opening; these studies may not correctly demonstrate permanent changes in property values but rather initial land speculation effects. Third, the sign and magnitude of capitalization due to transit should vary even among residences and offices near the new system. A
5 positive capitalization is expected for sites near (within ½ mile walking distance) a station, but there are possible negative effects for properties adjacent to the line itself and properties too close (i.e. less than 300 meters) to a station that could suffer the negative nuisance effects of noise, congestion, and increased crime. Fourth, the effects of capitalization may vary across low and high-income neighborhoods as each may have a different valuation of commute time-savings. Fifth, city-specific factors such as the presence of joint development transit projects, competing freeway access near the transit station, and overall employment trends and health of real estate market. 2) Methods to Estimate the Impact of Rail Transit Investment on Property Values Whether the predictions of theory occur in reality is an empirical question. There are two general categories of empirical public transit studies: predictive models that forecast future economic benefits of a transit project, and evaluative models that compare economic conditions before and after a transit investment is completed. There are many methods available within these two broad classifications but we limit our discussion to those methods that are best suited to estimate the impact of the CenterLine light rail project on commercial and residential property values. The three general methods presented here are case comparison, matched pairs, and multiple regression statistical models. Each method has advantages and disadvantages; there does not exist one best methodology for estimating property value changes from public transportation improvements. Although there is no one best methodology, hedonic multiple regression is the most extensively used technique to estimate the portion of property value attributable to transit access. Method 1) Case Comparison Many studies include a review of the experiences of other similar transit projects that have taken place in other cities. Case comparison studies usually comprise a coupling of a literature review with surveys and interviews with local business representatives, developers, and planners from other cities that have experience in transit investment projects. The comparisons allow investigators to get a general picture for how economic growth and development in other communities has been influenced by transit projects. Case comparisons are frequently used in predictive studies as supportive information demonstrating the impacts experienced in other communities from rail projects. Comparison studies are ideal for researchers to get an understanding of the potential impact of transit on particular economic variables such as the impact of a light rail transit system on property values. A detailed evaluation of transit studies and their economic impact on property values provides a sense of comfort to decision makers because they have a detailed summary of the experiences of other cities and transit agencies. Lastly, good comparison studies identify key steps (such as supportive parking policies, zoning regulations, joint development contracts) that could possibly be implemented to complement the proposed transit project.
6 There are disadvantages to case comparisons. Differences between communities are difficult to control for, so that the experiences of one city cannot usually be replicated in another. The overriding disadvantage of case comparisons studies is that each transit investment is unique in terms of its spatial characteristics as well as the demographic and economic circumstances, in which the transit investment is planned (Cambridge Systematics 1998). Method 2) Matched Pairs Method A simpler straightforward comparison of data on development patterns is also possible. Property prices before and after a transit system is constructed and operating can become the sole focus of analysis. Investigators compare the differences in property price in an area that is near transit development with property values in areas that has similar attributes but near another form of transit system such as a highway interchange as a control area (Cambridge Systematics 1998). The difference in sales price or rent is the value-added of a transit system. The general matched pair framework is set up in the following way: Where: Effect of Transit = (I TA -I TB) (I CA I CB ) I = Property values Impact T = Transit corridor being studied C = Transit system control area B = Before transit investment A = After transit Investment Simple difference comparison statistics can provide probability estimates of whether the observed changes in the comparison area are significantly different. If the differences are statistically significant than claims can be made that transit development had a positive impact over and above currently existing alternative transit areas. Also, the matched pairs method can be used to compare the new transit development area with a similar nearby location that lacks transit accessibility. Both the control and study areas are then studied over time or cross-sectionally. The matched-pairs method requires much less data than regression techniques and provides a level of transparency that is readily understandable for the average citizen. Data on property values can be obtained from the county assessor s office or from local government planning agencies that have an interest in the project. There are drawbacks of the matched pair methodology of assessing transit development s impact on property values. Because of the uniqueness of property location comparing two different properties inherently introduces the possibility of errors. How close the control area is to the study area, demographic characteristic variations, nearby highways, etc., can have an influence on the results. The matched pairs method does not have control mechanisms to account for the differences just mentioned, hence criticisms can arise as to how the researcher chose the control areas (Cambridge Systematics 1998).
7 Method 3) Multiple Regression Model: Multiple regression analysis is the chief methodology used in studies attempting to isolate the impact of transit s presence on property values because regression can separate and isolate out the influences of many non-transit factors (e.g., structural characteristics of the home, freeway proximity, neighborhood crime, school quality, etc.) that reasonably should influence price. Housing is heterogeneous in that each dwelling offers to consumers a set of characteristics that are unique. Two main categories of housing characteristics are: residence features and site features. The residence features are essentially the physical attributes of the dwelling. Residences differ in square footage and structural layout. They also have differences in interior living quality such as heating and air conditioning. Interior design differences are reflected in such things as types of windows, flooring, and cabinetry. Each home offers a unique grouping of home size, layout, and structural integrity and aesthetics. Site characteristics include such things as: job accessibility, nearness to shopping, and recreation choices. Other site characteristics are the available public services such as: fire services, police protection, and neighborhood schools. Environmental qualities unique to the home s geography include such things as air quality noise levels and possibly odors. Also, it must be noted that the exterior visual characteristics of other homes located in the vicinity are also purchased. In summary, the value of a home is made up of a bundle or set of characteristics that can be classified as: structural (the physical attributes of the home), environmental, and lastly neighborhood specific features. The hedonic valuation approach is founded on the idea that the individual dwelling components that make up the bundle of household characteristics each have an implicit price. All of these site-specific benefits are theorized to be embedded, or capitalized, into property values and in competitive land markets (Alonso, 1964; Muth, 1964). When multiple regression analysis is used to estimate the values of different characteristics of a property, the technique is often called hedonic price modeling (Rosen 1974). Rosen developed a theoretical competitive equilibrium model where buyers and sellers trade a multidimensional good, (residential dwelling/commercial property) which has a number of distinguishing characteristics. Buyers of property seek to achieve the highest level of satisfaction by choosing to purchase many goods, including properties with certain characteristics. Sellers of properties consider the costs of managing the property and set its price along property characteristics. A market equilibrium represents the maximum price that consumers will pay for a set of characteristics (Rosen 1974), (Freeman, 1979). Multiple regression hedonic models estimate observed property prices as a function of variables (including transit access) as suggested in the Rosen model. The general setup of hedonic models is similar in spirit for both commercial and residential property analysis. A profile of a typical hedonic model generally takes the following form outlined in Cervero and Duncan (2002); this profile is included in Appendix A. The drawback of regression methods is that they are very data intensive, which implies that costs can be relatively high.
8 B. Previous residential property studies These general lessons from the theoretical overview are relevant in considering the actual benefits transit has empirically conferred on residential values (single-family home sales prices and apartment rents) and commercial values (office and industrial buildings rents and sales prices). A summary of the case studies below is incorporated in the tables of Appendix B. We separate residential and commercial properties since these markets could be driven by a different underlying dynamic between transit and values. For instance, transit changes may be capitalized into residential values since commute time savings accrue to one house over another, whereas owners and leasees of commercial properties are more concerned about how the reduced commute time could lessen the need for employee wage increases and increase the employee pool. Additionally, transit may have a decentralization effect as rail can stimulate suburban residential development away from the Central Business District while it could have a clustering effect in downtown employment in older cities. A city-by-city summary of the literature on the residential property value studies is presented in Appendix C, with those cities operating light rail systems considered first. In most cases comparability across the cities and the studies is nearly impossible. We highlight the mitigating characteristics of the transit-property value link by drawing attention to the rail type, study timing, measure of distance, and city-specific issues. Additionally, each study represents a different real estate market location and period. The results of each study are not presented in percentage terms for comparability; many studies report absolute dollar changes per distance measure which represents a very different result depending on whether the city generally is a high-priced or low-priced real estate market. However, in an attempt at comparability, PriceWaterhouse Coopers (2001) reviewed many of these studies and conclude that residential properties near a station see a positive premium of 0-5% following the arrival of a transit system. The premium is highest for those properties located between ¼-1 mile from a station. However, for residential properties along segments between stations there is a potential negative valuation of 5-10%. Thus clearly some property owners will enjoy a positive valuation externality due to a public transit project while others will not. Nevertheless, the literature points to a general trend of a net positive capitalization of some residential property values following the arrival of a transit system. Studies in Boston, Philadelphia, Portland, San Francisco, Arlington/ Washington D.C., Atlanta and San Diego found that residential properties with close proximity to rail stations had higher property values than those farther away. But higher residential property values are not apparent in San Jose, Sacramento, and Miami. These rail systems probably weren t as high quality as the others or they enjoyed very low ridership. Higher system ridership tends to increase positive property premiums throughout the transit area.
9 C. Commercial property studies There has been far less analysis of the synergy between public transit and nearby commercial (retail, office and industrial) values. Theoretically the presence of a transit station should enhance a nearby commercial property by providing benefits to both employees and owners. Thus in some aspects a public rail transit system may have a larger impact on commercial values than residential, particularly if a high-quality rail transit system is in place. In several studies commercial property received the largest land value premiums and discounts (Cevero and Duncan, 2002). But commercial property data is more obtuse than residential home prices. Rental data (asking rents) could be linked to occupancy rates and hide many components of an office contract beyond comparable properties near or far from transit. The building sales price and land value is more appropriate (Cevero, 1994), yet few studies have incorporated this data. An summary of the relevant studies of how rail transit has impacted commercial property values in seven cities in included in Appendix C. Most studies of the link between public rail transit and commercial properties have compared properties within walking distance to a station (< 1000 ft.) and those farther away. The areas between stations along a line are generally not applicable. Price Waterhouse, Coopers (2001), in a lengthy review of commercial studies, suggest that the nearer properties generally command a premium of $2-4/ sq. ft. over those sites away from transit. But the results have been very different across cities and transit systems and study timing (anticipation of transit or post-transit). Studies of Washington, D.C. (Damm, et.al., 1980) and Atlanta (Cevero, 1994) demonstrate properties near the rail station rise faster than those farther away but no-long-term effects are observed. Retail and office properties in Dallas and Santa Clara have enjoyed 4-30% premiums for being located near rail transit stations (Weinberger, 2000; Weinstein and Clower, 1999). Commerical property values were shown to rise in anticipation of rail transit in Los Angeles (Fejarang, 1994) and San Fransisco (Dyett, 1979). Yet studies suggest the premiums in San Fransisco have not lasted over time or been limited to a very short distance from stations (Cambridge Systematics, 1998). Capitalization benefits have been limited to only a subset of hubs in San Diego (Ryan, 1997; Cevero and Duncan, 2002). D. Overall findings in the literature The published studies of the capitalization effects of public rail systems supports some of the theoretical ideas mentioned above, but draws out some nuisanced trends. Some of these issues may be particularly relevant for future light-rail programs in California, as the study results are affected by city and cultural differences as much as economic factors. Some of the important trends in literature as regards to the rail system scale, type of system, timing of analysis and study methodologies are discussed next. First, across the fourteen city systems analyzed, inherent cultural differences may be as important as system scale and type in mediating how a land market reacts to public transit. All of the systems incorporate at least 15 miles of track or 15 stations, with at
10 least 2 million passengers per year reported at the time of the study. Among all types of rail systems, the reality of a shared freight right-of-way has caused nuisance effects for some residences both near stations and lines, thus limiting property value increases postconstruction. And the presence of competing freeway access near a station tends to diminish the benefits of rail transit per se. As expected, the capitalization premiums appear to be higher with the establishment of heavy and commuter rail systems. Turning to the light-rail systems, Portland and Dallas have smallest scale operations as regards the distance and stations included. But the land market changes in Portland have been more extensive than those in cities with larger systems (such as Miami). A strong Smart Growth planning initiative in Portland and attempts to install a culture of reducing congestion have enhanced ridership and the premium residential owners place on being located near public transit. Authors of studies in both Miami and Dallas, where the effects on residential properties were actually negative, suggest that the car culture of these cities has limited public enthusiasm for light rail and thus dampened property value impacts. The studies of various systems in California indicate smaller, and highly varied, land market effects of public transit. Earlier studies focused on only average changes in residential housing prices, but later studies have picked up differences within the residential market. It now appears that the owners of single-family housing in California are not particularly enthusiastic about the presence of public transit, while residents of condominiums and apartments demand more of these services to reduce travel time. This trend could differ from other parts of the country where property capitalization premiums have tended to be larger in higher-income neighborhoods. Apparently wealthier residents in older cities place a high valuation on the travel cost savings public rail transit provides whereas California residents may be incorporating other concerns in their property valuations. Second, while there are fewer studies of transit s effects on commercial property values, the general trend has been positive for the well-established systems. In cities where both the residential and commercial markets have been analyzed (such as Washington, DC and Atlanta), higher premiums occur in the later. Yet the California studies of both light and heavy seem to represent some exceptions. In San Diego the commercial benefits appear to be concentrated only in the central business district rather than any suburban office zones. And in San Fransisco and Los Angeles few capitalization have appeared for commercial properties located nearer transit stations. Most analyses of these cases have suggested that varied city economic development policies could be contributing to these real estate trends. For instance, Atlanta and Washington have undertaken strong transit-oriented development programs while the Los Angeles Red line rail was placed within a less successful redevelopment zone. Finally, some lessons about methodology and analysis timing appear from the previous studies. The literature is equally divided into whether transit access should be just a positive variable in a statistical analysis (as a dummy ) or whether the actual linear distance from the station affects economic trends and property values in different
11 ways. Those studies which have statistically incorporated distance (as a continuous variable) still have not adequately explored if this geographic trend is strictly increasing so that nearer properties command higher premiums or whether some other mathematical relationship (i.e. reciprocal or cubic forms) are more appropriate. And the hedonic price regression technique is favored by most authors for its ability to control many factors reflected in property prices apart from transit access. The method, of course, assumes that property prices have adjusted to include the impact of transit location on the economy. Thus a consensus is emerging that such studies using hedonic methods are best employed well-after a public transit system has been constructed. The anticipation studies measuring property values before a system is operating show benefits to sites nearer transit stations, but this empirical result is difficult to explain. Studies completed right after a system construction tend to show small premiums for transit access, but after residents have adjusted their travel habits over time, property values increase further. Thus the group of studies taking place at least five years after a system opening reflect higher, and perhaps more accurate, representations of how transit affects the prices of homes and commercial spaces.
12 IV. Extending the analysis to the Orange County CenterLine Clearly, an empirical determination of how transit has changed property values historically is the preferred method to settle this debate. But this goal relies upon data collected several years in advance and after the construction of a transit system. For the Orange County case, this most rigorous hedonic price study of how the light-rail affects property owners would not be available until after the line has been open for an adequate period to allow real estate adjustments, i.e. 2015. In such a study, detailed sales transaction data for commercial and residential properties located with one mile of transit stations, and those farther away, would be required, alongside with general demographic and income characteristics of those census tracts. Thus either proprietary real estate database information, or property tax assessment data, would be combined with economic indicators possibly using a geographic information systems software. A matched pair study would analyze data from properties along the CenterLine with comparables farther away at two points of time, say in 2005 and 2015. Anticipatory studies rely upon a different timing, i.e. real estate data collected at system pre-announcement (before funding) up to a time midway through construction (pre-opening). The empirical results of this type of study could include some speculatory changes in property values appearing in residential properties, but the largest effect would most likely be in the commercial sector. Forecasting future property values alongside upcoming transit stations using simulation techniques with other region s data is another alternative. Generally, these methods are not the preferred approaches to accurately measure the link between transit and property values. However, such a study would most likely be done through looking at Metroscan or non-proprietary database of housing sales and/or commercial property rents/ sales say in 2000-1 and 2010 in Orange County, compared possibly with real estate market trends in Los Angeles County. Given the current concern among some county residents about their property values and the CenterLine, at least some salient observations from the previous literature can be applied. In the discussion below we draw out two important observations and questions which could assist planners in roughly predicting which census tracts of Orange County stand to see property value premiums. We also provide a general picture of current housing prices, commute times and demographics of three main proposed station nodes of the CenterLine project. This provides some indication that population cohorts may stand to benefit from the CenterLine. First, in reviewing the public documents of the Orange County project, it appears that the CenterLine is of a smaller scale than most previous light rail transit systems. However, it has a high projected ridership. Only Los Angeles Blue and Green LRT lines cover about the same lengths; but these zones offer very different demographic profiles which preclude easy comparison. Orange County s upcoming experience could be closest to that of Dallas, given the systems size, freeway competition, and partial freight right-of-way concerns. In addition, the literature on the Dallas case is a bit vague as regards to the system s primary users commuters or occasional riders. Residents of
13 both metropolitan areas rely heavily on freeway transit, and this reliance reduces the property value premiums. Thus, two outstanding questions are: 1) What is the profile of projected users on the CenterLine project, in its current alignment? How would this differ with an expansion of the system? 2) Is this clear commuting pattern between residents and workplaces compactly within the line? Or will notable stations such South Coast Plaza and the John Wayne Airport primarily be attracting tourists and residents interested in cultural and retail activities? Second, other California studies have highlighted great variation in the transitrelated premium or penalty across different residential and commercial property types. The local trend appears to be that commercial properties (especially retail properties) near stations have greatly benefited from the arrival of transit in places such as Santa Clara and San Diego, but only a subset of residential owners have enjoyed the premiums. Namely, condominium and apartment complexes, rather than single-family housing, have demonstrated significant appreciation in the years following the construction of a nearby public transit station. Very little data is available concerning any type of residence values located along rail line, particularly when a freight-of-way exists. Thus other important questions include: 3) What is the profile of residential property types within the projected Centerline corridor? What is the mix of single-family housing and apartment complexes near stations? 4) How many houses are located near the proposed rail line per se, not a transit station? And what will be the terms of right-of-way acquisition of some parts of the rail line will freight service be a small or significant burden to nearby residents? 5) And which properties would show a specific impact from the CenterLine apart from any previous freeway access or other transit construction project? The current profile of properties near the CenterLine project can at least begin a discussion on some of these issues. The attached graphics in Appendix D illustrate both the county average and subregional-specific trends in single-family residential home prices and demographics. A clear pre-light-rail transit differential emerges among three potential nodes of the system. The Civic Center Santa Ana node exhibits a trend of lower housing prices, lower family incomes, and higher commute times to various workplaces. The Irvine hub represents of an area of initially higher property prices, higher family incomes, and shorter commuting times compared to the county average. The SouthCoast Plaza node falls in the middle of these extremes. Next, the graphics turn to increasing distances from a central geographic hub point. Recall that properties closer to such a hub may demonstrate higher or lower values
over comparables depending on the mix of beneficial and nuisance effects. These current hubs are locations with both business concentration and existing transit such as buses and freeway access. In the 1995 Orange County data it appears that distance from a hub had varied impacts across the three sites. Properties farther from the central Civic Center node appeared to be priced higher per sq. ft. than those nearer, whereas those near the Irvine concentration point had premiums over comparables. Whether this trend will continue or be reversed after the arrival of a new light-rail transit system could be related to the residential mix and commuting patterns, as brought up in the questions above. 14
TRENDS IN PROPERTY VALUE CHANGES RELATED TO RAIL TRANSIT --Analysis includes 14 systems opened by 1996 with moderate to large-scale operations (i.e. a minimum of 15 stations or 15 miles of track) Residential Property Studies Light-rail transit has enhanced residential property values some 2-18% in Portland, Sacramento, San Diego, and Santa Clara Higher residential property premiums and penalties have occurred in cities with commuter rail systems (San Fransisco, Washington, Atlanta, Boston, Chicago and Philadelphia) The benefits are highest for properties located within 1 mile from a station, but not too near to receive nuisance effects No studies yet have determined if a strong penalty exists for homes located along a rail transit line, but not near a station. Some reference has been made to freight right-of-way hurting property values. Few premiums have occurred in cities with low ridership or a carculture (Dallas and Miami), and in some cases the premiums have been larger in high-income neighborhoods than in poorer areas California studies point to higher premiums for multifamily residential properties (apartments and condos)
Commercial Property Studies Fewer studies completed of this real estate market Premiums of 4-30% for office, retail and industrial buildings located near LRT in Santa Clara and Dallas, but not a widespread effect in San Diego (only concentrated in downtown and Mission Valley) Higher and more established premiums in cities with commuter rail systems (Washington, DC, Atlanta and San Fransisco); anticipatory premiums in Los Angeles have not held over time Transit-oriented development and central business district revitalization programs interact with the effect of transit in the observed property values Capitalization benefits usually limited to properties located within walking distance (< 1000 ft.) from a transit station
15 APPENDIX A HEDONIC PRICE REGRESSION PROFILE Pi = f(t, A, S, C), where: P i equals the estimated price of parcel i; T is a vector of transportation services, including proximity to transit and highways and accessibility via highway and transit networks; A is a vector of property (e.g., structure size and age) and land-use (e.g., type of commercial) attributes; S is a vector of neighborhood socio-demographic characteristics (e.g., mean household income); C is a vector of controls (e.g., municipality and time- series fixed effects) A. The specific model employed by Cervero and Duncan (2001, 2002) in their analysis of commercial property values and their relation to the Santa Clara Light-Rail transit is recreated for expositional purposes below. Dependent Variable (Pi): Land Value: Commercial parcel land value per square foot ($,1999) Independent Explanatory Variables (T, A, S, C): Rail/Hwy Proximity LRT: within ¼ mile of LRT station (1=yes;0=no) Commuter Rail: within ¼ mile of CenterLine station (1= yes; 0=no) Freeway: within ½ mile of grade separated freeway or highway interchange (1= yes; 0=no) Accessibility & Location Regional Labor Force Accessibility: Number of employed residents (in 100,000s) within 45 min. peak travel time of highway network Downtown Santa Ana: within ¼ mile of CenterLine station LRT station in Downtown Santa Ana (1= yes; 0 =no) South Coast Plaza: within ¼ mile of CenterLine station LRT station in South Coast Plaza (1= yes; 0 =no) Irvine Hub: within ¼ mile of CenterLine station LRT station in Irvine(1= yes; 0 =no)
16 Density & Land Uses Labor Force Density: No. employed residents per gross acre within one mile radius of parcel Non-Employed Resident Density: No. non-employed residents per gross acre within one mile radius of parcel Single-Family Housing Density: No. single-family housing units per gross acre within one mile radius of parcel Service Employment Density: No. service employees per gross acre within one mile radius of parcel Retail Employment Density: No. retail employees per gross acre within one mile radius of parcel Manufacturing Employment Density: No. manufacturing employees per gross acre within one mile radius of parcel Other Employment Density: No. other (not professional, service, retail, manufacturing) employees per gross acre within one mile radius Professional-Office Land Use (1=yes; 0=no) Institutional Uses: Square feet of public and institutional building area per gross acre within one mile radius of parcel Neighborhood Quality Proxies Building Values: Weighted average value of structures and improvements per square foot ($, 1999) for all properties within one mile radius of parcel Household Income: Mean household income (in $10,000, 1999) of households within one mile radius of parcel
Appendix B Summary Effects of Rail Transit on Residential and Commercial Property Values 1) Cases of Light-Rail Transit and Residential Property Values 2) Cases of Heavy-Commuter Rail and Residential Property Values 3) Cases of Light Transit and Commercial Property Values 4) Cases of Heavy-Commuter Rail and Commercial Property Values
SUMMARY EFFECTS OF RAIL TRANSIT ON RESIDENTIAL PROPERTY VALUES CASES OF LIGHT-RAIL TRANSIT SYSTEMS Authors Research Technique Timing of Study Type of Property Values 1) San Diego: 34 stations across 40 miles; opened in 1991 Landis, et.al. Hedonic Postconstruction SFH sales prices (1994) < Cevero & Duncan (2001) Hedonic Postconstruction > SFH, condo, apartment sales prices 2) Portland: 32 stations across 15 miles; opened in 1986 Al-Mosaind, et.al. (1988) Hedonic SFH prices Postconstruction < Chen (1998) Hedonic Post construction 5 years SFH prices 3) Santa Clara-San Jose: 33 stations across 39 miles; opened in 1987 Landis (1994) Hedonic Postconstruction SFH prices < 4) Sacramento: 28 stations across 36 miles; opened in 1986 Landis (1994) Hedonic Postconstruction < SFH prices Transit Access Measure Linear ft. to station; also dummy variable Distance rings ¼- 1/2 mile from stations Dummy for access in distance rings; linear distance < ¼ mile Linear ft. to station Linear ft. to station and distance category Linear ft. to station and distance category Premium/ Discount Found +$337 or.1% for every 1000 ft. closer to station; but 4% for close units < 900 ft. to station +2-18% for condo, apartment units if near; 0 to 4% lower for SFH +10.6% premium for 500 meters near station + premium for houses closer to station, but highest at 700 ft. distance +.1% for every 1000 ft. closer to station, but 10.8% penalty for units too close within 900 ft. 0.4% for every 1000 ft. closer to station, with 6.2% premium if very near station Comments City of San Diego only Controls for freeway access Distance measured only up to 1500 ft. from station Measures extent up to 7 miles from station Results not statistically significant 5) Los Angeles: 36 stations across 42 miles; opened in 1990 Cevero and Hedonic Duncan (2002) Postconstruction> 5 years SFH, condo, apartment sales prices 6) Dallas: 15 stations across 20 miles; opened in 1996 (expanded to 39 stations) Weinstein & Clower (1999) Matched pairs over time SFH prices Pre and postconstruction < Distance rings ¼- 1/2 mile from station Units < ¼ from station given transit access 1-3.5% premium for apartments, SFH near stations; -6% penalty for condominiums -5% penalty over time for units nearer stations Controls for freeway access, other public transit Value-added approach to premium