Citation
The Anticipatory Effects of Commuter Rail on Economic Development in Orange County, FL

Material Information

Title:
The Anticipatory Effects of Commuter Rail on Economic Development in Orange County, FL
Creator:
Lytle, Benjamin F
Place of Publication:
[Gainesville, Fla.]
Florida
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (19 p.)

Thesis/Dissertation Information

Degree:
Master's ( M.A.U.R.P.)
Degree Grantor:
University of Florida
Degree Disciplines:
Urban and Regional Planning
Committee Chair:
STEINER,RUTH LORRAINE
Committee Co-Chair:
ZWICK,PAUL D
Committee Members:
ALAKSHENDRA,ABHINAV
Graduation Date:
8/9/2014

Subjects

Subjects / Keywords:
Anticipation ( jstor )
Buildings ( jstor )
Counties ( jstor )
Housing ( jstor )
Jurisdiction ( jstor )
Land use ( jstor )
Rail transit ( jstor )
Statistical models ( jstor )
Statistics ( jstor )
Transportation ( jstor )
Urban and Regional Planning -- Dissertations, Academic -- UF
anticipatory -- building -- county -- development -- economic -- land -- orange -- orlando -- permit -- planning -- rail -- sunrail -- tod -- train -- transit -- use
City of Orlando ( local )
Genre:
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Urban and Regional Planning thesis, M.A.U.R.P.

Notes

Abstract:
This study quantifies the anticipatory effects of the SunRail commuter rail line on property values, building permits and land use in Orange County, FL from 2007 to 2013. Parcel data for each year from 2007 (when the interlocal agreements necessary for SunRail to become a reality were signed) was analyzed to determine the effect of proximity to SunRail on property values using a hedonic pricing model with linear distance, quarter, half, one and two mile buffer coefficients. Additionally, building permit data was assembled for each jurisdiction with a station to determine the share and total value of residential and nonresidential building permits near each station. Building permit data was investigated in greater depth for one station within each TOD typology present in Orange County (as described in Olore, 2011). The researcher found that the anticipatory effects of SunRail are predominantly related to increasing the value of residential properties located closer to stations. The share of building permits and building permit value located near stations increased between 2007 and 2013, with an increasing shares of permits near non-downtown stations. Finally, the areas near SunRail have maintained their original mix of land uses except for a reduction in institutional land uses. Overall this suggests that SunRail did not substantially change the mix of land uses near stations, but that residential properties near stations increased in value. Downtown and Urban Center stations had smaller gains in the share of building permits near stations than Village Center stations, because those have greater potential for growth. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis:
Thesis (M.A.U.R.P.)--University of Florida, 2014.
Local:
Adviser: STEINER,RUTH LORRAINE.
Local:
Co-adviser: ZWICK,PAUL D.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2016-08-31
Statement of Responsibility:
by Benjamin F Lytle.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Embargo Date:
8/31/2016
Classification:
LD1780 2014 ( lcc )

Downloads

This item has the following downloads:


Full Text

PAGE 1

http://eab.sagepub.com/Environment and Behavior http://eab.sagepub.com/content/43/6/789 The online version of this article can be found at: DOI: 10.1177/0013916510392030 December 2010 2011 43: 789 originally published online 17 Environment and Behavior Barbara B. Brown and Carol M. WernerStop: Do Residents Get What They Expect? The Residents' Benefits and Concerns Before and After a New Rail Published by: http://www.sagepublications.com On behalf of: Environmental Design Research Association can be found at: Environment and Behavior Additional services and information for http://eab.sagepub.com/cgi/alerts Email Alerts: http://eab.sagepub.com/subscriptions Subscriptions: http://www.sagepub.com/journalsReprints.nav Reprints: http://www.sagepub.com/journalsPermissions.nav Permissions: http://eab.sagepub.com/content/43/6/789.refs.html Citations: What is This? Dec 17, 2010 OnlineFirst Version of Record Oct 11, 2011 Version of Record >> at UNIV OF FLORIDA Smathers Libraries on March 3, 2014 eab.sagepub.com Downloaded from at UNIV OF FLORIDA Smathers Libraries on March 3, 2014 eab.sagepub.com Downloaded from

PAGE 2

Environment and Behavior 43(6) 789 þ­ The Author(s) 2011 Reprints and permission: http://www. sagepub.com/journalsPermissions.nav DOI: 10.1177/0013916510392030 http://eab.sagepub.com 392030 EAB 43 6 10.1177/0013916510392030B rown and WernerEnvironment and Behavior The Author(s) 2011 Reprints and permission: http://www. sagepub.com/journalsPermissions.nav 1University of Utah, Salt Lake City Corresponding Author: Barbara B. Brown, 225 S 1400 E RM 228, FCS Department, University of Utah, Salt Lake City, UT 84112-0080 Email: barbara.brown@fcs.utah.edu The Residents’ Benefits and Concerns Before and After a New Rail Stop: Do Residents Get What They Expect?Barbara B. Brown1 and Carol M. Werner1Abstract Transit-oriented developments are touted as providing a variety of social benefits, but personal benefits to residents are underresearched. The authors surveyed 51 residents before and after a new light rail stop was constructed in their revitalizing Salt Lake City neighborhood. Residents anticipated and then later experienced increased housing and neighborhood economic values, enhanced sense of community, and improved neighborhood reputation. Residents experienced greater than anticipated pedestrian and child safety after rail service started. Compared with resident perceptions of walkable neighborhoods elsewhere, the Salt Lake residents perceived their neighborhood to be denser, and offering less land-use diversity and more crime safety problems. Perceived walkability increased, with residents reporting greater land-use diversity and neighborhood satisfaction after rail stop completion. However, residents said more stores, parks and trails, and trees would improve walkability. These results show the personal benefits residents desire to make transit-oriented living a satisfying residential alternative. Article at UNIV OF FLORIDA Smathers Libraries on March 3, 2014 eab.sagepub.com Downloaded from

PAGE 3

790 þ En vironment and Behavior 43(6) Keywords transit-oriented development, neighborhood revitalization, perceived walkability, walking barriers and supports Developments around rail transit stops are often planned to mix compact housing with commercial and public spaces, designs intended to foster walking and rail transit use, especially within a half-mile of the rail stop (Calthorpe, 1993). Research shows transit-oriented developments (TODs) support multiple societal benefits, but less is known about whether TODs support residents’ personal benefits. We interviewed residents of an older city neighborhood before and after it received a new light rail stop as part of its multifaceted transition into more of a transit-oriented neighborhood. We explored how specific impacts of the new rail stop were both anticipated and experienced by residents. We examined residents’ perceptions of walkability measures for this neighborhood compared with walkable neighborhoods elsewhere. Finally, we summarized residents’ perceptions of barriers to walking and ways walkability could be improved. The societal benefits from TODs can include housing development and neighborhood revitalization, infrastructure efficiencies, decreased sprawl, and decreased automobile dependence and oil consumption (Dorn, 2004; Envision Utah, 2002). For example, with supportive policies and economic circumstances, TODs can facilitate the development of higher density housing (Handy, 2005), including affordable and subsidized housing (Hess & Lombardi, 2004). Cities, transit agencies, and others can build TODs to generate income (Cervero, 1994; Ridlington & Heavner, 2005) and increase transit ridership, especially among residents near stops (Lund, Cervero, & Willson, 2004). When light rail rides substitute for car rides, per passenger mile British Thermal Units [BTUs] of energy cost are about one fifth as high and pollution is substantially lower (Shapiro, Hassett, & Arnold, 2002). However, TODs are difficult forms of development to achieve because zoning, street standards, and lending practices typically favor other forms of development (Patton, 2007; Southworth & Ben-Joseph, 1995). Consequently, much of the research on TODs has focused simply on how to get the projects built, with special attention to zoning requirements and incentives (Belzer & Autler, 2002; Envision Utah, 2002). Although this work has been valuable, the ultimate success of TODs requires that they provide satisfying places to live. When areas transition into TODs, the expected and experienced neighborhood changes might be unwelcomed. The societal benefits of TODs, such as greater transit system use, might create negative personal impacts on at UNIV OF FLORIDA Smathers Libraries on March 3, 2014 eab.sagepub.com Downloaded from

PAGE 4

Brown and Werner þ 791 residents, such as elevated traffic safety concerns, perceived threats to property values, or loss of neighborhood satisfaction in the face of physical and demographic change. Indeed, in a sample of California residents, those living near heavy rail (Bay Area Rapid Transit) stations were more dissatisfied with their neighborhood, possibly because of rail noise (Baldassare, 1981). In London, residents anticipated that the construction of new underground transit stations would improve neighborhood conditions, but after the construction, residents reported less improvement than anticipated (Clark, Gatersleben, Uzzell, & Reeve, 2003). Adding affordable and/or denser housing to neighborhoods often evokes residents’ fears of negative consequences (Rose, 2004), even though these fears may dissipate once the housing is built (Briggs, Darden, & Aidala, 1999). Alternatively, providing rail service to residents who become light rail riders might enhance the fit of the neighborhood to personal preferences (Levine & Frank, 2007). Given the controversies associated with development in general and TOD in particular, surprisingly few studies have documented changes in residents’ perceptions, especially for light rail TODs in redeveloping neighborhoods (Hess & Lombardi, 2004). One challenge of transforming many neighborhoods into transit-oriented areas is that the neighborhoods were not designed for pedestrian use. Although some rail stops are constructed in previously undeveloped “greenfield” areas, many are placed where municipalities can easily aggregate land, such as along former freight rail corridors or highways. These major thoroughfares significantly hamper their walkability (Schlossberg & Brown, 2004). In such sites, the walkability and pedestrian friendliness of the rail stop area are expected to change and improve over time as the density and diversity of the land uses increase through redevelopment. However, it is unclear how those who already reside in the neighborhood will perceive the walkability of the neighborhood during its transition from underdeveloped to redeveloped. Thus, this study tracks perceived walkability before and after the rail stop becomes operational. In our study, perceived walkability was measured with the Neighborhood Environment Walkability Scale (NEWS; Saelens, Sallis, Black, & Chen, 2003), a scale intended to enable international comparisons in perceived walkability (IPEN [International Physical Activity and Environment Network], 2009). Scores from Salt Lake will be compared with scores residents reported in highand low-walkable neighborhoods in San Diego, California, and Adelaide, South Australia (Leslie et al., 2005; Saelens et al., 2003). In San Diego, high-walkability neighborhoods were chosen through researcher knowledge that they were characterized by relatively at UNIV OF FLORIDA Smathers Libraries on March 3, 2014 eab.sagepub.com Downloaded from

PAGE 5

792 þ En vironment and Behavior 43(6) high amounts of residential density, residential–commercial mix of land uses, and interconnectedness of street forms (Saelens et al., 2003). The low-walkability neighborhood, conversely, was chosen for its mostly residential land uses and less well-connected cul-de-sac street forms. In Adelaide, the highand low-walkable neighborhoods were chosen for similar features, but the features were identified through objective geographic information systems measures for density, land-use diversity, and interconnected street forms (Leslie et al., 2005). Perception of high walkability on the NEWS subscales could generally be argued as providing individual benefits for residents. Regardless of whether residents walk in their neighborhood, walkability subscales that describe perceived aesthetics, pedestrian accessibility, crime safety, and traffic safety are likely to be experienced as benefits. Perceived density and diversity are walkability subscales that might be more controversial and may or may not be appreciated by residents. Perceived walkability also relates to more walking in the neighborhood (Owen, Humpel, Leslie, Bauman, & Sallis, 2004; Saelens et al., 2003), which provides a health benefit for residents. The present study is part of a larger multimeasure investigation of the rail stop neighborhood. Our previous research indicated that rail users in the neighborhood experience multiple benefits, including a stronger sense of general neighborhood satisfaction, stronger place attachment, fewer car trips, lower obesity rates (Brown & Werner, 2009), and more objectively rated walkability on their residential blocks (Werner, Brown, & Gallimore, 2010). We also found that rail use was associated with more objectively measured healthy physical activity (Brown & Werner, 2007), especially among nonobese residents (Brown & Werner, 2008). In the present study, we test heretofore unexamined aspects of living in this neighborhood, namely, residents’ perceptions of rail-related neighborhood changes, perceived walkability, perceived barriers to walking, and desired changes that would enable them to walk more. We surveyed residents regarding a variety of neighborhood benefits and challenges before and after the addition of new neighborhood light rail stop along the TRAX line to ask the following: What substantial changes were anticipated (Time 1) and experienced (Time 2), and did these vary by rail ridership experience? Were individuals’ expectations (Time 1) later confirmed (Time 2)? What are the post-TRAX walkability characteristics of this evolving neighborhood? What barriers to walking remained and what changes would residents desire for greater walkability? at UNIV OF FLORIDA Smathers Libraries on March 3, 2014 eab.sagepub.com Downloaded from

PAGE 6

Brown and Werner þ 793Method Rail Station Construction and Neighborhood ContextThe new rail stop, part of a city revitalization initiative, began construction in April 2005. Residents were surveyed in summer 2005 (Time 1) and again the next summer 2006 (Time 2). The rail stop opened in September of 2005 on an existing rail line and had no dedicated parking spaces, which should encourage nearby residents to walk to the station. At Time 1, the nearest rail stop was on average 743 meters from home; at Time 2, the new stop brought service closer, on average 344 meters from home. In total, 90% of residents surveyed in Time 1 were aware that a TRAX stop was being constructed in their neighborhood, although construction was confined to the block of the actual station, given that the line already existed. The neighborhood had a legacy of both industrial and suburban influences from zoning, transportation, housing, and economic policies. The remnants of these policies were still manifested in the form of big box stores, industrial warehouses, vacant lots, and convenience stores designed for cars. The area also had a small park and a mix of singleand multifamily dwellings. Two subsidized apartment complexes had been constructed just prior to the study. Consistent with the city designation of the neighborhood as a revitalization area, residents in the target neighborhood had lower incomes than city averages (US$24,000 vs. US$43,367 household incomes inflation, adjusted to 2005).Sampling and ProceduresLetters alerting residents to the study were sent to all addresses within a half mile of the planned stop (obtained from the city, a crisscross directory, and a commercial telephone survey firm). All residential addresses were either phoned or visited door to door for recruitment (many residents did not own phones). Eligible households included adults who spoke Spanish or English and could walk more than a few blocks. Participants signed consent forms prior to participation and received US$20.00 for completing each phase of the study. Both English and Spanish language surveys were available. Of the 496 potentially eligible addresses, someone answered the door 43.35% (n 215) of the time. Among those invited, 102 were recruited at Time 1 (1 participant dropped out with health problems). At Time 2, 51 residents completed the survey, 38 had moved, 10 refused, and 1 was ineligible due to new health problems (additional recruitment details in Brown & Werner, 2007). at UNIV OF FLORIDA Smathers Libraries on March 3, 2014 eab.sagepub.com Downloaded from

PAGE 7

794 þ En vironment and Behavior 43(6) The participants who stayed in the sample did not differ demographically from those who dropped out, except that the longitudinal sample had marginally more employed residents% vs. 51%; F (1, 97) 2.99, p .087. The sample gender, ethnicity, and home ownership were comparable with neighborhood census statistics (for further details see Brown & Werner, 2007). In addition, there were no significant differences between the longitudinal and the Time 1 only participants for expected consequences from the new TRAX stop or perceived neighborhood walkability, according to t tests (all p s .05). All remaining analyses focus on the longitudinal sample, which is 47% female, 79% White, and 16% Hispanic, with a mean age of 41 (SD 13.82 years).MeasuresAlong with sociodemographic variables noted earlier, this study focused on expectations of and experiences with neighborhood changes related to light rail as well as perceived walkability, walking barriers, and walkability supports. Expectations and experiences with light rail. Nine questions addressed how the new stop was expected to and then was perceived to affect the neighborhood. At Time 1, the question asked, “Will the new TRAX stop change” the nine qualities?, and at Time 2, the question asked, “Has the new TRAX stop changed” the same nine qualities? Many residents took advantage of the “don’t know” option when they were unsure about the future or past consequences of the new TRAX stop. When comparing Time 1 expectations with Time 2 experiences, only residents who provided responses at both time were included in mean calculations (excluding those who said “don’t know” at either Time 1 or Time 2). Responses were on a (decrease a lot) to 2 (increase a lot; 0 no change) scale. The nine neighborhood qualities (and the sample size for each repeated measure calculation) include the following: child or pedestrian safety (n 38), car traffic (n 41), crime rates (n 29), sense of community (n 41), neighborhood reputation (n 42), property taxes (n 20), housing improvements (n 34), economic opportunities (n 37), and housing costs (n 31). At Time 2 only, we asked about the following postconstruction changes: neighborhood parking difficulties, noise levels, and the number of neighbors seen in the neighborhood. NEWS. We selected NEWS perceived walkability subscales that measured perceived housing density, land-use diversity, access to services, aesthetics, traffic safety, crime safety, and neighborhood satisfaction, following scoring conventions available at www.drjamessallis.sdsu.edu/ NEWSscoring.pdf. The neighborhood satisfaction NEWS score includes at UNIV OF FLORIDA Smathers Libraries on March 3, 2014 eab.sagepub.com Downloaded from

PAGE 8

Brown and Werner þ 795 17 items measuring satisfaction with a variety of specific neighborhood qualities, including school quality, commute time, neighborhood traffic, and neighborhood stores; thus, it measures many specific features that were not previously assessed with our more global three-item assessment of general satisfaction (Brown & Werner, 2009). We omitted NEWS subscales regarding street form and sidewalks, given that the new stop would not alter street forms or sidewalks. Past research with the NEWS (Saelens et al., 2003) has demonstrated moderate to high test–retest reliabilities (most items 0.75). Ridership groups. Previous analyses in this neighborhood indicated that long-term TRAX users were more favorable toward transit and TOD (Brown & Werner, 2009), and neighborhoods may become more satisfying if changes fit residents’ preferences (Levine & Frank, 2007), so we included TRAX use categories in the present analyses. Although the average distance from TRAX to participants’ residences was 743 meters at Time 1, almost half the participants were continuing riders who reported riding TRAX in the previous 2 weeks at both Time 1 and Time 2 (45.8%, n 22; six had no car access). New riders were the 22.9% (n 11) who reported riding TRAX only at Time 2, when the new, closer stop opened. A final group we classified as nonriders (31.3%, n 15; one had no car access, two had ridden TRAX at Time 1 but not Time 2 but were similar to and classified with nonriders). Three participants did not answer the TRAX ridership question.Results Are Significant Changes Expected (Time 1) and Subsequently Experienced (Time 2)?This question focuses on the magnitude of changes expected and experienced by the group of residents, compared with no change (i.e., the length of the bars in Figure 1). One sample t tests assessed whether Time 1 expectations and, separately, Time 2 experiences were significantly different from the 0 value of “no change.” As shown in Figure 1, residents anticipated (Time 1) and later perceived (Time 2) that the biggest changes due to the TRAX stop involved economic conditions. They anticipated and later perceived increases in housing costs (Time 1: one sample t [41] 6.86, p .000, d 1.06; Time 2: t [30] 6.02, p .000, d 1.08), housing improvement investments (Time 1: t[41] 4.29, p .000, d 0.66; Time 2: t[33] 5.26, p .000, d 0.90), and property at UNIV OF FLORIDA Smathers Libraries on March 3, 2014 eab.sagepub.com Downloaded from

PAGE 9

796 þ En vironment and Behavior 43(6) taxes (Time 1: t[38] 3.89, p .000, d 0.62; Time 2: t[19] 5.14, p .000, d 1.15) along with economic opportunities in the neighborhood (Time 1: t[42] 6.44, p .000, d 0.98; Time 2: t[36] 4.50, p .000. d 0.74). There were more modest but significant anticipations and perceptions that the stop will/did increase neighborhood reputation (Time 1: t [42] 3.33, p .002, d 0.51; Time 2: t[41] 5.27, p .000, d 0.81) and sense of community (Time 1: t [44] 3.04, p .004, d 0.45; Time 2: t [41] 4.05, p .000, d 0.63). Neither crime nor car traffic was expected or perceived to change much, all ts nonsignificant. At Time 1, child and pedestrian safety were expected to remain the same, but at Time 2, residents felt that child and pedestrian safety had improved, t(37) 3.52, p .001, d 0.57. Other perceived changes were not asked at Time 1 but were at Time 2, using the same to 2 scale. Residents report that, after the TRAX stop opened, they saw more neighbors (M 0.57, SD 0.80; one sample t [46] 4.92, p .000, d 0.72), had more difficulties parking in the neighborhood (M 0.30, SD 0.80; t[42] 2.47, p .018, d 0.38), and reported more neighborhood noise (M 0.26, SD 0.73; t[41] 2.31, p .026, d 0.36). 0.00 0.50 1.00 1.50 Housing costs Economic opportunities Housing improvements Property taxes Neighborhood reputation Sense of community Crime Car traffic Child/pedestrian safety Time 2 Time 1 Figure 1. Resident expectations (Time 1) and experiences (Time 2) on a (decrease) to 2 (increase) scale at UNIV OF FLORIDA Smathers Libraries on March 3, 2014 eab.sagepub.com Downloaded from

PAGE 10

Brown and Werner þ 797Do Expectations (Time 1) Differ From Later Experience (Time 2) Among Ridership Groups?The previous analysis asked whether participants expected neighborhood changes (Time 1) and whether they experienced changes (Time 2), each time tested relative to “zero,” or no change. The present analysis uses a withinparticipants’ design to ask whether each participant’s opinions changed between Times 1 and 2 and whether these differ for the three ridership groups. A set of 3 2 ridership group by time-repeated measures general linear models (GLMs) tested the ratings for each of the nine expectations measured before and after the new light rail stop, controlling for income and employment differences between ridership groups. With one exception, differences across ridership groups were not significant. The one significant group main effect showed continuing riders (adjusted mean 0.81), and new riders (adjusted mean 0.62) believed the rail stop improved neighborhood reputation whereas nonriders did not (adjusted mean 0.07), F(2, 31) 4.04, p .03, partial 2 .21. There were three differences between Times 1 and 2, all of them more favorable with the TRAX stop in place. At Time 2, residents reported higher than expected sense of community (adjusted means: Time 1 0.34, Time 2 0.48), F(1, 32) 5.81, p .02, partial 2 .15, and pedestrian and child traffic safety (adjusted means: Time 1 .13, Time 2 0.40), F(1, 29) 19.30, p .001, partial 2 .15, but lower than expected perception of crime (.30 vs. .02), F (1, 21) 4.31, p .05, partial 2 .17. Group by time interactions were not significant, indicating that all three ridership groups held these opinions. Perceived walkability and satisfaction. The availability of survey data from highand low-walkability neighborhoods in San Diego and Adelaide provided comparative case studies for understanding perceived walkability in the Salt Lake rail stop neighborhood. Following their procedures (Leslie et al., 2005; Saelens et al., 2003), unadjusted means for seven perceived walkability subscales are provided in Table 11. A GLM compared changes in perceived walkability over time in Salt Lake City (comparing columns 2 and 3 of Table 1); a preliminary test confirmed that there were no significant ridership group differences on these measures. Overall, Salt Lake City residents’ perceptions changed significantly over time, F(6, 43) 2.79, p .02, partial 2 .28. Follow-up univariate tests of NEWS subscales revealed that, after the TRAX stop installation, residents perceived greater land use diversity and more satisfaction with the neighborhood. We note that although Bonferroni tests were not used in the comparison studies using the NEWS scores, if adopted in this study, the effect for diversity would be marginally significant at UNIV OF FLORIDA Smathers Libraries on March 3, 2014 eab.sagepub.com Downloaded from

PAGE 11

798 þ En vironment and Behavior 43(6) at the p .007 standard for seven tests; satisfaction effects would still be significant. A series of one-sample t tests allowed testing for differences between Salt Lake (using Time 2 NEWS scores in column 3 of Table 1) and the means provided for highly walkable neighborhoods in previous research, that is, the high-walkable neighborhoods in San Diego (column 5 of Table 1) and Adelaide (column 7 of Table 1). The most striking differences show that the Salt Lake neighborhood had higher perceived density (San Diego: t [50] 4.14, p .000, d 0.58; Adelaide density not available) and lower land-use diversity (San Diego: t [50] .17, p .000, d 0.72; Adelaide: t [50] .02 p .000, d 1.39) than the walkable neighborhoods elsewhere. Perceived crime safety in Salt Lake was also lower than for the walkable neighborhoods elsewhere (San Diego: t [50] .51, p .015, d 0.35; Adelaide: t [50] .51, p .015, d 0.35). Table 1. NEWS Scores of Perceived Walkability Before and After Rail Stop Construction and in Comparison Lowand High-Walkable Neighborhoods Salt Lake San Diego Adelaide Perceived quality (scoring range) Time 1Time 2 Low walkability High walkability Low walkability High walkability Density (5-375) 226.8233.5194.4203.2*NANA Land-use diversity (1-5) þ 2.7 3.0* þ 2.8 þ 3.5* 3.44.0** Access to services (1-4) þ 3.2 3.2 þ 2.8 þ 3.2* 2.93.6** Aesthetics (1-4) þ 2.8 2.7 þ 2.8 þ 3.0* 3.12.7** Traffic safety (1-4) þ 2.8 2.8 þ 2.7 þ 3.1* 2.42.5 Crime safety (1-4) þ 2.9 2.9 þ 3.1 þ 3.1 3.03.1 Neighborhood satisfaction (1-5) þ 3.5 3.8**NANANANANote: NEWS Neighborhood Environment Walkability Scale. Significance levels indicate significance tests within each of the three studies; see original articles for details on the San Diego and Adelaide analyses. *p .05. **p .001. at UNIV OF FLORIDA Smathers Libraries on March 3, 2014 eab.sagepub.com Downloaded from

PAGE 12

Brown and Werner þ 799 Other perceived walkability features do not consistently differentiate Salt Lake from high-walkability neighborhoods in San Diego or Adelaide. The Salt Lake neighborhood was just as accessible as the San Diego neighborhood (p ns) but less accessible than the walkable Adelaide neighborhood, t (50) .53, p .000, d 0.90. The Salt Lake neighborhood was judged less aesthetically pleasing than the walkable San Diego neighborhood, t (50) .71, p .001, d 0.52, but similar to the Adelaide neighborhood (p ns). The Salt Lake neighborhood had lower traffic safety than the San Diego neighborhood, t (50) .68, p .000, d 0.66, but higher traffic safety than the Adelaide neighborhood, t (50) 4.36, p .000, d 0.62. In sum, some of the requirements of walkability were perceived to be clearly present (e.g., density) but others were lacking (i.e., diversity and crime safety) and still others showed Salt Lake to be similar to at least one of the two comparison high-walkability neighborhoods. Walking barriers and supports. Residents also reported on perceived walking barriers and supports (Figure 2). Resident perceptions of why they did not walk more (walking barriers) showed that having “no time to walk” was their most common reason, cited by 58%. In contrast, the least frequently cited barriers were crime (16%) and traffic (11%). This result may be surprising given that the area contains a mix of industrial areas, private market housing, and subsidized public housing at the edge of the city, where traffic is generally higher than in the suburbs. The most frequently noted modifiable environmental feature of the neighborhood that made walking difficult was the absence of nearby walkable destinations, reported by 31% of residents. 0.11 0.16 0.2 0.23 0.24 0.3 0.31 0.38 0.58 0 Feel unsafe due to traffic Feel unsafe due to crime Other walk barrier Extreme weather Health reasons Get enough exercise elsewhere Not enough destinations No interest/don’t think about it No time/busy 0.10.20.30.40.50.60.7 Figure 2. Proportion of residents reporting barriers that keep them from walking more in the neighborhood at UNIV OF FLORIDA Smathers Libraries on March 3, 2014 eab.sagepub.com Downloaded from

PAGE 13

800 þ En vironment and Behavior 43(6) When asked what changes in the neighborhood would encourage them to walk more (walkability supports), most residents wanted more stores (Figure 3), which is consistent with the relatively low scores on NEWS-scored perceptions of diversity of destinations and the 31% reporting that few walkable destinations existed in the neighborhood, as noted earlier. A majority of residents also wanted more parks and trails as well as trees. Over a third of residents wanted cleaner streets. Despite saying that they would not avoid walking because of traffic or crime (from walking barrier reports and above), about one fourth of residents said they would walk more with more/better sidewalks, night lighting, crosswalks, and police enforcement. Thus, although most residents said time constraints keep them from walking more, a number of modifiable environmental features could be targeted to enhance walking conditions.DiscussionAfter the rail stop was constructed in this older, revitalizing area that is transitioning into a TOD, residents’ preand postconstruction reports indicated that that they had accurately predicted most of the changes they subsequently experienced. Residents anticipated and then reported increases in neighborhood economic values (e.g., taxes and housing costs), suggesting that the city is achieving their revitalization goals. Residents also expected and then later perceived enhancements of neighborhood reputation and sense of community. 0.24 0.24 0.27 0.28 0.37 0.53 0.57 0.6 0 Better police enforcement Better or more crosswalks Better night lights Better or more sidewalks Cleaner streets More trees More parks and trails More stores 0.10.20.30.40.50.60.7 Figure 3. Proportion of residents identifying changes that would encourage them to walk more at UNIV OF FLORIDA Smathers Libraries on March 3, 2014 eab.sagepub.com Downloaded from

PAGE 14

Brown and Werner þ 801 Furthermore, the postconstruction surprises were pleasant ones: more sense of community, less crime, and more child and pedestrian safety than antici pated. One resident’s survey explained, “I think pedestrian safety has increased because professional TRAX drivers are more cautious than some drivers in the neighborhood.” Thus, residents report tangible economic investments in the neighborhood as well as some social and safety benefits. Less favorably, they reported more neighborhood noise and parking difficulties. Overall, however, residents appear to perceive more personal benefits associated with the new TRAX station than did Londoners who received an underground transit station (Clark et al., 2003). Given that personal benefits of newly popular neighborhoods near transit are underresearched, this study suggests that residents and society have a “win-win” situation with light rail neighborhoods. Residents benefit from perceiving positive consequences of the new stop, found in the current study, as well as health benefits associated with walking to the stop, found earlier (Brown & Werner, 2007, 2008, 2009). Results suggest that the city did several things right in the zoning, designing, and phasing of this TOD development. The city invested in building subsidized apartments very close to the rail stops prior to rail stop construction. This provided residents with satisfying housing prior to housing cost increases that were perceived by residents in this study and that are part of more general upward economic pressures that often accompany rail stop construction (Cervero et al., 2004). This strategy may also encourage secondary benefits to poor residents, such as reduced need to purchase private automobiles and easier access to jobs. Observation of the rail stop area since data collection ended suggests that the subsidized housing did not deter market rate housing investments either. The stop area has newly constructed livework housing, where living and studio or shop spaces are combined, which may attract “creative-class” residents who are seen as the key to the economic and cultural health of many U.S. cities (Arrington & Cervero, 2008). Finally, the new stop was deliberately designed without dedicated parking spaces, and parking requirements were reduced for the subsidized housing. Many cities attempting to accommodate rail transit fail to reduce TOD area parking requirements, which drives up land and construction costs, reduces walkability, and is believed to make rail use less appealing, convenient, and safe. These benefits need to be weighed against resident complaints that, after TRAX service started, parking difficulties and neighborhood noise were both significantly higher than a neutral point of “no change.” Residents’ desires for improved neighborhood conditions are feasible but will require joint efforts between the city and residents and can be aided by the type of postconstruction surveys we provided. Simple things like cleaner streets at UNIV OF FLORIDA Smathers Libraries on March 3, 2014 eab.sagepub.com Downloaded from

PAGE 15

802 þ En vironment and Behavior 43(6) have been shown to be of great importance to residents. For example, less litter, better lighting, and less crime were the top three preferences among 24 different options for neighborhoods among a random sample of residents interviewed in New Orleans after the destruction caused by hurricane Katrina (Hong & Farley, 2008). Similarly, litter has been shown to predict greater crime vulnerability in a neighborhood a few miles away from the one in this study (Brown, Perkins, & Brown, 2004). Additional green spaces and trees also have been associated with a wide variety of benefits for residents (Kuo, 2003) and are a key amenity in making higher density living options attractive (Cervero & Bosselmann, 1998). As this area transitions to more of a transit-oriented neighborhood, the greater intensity of land use associated with transit area development may achieve residents’ desires for more attractive and diverse destinations. Governmental action will be needed to provide more green spaces; an initial plan for a bike path and walking trail may provide part of the green amenities that residents desire. Although resident perceptions of land-use diversity (as indexed by the NEWS) increased over time, residents perceived less land-use diversity than residents reported in walkable neighborhoods in San Diego and Adelaide. Similarly, when asked what would encourage them to walk more in the neighborhood, residents reported that more stores was their top choice. When asked about specific walking destinations (in additional unreported data), many residents said they walked to a neighborhood 7-11 convenience store and to a Walmart superstore just outside the neighborhood (about 1 mile from the new rail stop). Neither of these options is the type of pedestrian-friendly destination envisioned by TOD advocates; as the rail stop area continues to grow, new pedestrian-oriented destinations would be welcomed by residents. Because commercial development often lags housing development, we suspect residents of new rail stop neighborhoods elsewhere may share these complaints. Cities may want to seek out ways to assure sufficient numbers of early rail stop businesses to encourage walking habits among residents and to prevent them from feeling compelled to buy cars to attain needed goods. This study also has limitations worth noting. We had a small sample size of 51 compared with the (between-participants) sample sizes of 87 by Leslie et al. (2005) and 109 by Saelens et al. (2003), although our within-participants design requires fewer participants for adequate power. The small sample size also makes our tests of differences conservative. Although we presented demographic comparisons showing participants were fairly similar to nonparticipants, we cannot rule out the possibility of a biased sample. Our analyses showed that those who stayed in the study were comparable on most variables with those who dropped out; specifically, they had similar scores on most demographic variables and on all the expectations and perceived walkability at UNIV OF FLORIDA Smathers Libraries on March 3, 2014 eab.sagepub.com Downloaded from

PAGE 16

Brown and Werner þ 803 subscales. Similarly, the longitudinal sample had similar demographic charac teristics to the entire neighborhood (Brown & Werner, 2007). The Time 1 survey did not inquire about walking barriers and supports so we do not know whether those who dropped out of the sample differed in any substantial way from our remaining sample on those perceptions. Those who stayed in the sample were marginally more likely to be employed (p .087), so we are more confident of the findings for employed residents. In addition, the sample was poorer than the city average and their experiences may not generalize to experiences of all residents who receive light rail service in their neighborhood. As the United States and other countries seek more sustainable energy resources, many advocate designing neighborhoods to make cars less necessary. Light rail stop neighborhoods will be one focus for new research and design, given the growth of light rail usage in the United States; rides on rail increased from 251 million in 1995 to 381 million in 2005 (American Public Transportation Association, 2009). Even though objective measures of physical activity intensities (Troiano et al., 2007) reveal that few adults in the United States achieve the center for disease control and prevention’s recommended amount and level of physical activity (30 min of moderate activity per day on most days of the week in 10-min bouts), Salt Lake residents of this neighborhood who rode rail did accrue more healthy bouts of activity using TRAX (Brown & Werner, 2007, 2008). Although these benefits are promising, future research should address the benefits and disadvantages of rail stop neighborhoods in varied situations across the country. In sum, this evaluation of a new light rail stop suggests that fears experienced in some neighborhoods transitioning to more intensive use did not materialize (cf. Downs, 2005). Residents largely anticipated the light rail service-related changes to their neighborhood. The fact that residents expected and experienced most strongly the increase in economic opportunities and val ues in their neighborhood suggests that it may be especially important to pre serve affordable housing prior to allowing rail stop construction, lest the area’s new popularity begin to exert steep price increases. This type of development—city-owned land with multiple sources of financing supporting subsidized housing near rail stops—may provide a good quality affordable housing to residents while also providing environmental benefits (e.g., less pollution, car use, and transportation service demand) to the region. AcknowledgmentThe authors appreciate the anonymous reviewers and the data collection assistance from Stephanie Nalbone, Jonathan Gallimore, Melissa Napier, Elisa Hamblin, Chad Killpack, Bekah Larson, and Edward Cusack. at UNIV OF FLORIDA Smathers Libraries on March 3, 2014 eab.sagepub.com Downloaded from

PAGE 17

804 þ En vironment and Behavior 43(6) Authors’ NoteAny opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the view of the National Science Foundation. This article was accepted by Robert Bechtel, Editor.Declaration of Conflicting InterestsThe authors declared that they had no conflicts of interests with respect to their authorship or the publication of this article.FundingResearch was funded by the University of Utah’s Institute of Public and International Affairs and the University Research Committee; this research was also supported in part by the Research Experience for Undergraduates Program, the National Science Foundation, under Grant ATM 0215768.Note1. þ A preliminary test confirmed that there were no significant ridership group dif ferences on these measures.ReferencesAmerican Public Transportation Association. (2009). Unlinked passenger trips by mode. Washington, DC: Author. Arrington, G. B., & Cervero, R. B. (2008). TCRP 128: Effects of TOD on housing, parking and travel. Washington, DC: Transportation Research Board. Baldassare, M. (1981). The effects of a modern rapid-transit system on nearby residents: A case study of BART in the San Francisco area. In I. Altman, J. F. Wohlwill, & P. B. Everett (Eds.), Transportation behavior (pp. 203-238). New York, NY: Plenum. Belzer, D., & Autler, G. (2002). Transit-oriented development: Moving from rhetoric to reality. Washington, DC: Brookings Institution Center on Urban and Metropolitan Policy. Briggs, X. D. S., Darden, J. T., & Aidala, A. (1999). In the wake of desegregation: Early impacts of scattered-site public housing on neighborhoods in Yonkers, New York. Journal of the American Planning Association, 65, 27. Brown, B. B., Perkins, D. D., & Brown, G. (2004). Crime, new housing, and housing incivilities in a first-ring suburb: Multilevel relationships across time. Housing Policy Debate, 15, 301-345. Brown, B. B., & Werner, C. M. (2007). A new rail stop: Tracking moderate physical activity bouts and ridership. American Journal of Preventive Medicine, 33, 306-309. at UNIV OF FLORIDA Smathers Libraries on March 3, 2014 eab.sagepub.com Downloaded from

PAGE 18

Brown and Werner þ 805Brown, B. B., & Werner, C. M. (2008). Using accelerometer feedback to identify walking destinations, activity overestimates, and stealth exercise in obese and nonobese individuals. Journal of Physical Activity and Health, 5, 882-893. Brown, B. B., & Werner, C. M. (2009). Before and after a new light rail stop: Resident attitudes, travel behavior, and obesity. Journal of the American Planning Association, 75, 5-12. Calthorpe, P. (1993). Next American metropolis: Ecology, community, and the American dream. Princeton, NJ: Princeton University Press. Cervero, R. (1994). Rail transit and joint development: Land market impacts in Washington, DC and Atlanta. Journal of the American Planning Association, 60, 83-94. Cervero, R., & Bosselmann, P. (1998). Transit villages: Assessing the market potential through visual simulation. Journal of Architectural and Planning Research, 15, 181-196. Cervero, R., Murphy, S., Ferrell, C., Goguts, N., Tsai, Y., Arrington, G. B., . . . McKay, S. (2004). Transit-oriented development in the United States: Experiences, challenges, and prospects (Transit Cooperative Research Program Report No. 102). Washington, DC: Transportation Research Board. Clark, C., Gatersleben, B., Uzzell, D., & Reeve, A. (2003). Perception study comparison report (Working Paper No. 44). London, UK: University of Westminster. Dorn, J. (2004). Hidden in plain sight: Capturing the demand for housing near transit.Washington, DC: Federal Transit Administration. Downs, A. (2005). Smart growth: Why we discuss it more than we do it. Journal of the American Planning Association, 71, 367-378. Envision Utah. (2002). Wasatch front: Transit oriented development guidelines. Salt Lake City, UT: Author. Handy, S. (2005). Smart growth and the transportation: Land use connection: What does the research tell us? International Regional Science Review, 28, 146-167. Hess, D. B., & Lombardi, P. A. (2004). Policy support for and barriers to transitoriented development in the inner city: Literature review. Transportation Research Record, 1887, 26-33. Hong, T., & Farley, T. A. (2008). Urban residents’ priorities for neighborhood features. American Journal of Preventive Medicine, 34, 353-356. IPEN. (2009). International physical activity and the environment network. Retrieved from http://www.ipenproject.org Kuo, F. E. (2003). The role of arboriculture in a healthy social ecology. Journal of Arboriculture, 29, 148-155. Leslie, E., Saelens, B., Frank, L., Owen, N., Bauman, A., Coffee, N., & Hugo, G. (2005). Residents’ perceptions of walkability attributes in objectively different neighbourhoods: A pilot study. Health & Place, 11, 227-236. at UNIV OF FLORIDA Smathers Libraries on March 3, 2014 eab.sagepub.com Downloaded from

PAGE 19

806 þ En vironment and Behavior 43(6)Levine, J., & Frank, L. D. (2007). Transportation and land-use preferences and residents’ neighborhood choices: The sufficiency of compact development in the Atlanta region. Transportation, 34, 255-274. Lund, H., Cervero, R., & Willson, R. W. (2004). Travel characteristics of transitoriented development in California. Sacramento, CA: Caltrans. Owen, N., Humpel, N., Leslie, E., Bauman, A., & Sallis, J. F. (2004). Understanding environmental influences on walking: Review and research agenda. American Journal of Preventive Medicine, 27, 67-76. Patton, J. W. (2007). A pedestrian world: Competing rationalities and the calculation of transportation change. Environment and Planning A, 39, 928-944. Ridlington, E., & Heavner, B. (2005). Transit-oriented development: Strategies to promote vibrant communities. Baltimore, MD: MaryPIRB Foundation. Rose, D. (2004). Discourses and experiences of social mix in gentrifying neighbourhoods: A Montreal case study. Canadian Journal of Urban Research, 13, 278-316. Saelens, B. E., Sallis, J. F., Black, J. B., & Chen, D. (2003). Neighborhood-based differences in physical activity: An environmental scale evaluation. American Journal of Public Health, 93, 1552-1558. Schlossberg, M., & Brown, N. (2004). Comparing transit-oriented development sites by walkability indicators. Transportation Research Record, 1887, 34-42. Shapiro, R. J., Hassett, K. A., & Arnold, F. S. (2002). Conserving energy and preserving the environment: The role of public transportation. Washington, DC: American Public Transportation Association. Southworth, M., & Ben-Joseph, E. (1995). Street standards and the shaping of suburbia. Journal of the American Planning Association, 61, 65-81. Troiano, R. P., Berrigan, D., Dodd, K. W., Masse, L. C., Tilert, T., & McDowell, M. (2007). Physical activity in the United States measured by accelerometer. Medicine & Science in Sports & Exercise, 40, 181-188. Werner, C. M., Brown, B. B., & Gallimore, J. (2010). Light rail use is more likely on “walkable” blocks: Further support for using micro-level environmental audit measures. Journal of Environmental Psychology, 30, 206-214. doi: 10.1016/ j.jenvp.2009.11.003BiosBarbara B. Brown is an environmental psychologist who studies human behaviors related to neighborhood and housing design, including new urbanism and neighborhood revitalization, place attachment and neighborhood satisfaction, and physical activity, transit use, and obesity. Carol M. Werner is a social/environmental psychologist who studies psychological processes underlying environmentally sustainable behaviors, such as transit use, recycling, and toxics reduction. at UNIV OF FLORIDA Smathers Libraries on March 3, 2014 eab.sagepub.com Downloaded from



PAGE 1

THE ANTICIPATORY EFFECTS OF COMMUTER RAIL ON ECONOMIC DEVELOPMENT IN ORANGE COUNTY, FL By BENJAMIN LYTLE A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FO R THE DEGREE OF MASTER OF ARTS IN URBAN AND REGIONAL PLANNING UNIVERSITY OF FLORIDA 2014

PAGE 2

© 2014 Benjamin Lytle

PAGE 3

3 ACKNOWLEDGMENTS I would like to thank my committee for their guidance throughout my graduate studies. Specifically I would like to thank Dr. Ruth Steiner for introducing me to new ways to look at issues I thought I understood, Dr. Paul Zwick for introducing me to the joys of ArcGIS ModelBuilder, and Dr. Abhinav Alakshendra for assigning the project that lead to this t hesis. I would also like to thank the professionals who assisted with the collection of data. In particular, I would like to thank Sara Forelle for giving me the names of the best people to c ontact for building permit data : Dean Salmons, Elizabeth Dang, Sa ra Blanchard, and Keith Gerhartz who kindly arranged for me to have access to the data necessary for this project and answered questions regarding the permit records. I am also grateful to anyone within the planning offices who may have supported those lis ted above with the creation or packaging of this data. M y parents and grandparents deserve thanks for their love and support. Without them, I don't know if I would have been able to complete this program. I am also grateful for the advice and conversation of David Wasserman and Marcos Bastian. David showed me that it was possible to graduate on time and Marcos convinced me that if I didn't it would take seven years . F inally, I would like to thank my fiancée Tracey for putting up with me over the last two ye ars. Her patience and love have probably been more than I deserve.

PAGE 4

4 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 3 LIST OF TABLES ................................ ................................ ................................ ............ 6 LIST OF FIGURES ................................ ................................ ................................ .......... 7 LIST OF ABBREVIATIONS ................................ ................................ ............................. 9 ABSTRACT ................................ ................................ ................................ ................... 10 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 12 SunRail and Orange County ................................ ................................ ................... 12 Research Questions ................................ ................................ ............................... 15 Organization ................................ ................................ ................................ ........... 16 2 LITERATURE REVIEW ................................ ................................ .......................... 17 Theoretical Impact of Transit ................................ ................................ .................. 17 Case Studies ................................ ................................ ................................ .......... 1 8 3 METHOLOGY ................................ ................................ ................................ ......... 26 Data Availability ................................ ................................ ................................ ...... 26 Parcel Data ................................ ................................ ................................ ....... 26 Creating proximit y indicators ................................ ................................ ...... 27 Creating noise indicators ................................ ................................ ............ 29 Creating land use indicators ................................ ................................ ....... 30 Creating control group indicators ................................ ............................... 30 Permit Data ................................ ................................ ................................ ...... 31 Missing Data ................................ ................................ ................................ ..... 32 A nalysis Procedure ................................ ................................ ................................ . 33 Parcel Data Analysis ................................ ................................ ........................ 33 Permit Data Analysis ................................ ................................ ........................ 34 4 R ESULTS ................................ ................................ ................................ ............... 36 Overall Land Use Change ................................ ................................ ....................... 36 Land Value Change: All Land Use Types ................................ ............................... 37 Two Mile Buffer ................................ ................................ ................................ 37 One Mile Buffer ................................ ................................ ................................ 38

PAGE 5

5 Other Distance Measures ................................ ................................ ................. 39 Lan d Value Change: Residential Land Use ................................ ............................ 41 Two Mile Buffer ................................ ................................ ................................ 41 One Mile Buffer ................................ ................................ ................................ 42 Half Mile Buffer ................................ ................................ ................................ . 42 Quarter Mile Buffer ................................ ................................ ........................... 43 Linear Distance ................................ ................................ ................................ 44 Land V alue Change: Commercial Land Use ................................ ........................... 44 Linear Distance ................................ ................................ ................................ 44 Half Mile Buffer ................................ ................................ ................................ . 45 Quarter Mile Buffer ................................ ................................ ........................... 46 Land Value Change: Institutional Land Use ................................ ............................ 46 Land Value Change: Industrial Land Use ................................ ............................... 46 Building Permits by Jurisdiction ................................ ................................ .............. 47 Maitland ................................ ................................ ................................ ............ 48 Winter Park ................................ ................................ ................................ ....... 50 Orlando ................................ ................................ ................................ ............. 51 Orange County ................................ ................................ ................................ . 53 Building Permits by Station ................................ ................................ ..................... 55 Downtown (Church Street Station) ................................ ................................ ... 55 Urban Center (Florida Hospital Health Village Station) ................................ .... 57 Village Center (Maitland Station) ................................ ................................ ...... 60 Neighborhood Center (Meadow Woods Station) ................................ .............. 63 5 DISCUSSION ................................ ................................ ................................ ......... 65 6 CONCLUSION ................................ ................................ ................................ ........ 68 Further Research ................................ ................................ ................................ .... 69 Limitations ................................ ................................ ................................ ............... 70 APPENDIX A PARCEL DA TA ................................ ................................ ................................ ....... 71 B JURISDICTION LEVEL PERMIT DATA ................................ ................................ . 85 C STATION LEVEL DATA ................................ ................................ ......................... 91 LIST OF REFERENCES ................................ ................................ ............................. 102 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 107

PAGE 6

6 LIST OF TABLES Table page 4 1 Land use changes wi thin a half mile of stations compared to total parcels in Orange County. ................................ ................................ ................................ .. 36 4 2 Permits issued within a half mile of SunRail stations compared to all permits issued by the City of Maitland. ................................ ................................ ............ 49 4 3 Permits issued within a half mile of SunRail stations compared to all permits issued by the City of Winter Park. ................................ ................................ ....... 51 4 4 Permits issu ed within a half mile of SunRail stations compared to all permits issued by the City of Orlando. ................................ ................................ ............. 53 4 5 Permits issued within a half mile of SunRail stations compared to all permits issued by Ora nge County. ................................ ................................ .................. 55 4 6 Permits issued within a half mile of Church Street station compared to all permits issued from 2007 to 2013. ................................ ................................ ..... 57 4 7 Permits issued within a half mile of Florida Hospital Health Village station compared to all permits issued from 2007 to 2013. ................................ ............ 60 4 8 Permits issued within a half mile of Maitland station compar ed to all permits issued from 2007 to 2013. ................................ ................................ .................. 62 4 9 Permits issued within a half mile of Meadow Woods station compared to all permits issued from 2010 to 2013. ................................ ................................ ..... 64 A 1 Land Use Code Classifications. ................................ ................................ .......... 71 A 2 Summary of all Phase 1 regression models. ................................ ...................... 75 A 3 Summary of all Phase 2 regression models. ................................ ...................... 80 A 4 Orange County parcel mean just value, all land uses ................................ ......... 84 A 5 Land use types as a percent of total parc els within a half mile of stations. ......... 84 B 1 Summary of building permit data by jurisdiction. ................................ ................ 85 C 1 Total SunRail boardings by station for the week of May 19, 2014. ..................... 91 C 2 Summary of building permit data by station. ................................ ....................... 92

PAGE 7

7 LIST OF FIGURES Figure page 4 1 Two mile coefficient model, all land use types. ................................ ................... 38 4 2 One mile coefficient model, all land use types. ................................ ................... 38 4 3 Half mile coefficient model, all land use types. ................................ ................... 39 4 4 Orange County mean parcel just value. ................................ ............................. 40 4 5 Quarter mile coeffi cient model, all land use types. Note that the 2008 coefficient was not statistically significant, so the trend line connects 2007 and 2009. ................................ ................................ ................................ ........... 41 4 6 Two mile coefficient model, residential land types. ................................ ............. 42 4 7 One mile coefficient model, residential land types. ................................ ............. 42 4 8 Half mile coefficient model, residential land types. ................................ ............. 43 4 9 Quarter mile coefficient model, residential land types. ................................ ....... 44 4 10 Linear distance coefficient model, commercial land types. ................................ . 45 4 11 Half mile coefficient model, commercial land types. ................................ ........... 45 4 12 Quarter mile coefficient model, commercial land types. ................................ ..... 46 4 13 Linear distance coefficient model, industrial land types. ................................ ..... 47 4 14 Building permits issued by City of Maitland over study period. ........................... 48 4 15 Percent of total building permit value issued by City of Maitland for each distance measure over study period. ................................ ................................ .. 49 4 16 Building permits issued by City of Winter Park over study period. ...................... 50 4 17 Percent of total building permit value issued by City of Winter Park for each distance measure over study period. ................................ ................................ .. 51 4 18 Building permits issued by City of Orlando over study period. ............................ 52 4 19 Percent of total building permit value issued by City of Orlando f or each distance measure over study period. ................................ ................................ .. 53 4 20 Building permits issued by Orange County from 2010 to 2013. .......................... 54

PAGE 8

8 4 21 Building p ermits issued near Church Street station over study period. ............... 56 4 22 Building permit value distribution issued near Church Street station over study period. ................................ ................................ ................................ ....... 56 4 23 Building permits issued near Florida Hospital Health Village station over study period. ................................ ................................ ................................ ....... 59 4 24 Building permit value distribution issued near Florida Hospital Health Village station over study period. ................................ ................................ ................... 59 4 25 Building permits issued near Maitland station over study period. ....................... 61 4 26 Buildin g permit value distribution issued near Florida Hospital Health Village station over study period. ................................ ................................ ................... 62 4 27 Building permits issued near Meadow Woods station over study period. ........... 63

PAGE 9

9 LIST OF ABBREVIATIONS BART Bay Area Rapid Transit. BART is a much researched heavy rail transit system in the San Francisco Bay Area. BRT Bus Rapid Transit. BRT is a type of premium bus transportation system usually characterized by high frequencies, dedicated right of way, traffic signal priority, and other efficiency improvements over conventional bus service. FDOT Florida Department of Transportation. The state agency responsible for owning and operating regionally significant tra nsportation facilities in Florida, including SunRail. FGDL Florida Geographic Data Library. The FGDL was a source for some of the base map GIS data and for the 2012 parcel just value data. GIS Geographic Information System. GIS data consists of databases linked to geographic information that can be analyzed and processed. MARTA Metropolitan Atlanta Rapid Transit Authority. MARTA is a heavy rail rapid transit system in the Atlanta region. TOD Transit Oriented Development. TOD is a concept in land use pla nning to refer to high density mixed use development located adjacent to transit with an urban design intended to encourage transit ridership and walkability. USA Urban Service Area. The USA includes the areas in the county where urban development is plan ned and where utilities are usually provided.

PAGE 10

10 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Arts in Urban and Regional Planning THE ANTICIPATORY EFFECTS OF COMMUTER RAIL ON ECONOMIC DEVELOPMENT IN ORANGE COUNTY, FL By Benjamin Lytle August 2014 Chair: Ruth Steiner Cochair: Paul Zwick Major: U rban and R egional P lanning This study quantifies the anticipatory effects of the SunRai l commuter rail line on property values, building permits and land use in Orange County, FL from 2007 to 2013 . Parcel data for each year starting in 2007 (when the interlocal agreements necessary for SunRail to become a reality were signed) was analyzed to determine the effect of proximity to SunRail on property values using a hedonic pricing model with linear distance, quarter, half, one and two mile buffer coefficients . Additionally, building permit data was assembled for each jurisdiction with a station to determine the share and total value of residential and nonresidential building permits near each station. Building permit data was investigated in greater depth for one station within each TOD typology (as described in Olore, 2011) present in Orange Cou nty . The researcher found that the anticipatory effects of SunRail we re predominantly related to increasing the value of residential properties located closer to stations. The share of building permits and building permit value located near stations increa sed between 2007 and 2013, with an increasing shares of permits near non downtown stations. Finally , the areas near SunRail have maintained their original mix of land uses except for a reduction in

PAGE 11

11 institutional land uses. Overall this suggests that SunRai l did not substantially change the mix of land uses near stations, but that residential properties near stations increased in value . Downtown and Urban Center stations had smaller gains in the share of building permits near stations than Village Center sta tions, because those have greater potential for growth.

PAGE 12

12 CHAPTER 1 INTRODUCTION As funding for transportation projects has become more scarce due to reductions in per vehicle revenue from the gas tax, t ransportation projects are routinely justified on the grounds that they are catalysts for economic development. Often, such projects take years to plan and construct. During that time, it stands to reason that the real estate market would adjust based on the expectation of the new infrastructure's presence an d generate much of that change before the infrastructure becomes available . This is called the anticipatory effect, and it was first investigated within the context of the Washington DC Metro, where it was found to have a significant effect on retail and r esidential property values (Damm et al., 1980) . Using Orange County parcel data from the Orange County Property Appraiser's Office and building permit data from each jurisdiction with a SunRail station, the researcher analyze d the anticipatory effect of Su nRail on Orange County, FL. SunRail and Orange County Premium transit systems such as commuter rail, light rail and bus rapid transit have been proposed for most larger cities in the southeast United States in recent years. Unlike older cities in the north east or Midwest, most of these cities became large metropolitan areas after the decline of the historic streetcar systems and the rise of the automobile after World War II. Because of this, these cities grew predominately when it was assumed that transport ation would be provided only by automobile or bus. Because of this, these cities often lack the smaller scale, pedestrian oriented areas around transit stations in older cities. Therefore , the viability of such systems are often questioned by

PAGE 13

13 citizens and the public alike on the grounds that once a passenger gets off the train, they would have no way of getting to their destination without a car. Orange County, FL is the largest county in the Orlando metro area with an estimated population of over 1.2 milli on people (US Census Bureau, 2014). Before SunRail opened, it was one of the largest metro areas in the United States wit hout some form of rail transit. SunRail was not the first regional premium transit system proposed in Orlando. In the late 1990s, a lig ht rail system along much of the same alignment as SunRail was studied and federal money was received, however it was cancelled when the Orange County Commission voted not to support the project in 1999. Studies on the feasibility of such a project continu ed throughout the early 2000s, but they seemed certain to face strong political opposition (Krueger, 2001). In July 2007, Orange, Osceola, Volusia, and Seminole counties and the City of Orlando entered into an interlocal agreement with each other and FDOT to create and provide funding for the Central Florida Commuter Rail (SunRail) system (USDOT, FTA & FDOT, 2008). Phase 1 of this system, identical to what opened in May 2014, was scheduled to open in 2010, followed by a Phase 2 opening in 2013 (Hamburg & Pi no, 2007). This system was conceived as a way to offer alternative transportation modes as a way to avoid congestion along the Interstate 4 corridor, to allow for an increase in the densities and to better serve high density residential and employment area s as called for in the local comprehensive plans, and as a way to prevent congestion that stifles economic growth ( USDOT, FTA & FDOT, 2008).

PAGE 14

14 Following this agreement, FDOT began negotiating with CSX to purchase the right of way for the train. Negotiations stalled in 2008 when the two parties could not come to an agreement about the distribution of liability (Tracy, 2009). Because the tracks were to be owned by the FDOT but CSX would have the right to operate freight trains, concerns over who would be liable in the event of an accident involving CSX prevented the state from approving the purchase of the right of way (2009). Eventually, an agreement was reached between CSX and the FDOT where SunRail would be held liable for CSX's equipment in case of an accide nt on SunRail's tracks, regardless of who is at fault. This agreement was approved by the Florida legislature when it was bundled together with additional funding for S outh Florida's Tri Rail commuter rail system. As soon as an agreement was reached betwee n the FDOT and CSX, Amtrak contested the deal for the same liability concerns as CSX by lodging a complaint with the US Surface Transportation Board, which would prevent SunRail from receiving the $300 million in federal funding set aside for the project. In December 2010, Amtrak agreed to end its opposition to SunRail in a meeting brokered by the US Departm ent of Transportation Secretary (Tracy, 2010). With the sale of the tracks finally approved in late 2010, the projected opening date had been moved to s ummer 2013 ( Tracy & Deslatte , 2011a). In January 2011, newly elected Governor Rick Scott froze all state contracts worth over $1 million for review by the newly created Office of Fiscal Accountability, including four critical SunRail contracts ( Tracy & Des latte , 2011a). After a six month review, the project was given permission to proceed, and the opening date of s pring 2014 was established (Tracy & Deslatte, 2011b).

PAGE 15

15 On May 1st, 2014, the first phase of SunRail running north south through the Orlando metro politan area open ed to passengers. This commuter rail system runs approximately parallel to the busy Interstate 4 corridor along the former CSX A line right of way. When the second phase is complete in 2017, i t will be approximately 61 miles long and have 17 stations (Tracy, 2013). Currently, the system is 31 miles long with 12 stations and is expected to be carrying about 4,300 passengers per day by the end of the year (Tracy, 2014). Each Phase 1 station in Orange County has between 666 and 3707 boardings per week (Table C 1). Research Questions In addition to providing additional capacity along this busy north south corridor, SunRail was hoped to spur economic development along the line, but a recent report in the Orlando Sentinel suggest ed that developmen t near stations has been slow (Shankin, 2013). The purpose of this study is to determine the extent to which this has actually been the case. Because Orange County has a car dependent growth pattern typical of other cities in the southeast, and because the SunRail stations in Orange County range from suburban park and ride type stations to the stations in the heart of downtown Orlando, this county was seen as a useful case study that may apply to other places with similar growth patterns considering commute r rail. The purpose of this study was to determine the impact of proximity to SunRail on property values, building permit rates, permitted project values, and land use changes from 2007 to 2013. If there wa s an impact on these economic development indicato rs, at what distance from the station was it significant, and did it affect some land uses more than others?

PAGE 16

16 Organization This research is presented in six chapters. Chapter 2 includes a literature review that gives an overview of the current state of know ledge on the relationship between transit access and property values, the viability of transit oriented design, and the anticipatory effect of transit in different contexts. Chapter 3 is a description of the data collected for this study and the methodolog y used to analyze the data and answer the research question. Chapter 4 is a discussion of the results, including the land use changes on property close to SunRail, the impact of SunRail proximity on property values and the changes in the number of building permits issued and their values at a jurisdictional and at an individual station level. Chapter 5 discusses some of the results within the context of the history of the project. Chapter 6 discusses the results and synthesizes some of the findings that cou ld be applied to other transit projects in an similar cities. Additionally this chapter discusses changes and improvements to the methodology that can be used for further research.

PAGE 17

17 CHAPTER 2 LITERATURE REVIEW Theoretical Impact of Transit Prior to the 19 60s, economists and policy analysts understood the distribution of property values to be largely a function of history. As described by Alonso (1964), this idea was that as a city grows, the wealthier citizens desire newer, bigger houses. But because land is scarce in the city, they have to move to the periphery to be able to assemble enough land to build their desired houses. Their previous houses are sold to less wealthy people, and because those houses are older the less wealthy people can afford them ( A lonso, 1964) . Alonso added the idea that households also value open space and lower densities in addition to proximity to the center. His major contribution to this question was the idea that proximity is an inferior good, meaning that if given the choice at a given price between proximity and lower density, the wealthy will choose lower density. An implication of this is that if a new transportation facilities such as highways are built that make it easier to get to the center from a given location, the pr operty value would raise because the property now has both proximity and space ( Alonso, 1964). This framework is the theoretical basis for the idea that transportation infrastructure improvements increase property values where accessibility is improved (Da mm et al. 1980) . The economic theory behind this idea was developed long after it was commonly understood to be true. In 1930, E. H. Spengler published his conclusions on the effects of the rail transit opening in New York City in the early 1900s: (1) New transit lines tend to shift value rather than to create increased aggregate value. While owners of land in the vicinity of a new transit line may benefit, owners of land elsewhere may be disadvantaged. (2) Transit lines are only one of the numerous factors

PAGE 18

18 influencing land values, and they often cannot outweigh the effects of other factors which are acting to depress land values. (3) Transit acts to enhance land values in centres of concentration at the ex pense of outlying areas. (4) Areas already developed do not generally show a marked increase in land value when new transit lines are opened. (5) In areas already supplied with a number of transit lines, addition of another one will have only a mild stimulative effect compared with the effect it would have in an area not already supplied with transit. (6) In newly developing areas with transit service, increased land values are likely to be attributable in large part to the process of subdivision rather than to transit access (as cited in Damm et al., 1980, p. 317). Although the connection between property values and transit facilities has long been taken as a given by many politicians and decision makers, t he literature is decisively less conclusive. To best understand the actual effect of transit on property values, the researcher consulted a variety of case studies on premium transit facilities like commuter rail, heavy rail, light rai l, and bus rapid tran sit (BRT). The majority of the literature focused on the heavy rail systems built in the United States after World War II or light rail systems built in the last 30 years. Heavy rail systems included the Washington Metro, the Metro politan Atlanta Rapid Tra nsit Authority (MARTA) rapid transit system , the Bay Area Rapid Transit (BART) system in San Francisco and a handful of international examples selected because of the frequency of the literature from which their methodology was found. Light rail systems in cluded studies on systems in automobile oriented cities such as Los Angeles, Charlotte, Dallas and others. Very few studies on commuter rail were reviewed because of their rarity in the literature. Case Studies The first study on the anticipatory effects o f transit was conducted by Damm et al. (1980) on the Washington Metro. This study adapted methodologies originally developed to describe the effects of freeways on property values. This study replaced

PAGE 19

19 the highway proximity variables with proximity to trans it stations for parcel and real estate transaction data within the District of Columbia between 1969 and 1978. A hedonic pricing model was used to identify the impact of the Washington Metro system on property values before the system opened . Like several later studies, commercial and retail properties had the strongest increase in property values at a close proximity to stations while residential property values increased modestly ( Damm et al., 1980). The stronger effect of rail infrastructure on commerci al property was not found by a study by Cervero and Landis (1993). Their quasi experimental research methodology matched sites in Atlanta and Washington DC and found that office rents were slightly higher in areas served by rail transit than those that wer e not. However, the magnitude of the effect was not great enough to say that owners of buildings near rail transit were able to capture a monetary benefit from their proximity to rail transit ( Cervero & Landis, 1993). A large scale study by Landis et al. ( 1994) was conducted on five fixed rail transit systems in California to provide a consistent methodology that allows a comparison of the effects of each system. The five systems studied included the BART heavy rail system, the CalTrain commuter rail system , and three light rail systems in Sacramento, San Diego, and San Jose. Homes located close to BART stations sold at a $2.29 premium for every meter closer they were to the station ( Landis et al., 1994, p. 21) , however there was not a statistically signific ant increase in property values nea r the other transit system s . Landis et al. speculate that this difference was related to the lower frequency of service and a relative lack of parking capacity near the other four

PAGE 20

20 systems in suburban areas , thereby limiti ng their potential ridership outside of walking distance ( Landis et al., 1994). The positive impact of additional parking was also found in a study on the MARTA rapid transit system in Atlanta by Bowes and Ihlanfeldt (2001). This study also used a hedonic price model in addition to models of neighborhood crime and retail activity to determine the impact of rail transit stations on property values. This analysis found that in lower income areas, property values were negatively correlated with proximity. Howe ver, in higher income areas, there was a premium for residential units between 1 and 3 miles away, suggesting that most gains from proximity to rail transit were more closely connected to park and ride than to pedestrian access ( Bowes and Ihlanfeldt, 2001) . However, the impact of transit access was generally found to be less than that of freeway access within the first 20 years of San Francisco's BART system (Cervero & Landis, 1997). Additionally, this study found that the effect of transit access was limit ed significantly by public policy. In many station areas, greater density growth was prevented because of successful opposition from local residents. However, BART was found to have a positive economic development impact on the traditional downtown areas d ue to supportive public policy and a relative lack of opposition to densification ( Cervero & Landis, 1997). Internationally, transit systems have been shown to have potentially massive impacts on property values. A study by Cervero and Kang (2011) on land use changes and property values in Seoul, Korea found that land values increased up to 10% for

PAGE 21

21 residential land uses within 300 meters (984 feet) and 25% for commercial land uses within 150 meters (492 feet) of bus rapid transit (BRT) stations ( Cervero & K ang, 2011). Property around Line 5 of the Seoul subway (Bae, Jun & Park, 2003) found that proximity to stations increased property values before the subway opened, but property values flattened or declined upon the opening of the line. This suggested that the majority of the effects of property values from transit improvements were realized before the system opens and that potentially the market corrected itself by lowering slightly after the system opened ( Bae, Jun & Park, 2003). Although Seoul has few si milarities to a Sunbelt , auto oriented metro politan area like Orlando, this study was relevant be cause it showed an extreme case of the potential for property value change caused by transit in a place with significant latent demand and strong levels of gro wth (two factors frequently identified as being necessary for changes in property values) . A study of the anticipatory effects of light rail in Sheffield , England ( Henneberry, 199 8 ) utilized hedonic models and found a negative anticipatory effect followed by a lack of statistical significance two years after the system opened . This was likely to have been caused by the nuisance effects of the construction of the system. Once the system opened, property values may have not risen due to a relatively low growt h level and potentially a lack of latent demand ( Henneberry, 199 8 ). Additionally, Cervero and Kang (2011) note d that if demand for the system and growth exist ed , theoretically the land value premium for proximity to transit was the result of an increase in the accessibility of a parcel. This accessibility was given by an increase in level of service . Therefore, the technology that provides the increase in level

PAGE 22

22 of service ( commuter rail, heavy rail, light rail, BRT ) is not as important as the actual increas e in level of service ( Cervero & Kang, 2011). This might at first appear to be at odds with the findings of the meta analysis by Debrezion, Pels & Rietveld (2007), which found that commuter rail has a consistently higher impact on property values than metr o rail systems, but th is could instead be a result of differences in level of service and potentially the provision of parking at stations as described by Landis et al. (1994) . Finally, Cervero and Kang (2011) conclude that land use regulation needs to all ow an increase in density, or else such increases in property value are unlikely to take place. The City of Orlando and other jurisdictions along SunRail have provisions in their comprehensive plans to allow for TOD and other higher density developments ne ar stations (City of Orlando, 2012, Orange County Community, Environmental & Development Services Planning Division, 2012, City of Maitland, 2010). A literature review on the effects of transit on land use and travel mode by Badoe & Miller (2000) found tha t there is a variation in the literature between no effect and some effect. The differences in the studies outcomes depends on whether or not the researchers used an integrated model that takes into account the different actors and interactions driving dev elopment. However, studies including other accessibility variables like highways reduces the reported impact of railway proximity. In general, the meta analysis revealed that the impact of railway stations is different for separate land use categories. Res idential properties at a greater distance from stations are influenced than commercial properties. However, commercial properties are often reported to have greater effects at close proximity ( Debrezion, Pels & Rietveld, 2007). These studies

PAGE 23

23 were conducted using hedonic pricing models which are a common statistical regression technique that allows one to determine the value of an individual component of a value derived from multiple attributes ( Debrezion, Pels & Rietveld, 2007). To understand the mechanics of the economic development near transit stations, the researcher briefly reviewed studies including alternative data sources such as surveys and building permit data. A study by Loukaitou Sideris & Banerjee (2000) explore d the difficulties of creating eco nomic development along new transit corridors like the Los Angeles Blue Line light rail system through inner city and industrial areas. Like many recent transit projects, t his system alignment was opportunistically selected to save time and money on real e state acquisition, but little thought was given to the utility the line would have to the communities along its length. In addition to the low income single family housing present along much of the line, significant portions are surrounded by heavy industr ial uses while each end of the line has commercial, light industrial and some mixed use development ( Loukaitou Sideris & Banerjee, 2000). Ten years after the Blue Line opened, much of the corridor along the line was in the same state as before: large areas lacking any type of destination or amenity, poverty and underinvestment. An analysis of building permit data show ed that in all but one year, the areas around Blue Line stations had proportionally less investment than the cities they are in, and that the station areas did not generally participate in the economic upswing in the mid 1990s ( Loukaitou Sideris & Banerjee, 2000). The authors conclude that The Blue Line had not succeeded in its goal of economic development in the first ten years of its operation . The primary reasons of this failure were the location of the route along the backs of buildings and through

PAGE 24

24 nondescript industrial areas (the "back door problem"), low population densities near stations, difficulty in accessing stations as a result of th e distance from dense areas combined with almost no feeder bus service to stations or park and rides at stations, poor urban design, land cost, regulatory barriers, lack of institutional commitment, and lack of community participation in the planning proce ss ( Loukaitou Sideris & Banerjee, 2000). Another analysis by Loukaitou Sideris (2010) looks at the anticipatory effect of transit in its environment directly. The author reviewed the area around the Metro Gold Line in Los Angeles with building permit data to identify the changes in the years leading up to its opening. These changes in the land use and ownership escalated in the years immediately prior to the line's opening ( Loukaitou Sideris, 2010). Because of better urban design, fewer regulatory barriers to transit oriented development (TOD) and better institutional commitment, the economic development and ridership goals were reached by the Gold Line ( Loukaitou Sideris, 2010). Overall, this literature review revealed that SunRail could have a positive imp act on property values and building permit activity near stations . Theoretically, Alonso (1964) suggests that transit like SunRail can have an effect on property values if it improves the accessibility of property near the stations. But property in Orange County already generally has good accessibility because of the existing and expansive roadway and expressway network. Therefore, the low service frequency and relatively small number of destinations (compared to existing automobile accessibility) would sug gest a relatively modest premium. This premium is likely to be different for different land use types. Commercial properties often exhibit large premiums at short distances while

PAGE 25

25 residential properties exhibit smaller premiums at much larger distances (Deb rezion, Pels & Rietveld, 2007). Transit through built out corridors of an industrial nature like the Los Angeles Blue Line may fail to generate any significant changes because of the location of the line and barriers to changing land use (Loukaitou Sideris & Banerjee, 2000). But where land use regulation changes along with the introduction of the transit service in the case of the Los Angeles Gold Line (Loukaitou Sideris, 2010) and BART in traditional downtown areas (Cervero & Landis, 1997) it is possible f or economic development to occur .

PAGE 26

26 CHAPTER 3 METHOLOGY Data Availability To understand the anticipatory effects of SunRail, the researcher utilized a methodology based on the Loukaitou Sideris Gold Line study (2010). The impacts of transportation facili ties on land use and economic development were understood through a combination of factors including information about the use and value of land throughout the study period, information about construction and renovation on those properties. The indicators identified by Loukaitou Sideris included the following: P opulation Density Race Population Average Age Proportion Foreign born Poverty Rate Educational Achievement Land Use Type: Commercial, Institutional, Residential Percent change in land use type Parcel sale rates Parcel Value Built Square Footage Building Permit Issuance Building Permit Value The following section describes the data publically available and the data made available to the researcher for this project. It also describes the data created b y the researcher for the use in this project and the ways in which the data was processed. Parcel Data GIS Parcel data from 2007 to 2013 from the Orange County Property Appraiser's Office was provided by the GIS Division of the Orange County Government. Th is data contained polygons representing the property boundaries of all the parcels in Orange County along with data for each parcel. Information about the physical characteristics of

PAGE 27

27 the parcel such as the size of the parcel in acres, the combined living a rea of any structures on the property in square feet, and the Department of Revenue land use code (DOR Code) were included with this data . Additionally, this data included th e just v alue of each parcel for every year except 2012 and 2013 . The just v alue is the assessed value of the parcel and all structures contained on it before January 1st of that year. The assessed value represents the property appraiser's best estimate of the probable sale price of the property if the property was sold on the open marke t with adequate sale time and buyers and sellers behaving in a rational, self interested manner, free of duress (Value Adjustment Board, n.d.). Just value data from 2012 was collected from the statewide parcel data available on the Florida Geographic Data Library (FGDL) (Panda Consulting, 2012) and joined to the parcel data provided by Orange County . Just value data for 2013 was calculated based on two fields included in the data : just value change and previous year just value . The two fields were added tog ether for each parcel to determine the 2013 just value. Creating p roximity i ndicators To identify the relationship between SunRail stations and property values, the researcher first had to create GIS data including the location of SunRail. This was accompl ished by creating a new line feature within ArcGIS 10.1 along the center of the CSX/SunRail right of way parcel. Referencing satellite imagery in Google Maps, and station design information fro m the SunRail Corporate Website (Project Documents, 2013), the researcher created point data representing the center of each station along

PAGE 28

28 the line within Orange County. Additionally, each grade crossing was documented following a similar methodology to allow the creation of a noise indicator. Once the location of eac h station was identified, the researcher created two models within ArcGIS ModelBuilder to automatically generate the proximity indicators. Each proximity indicator was created based on straight line distance because of the relative simplicity of the proces s compared to network distances and because there is some evidence suggesting that the differences between the outcomes of the two approaches are insignificant from a ridership perspective (Guerra, Cervero & Tischler , 201 1 ). This model generated a raster d ataset of the Euclidian distance from the center of each station at 25 foot intervals, generating a surface showing the straight line distance from any point within Orange County to the nearest SunRail station. Such linear distance based proximity indicato rs were used by several studies on the impacts of transportation facilities on property values ( Baum Snow & K a h n, 2000; Billings, 2011; Celik & Yankaya, 2006 ; Cervero & Kang, 201 1 ; Damm et al., 1980; Grimes & Young, 2010; Henneberry, 1997 ; Hanneberry, 199 8 ; McMillen & McDonald, 2004 ) . The parcels were converted to points (the p oints were created at the centroid of each parcel) and the value of the raster at each point was added to the point parcel dataset . In addition to determining the linear distance from each parcel to a SunRail station, the researcher created a model to identify which parcels fall within quarter, half, one and two mile buffers around stations. Many studies have found significant impacts on commercial land uses within a quarter of a mile of transit stations ( Bowes & Ihlanfeldt, 2001 ; Cervero & Duncan, 2002; Guerra, Cervero & Tischler, 2011 ; Petheram

PAGE 29

29 et al., 201 3 ; Weinstein & Clower, 2002 ) and impacts on residential land uses for half a mile from stations ( Bowes & Ihlanfeldt, 2001 ; Knaap et al., 2001 ; Loukaitou Sideris, 2010 ; Petheram et al., 2013 ) . Because SunRail is a commuter rail system with park and ride facilities at several stations, longer buffers of one and two miles were created to account for the greater distances riders arriving by car may have to their homes like several other studies ( Billings, 2011 ; Bowes & Ihlanfeldt, 2001 ; Garrett, 2004 ; Knaap et al., 2001 ; Petheram et al., 2013 ) . Creating noise i ndicators The increase in accessibility to a parcel located close to a commuter rail station may be offset by the nuisance created by the noise of trains traveling along the tracks and sounding their horns at grade crossings. A study by Bellinger (2006) found that residential properties lo se an estimated $48,000 for each 10dB increase in horn noise over 50dB. Research on the effects of rail noise on property values generally use similar methodologies as those identifying the impacts of airport or highway noise on property values. Generally, these studies calculate the noise level above a given threshold at a given parcel as an input to a hedonic pricing model. The coefficient within the model is called a Noise Depreciation Sensitivity Index (NDSI) (Brons, et. al, 2003). A similar methodology was used here. The linear distance from the S unRail track and grade crossings were input into separate Euclidian distance rasters. The distance to the nearest grade crossing was used to calculate the horn noise in dB using the following variation of the sound pressure level equation (Equation 3 1). (3 1)

PAGE 30

30 where is the estimated noise level at each parcel at a distance of from the source of the horn noise; and is the known higher end estimated sound level of a train horn of 1 50dB when measured at , a distance of 100 feet ( FRA, n.d.). Equation 3 1 was used to model the noise generated by the sound of the train, with an of 95dB as the known noise level of a locomotive at a distance of 100 feet (FRA, n.d.). Once th e noise level of the train noise and horn noise were calculated, they were converted to a scale from zero to one, with zero representing 50dB or less and one representing the highest possible noise level from a train of 150 dB. Train noise less than 50 dB was disregarded because it was below the average background noise level where train noise does not affect property values (Bellinger, 2006). Creating land use i ndicators The parcel data included a Department of Revenue land use code (DOR Code) tha t include d 188 different land use classifications within Orange County. The DOR Code was used to create a dummy variable identifying Residential, Commercial, Institutiona l and Industrial land uses . Table A 1 includes th e complete classification table . Creating cont rol g roup i ndicators Orange County has a wide range of land use intensities, from wild swamplands to relatively dense urban districts. Therefore, trends in property values would be expected to vary based on the location of the parcel. This diversity of con texts was the primary reason for choosing Orange County for this study. However, this variety of land uses could potentially interfere with the effectiveness of the hedonic modeling. To avoid

PAGE 31

31 this problem, properties outside of the Urban Service Area (USA) were excluded from this analysis. Orange County discourages development outside the USA by restricting the services provided by the county. Because of this, t he majority of new development wa s contained within the USA, so it was used as a good indicator t hat a parcel could potentially become developed and would therefore be subject to the market. The USA GIS shape file provided by the Orange County Planning Division was used to generate a raster indicating whether a given point was inside or outside the US A. Parcels located within the USA were given a dummy value of one and were included in the parcel data analysis. Permit Data Permit data was collected from the cities of Orlando, Maitland and Winter Park including all building permits issued from 2007 thro ugh 2013 containing the land use type (residential or nonresidential), the address or location of the permitted project, the project type, and the value of the project. Additional permit data from unincorporated Orange County was provided from 2010 to 2013 , however the Orange County data did not include the approximate value of each project. Combined, this data covers the building permits approved within the jurisdictions of all eight Orange County SunRail stations. Data from Maitland , Orlando and Orange Co unty were geocoded using the addresses contained within the data. The address locator identified the actual location of 83% of the permits from Orlando, 96% of the permits from Maitland, and (according to the variables left in the geocoded data provided by Orange County) 87% of the permits from Orange County . Data from Winter Park included the latitude and longitude of each permit parcel.

PAGE 32

32 Finally, the distance, noise and control group indicators created for the parcel data were generate d for the permit data . Missing Data E mployment data was not considered for this study because of the difficulty of collecting the data and inconsistencies within the commonly used Info USA employment data . Likewise, p arcel sale rates were absent from this analysis due to the d ifficult y of consistently identifying sales rates from the parcel data. Because this study covers t he years 2007 to 2013, no useful source of demographic data exists to analyze trends . Census 2000 was seven years before the study, so it would not make a fa ir baseline for the purposes of this study. The 2010 census occurred in the middle of the sample, so it is useful for neither the before or after sample. Likewise, American Community Survey dat a from 2009, 2010 and 2011 fall neatly into the middle of the time period being analyzed. Additionally, the margins of error are relatively high for many of the factors called for by this study. Therefore, demographic data such as total population, race, average age, proportion foreign born, poverty rate, educational attainment, and household vehicle availability were not analyzed in this study. In regard to the parcel data, t he researcher hoped to utilize additional building characteristics common to other hedonic models in the literature such as number of bedrooms a nd bathrooms . However, only 2012 and 2013 property parcel data included those variables. Therefore, these variables were not included in this study.

PAGE 33

33 Analysis Procedure Parcel Data Analysis This study identifies the changes in just value based on f ive diffe rent measures of proximity for each year 2007 to 2013 using property parcel data from the Orange County Property Appraiser's Office . The impact of these proximity measures was calculated for residential, commercial, industrial, institutional and all land u ses. T he researcher created a linear regression for each combination of distance variables and land use types for each year using SPSS Statistics 22 . For the buffer based distance models, the regression was run using just the Phase One Orange County SunRai l stations (every station in Orange County opened with Phase 1 except for the Meadow Woods station on the south side of the county) and using all Orange County stations. For all land use types, the regression was based on Equation 3 2. (3 2) Where: Acres is the size of the parcel in acres; LivingArea is the interior square footage of all buildings on the parcel; Residential is a dummy variable for whether the parcel has a residential use; Institutional is a dummy variable for whether the parcel has an institutional use; Commercial is a dummy variable f or whether the parcel has a commercial use; Industrial is a dummy variable for whether the parcel has an industrial use; and Distance is either the distance to the nearest SunRail station in feet or is a dummy variable for Half Mile, Quarter Mile , One Mile or Two Mile.

PAGE 34

34 Each distance variable was also used in a model for each individual land use type, Residential, Commercial, Institutional or Industrial. Models were referred to by their distance coefficient type followed by their land use type: (3 3) Where the coefficients are the same as above and land uses are selected before the regression is run. Similar to the analysis technique employed by Henneberry (19 98) and Bae, Jun & Park (2003) each year's regression coefficients were compared to identify trends in the distance coefficients across the analysis period. For each set of regressions, the distance coefficient and t statistic was noted for every year in a ddition to the adjusted R square for the model. The model combinations that made statistically significant distance coefficients for each year were analyzed to determine the changes in the relationship between distance to SunRail and the property values be tween the project announcement in 2007 and 2013. Permit Data Analysis Overall trends in building construction along the SunRail line was determined by calculating the total number and total value of permits issued within the quarter mile, half mile, one m ile and two mile buffers around stations within each jurisdiction. Additionally, the number and value of all permits within the buffers around each individual station was analyzed to determine the changes in the percent of total development work contained within the permit data was near each individual station.

PAGE 35

35 This station level permit data allowed the researcher to identify station neighborhoods that experienced an increase in development and to identify stations that did not. Station design and local are a characteristics were explored as possible explanations for the local development patterns. This quasi experimental case study allowed the researcher to identify t rends at a countywide level for three important characteristics identified in the literature review as possible effects of new transit services. Longitudinal l and use information was identified and simplified into four basic categories using countywide parcel data from 2007 to 2013 to compare changes occurring within half a mile of stations to al l parcels countywide. A retrospective longitudinal methodology utilizing h edonic regression models for each land use to determine the share of the value of parcels attributable to proximity to SunRail stations using five different proximity indicators. Add itionally, p ermit data was used to develop a description of the economic development impacts of SunRail in terms of the number and value of permits. The building permit analysis was conducted at a jurisdiction al level with the permits located outside the p roximity indicators within each city acting as the control group. At a neighborhood level , the process was repeated for one station in each neighborhood type utilizing all building permits collected outside the proximity indicators as the control group . Wh en considered together, these mixed methods describe the anticipatory effects of property values and permitted building activity within Orange County near SunRail stations.

PAGE 36

36 CHAPTER 4 RESULTS Overall Land Use Change As seen in Table 4 1 , the percentage o f residential and industrial parcels within a half mile of stations remain ed relatively constant over the study period. Likewise, the percentage of commercial properties remain ed fairly constant except for a potentially anomalous bump in Orange County as a whole and within half mile of stations in 2012. The percentage of institutional parcels declined from 8.59% in 2007 to 4.57% in 2013 while institutional parcels in Orange County as a whole declined by only 0.9 %. Industrial land uses declined within a half mile of stations in 2012 and 2013, but between 2007 and 2013 they only dropped 0.2%. This shows that the land use has mostly remained stable across the study period across the county without much major changes near stations. Table 4 1. Land use changes w ithin a half mile of stations compared to total parcels in Orange County. Year 2007 2008 2009 2010 2011 2012 2013 All Land Uses Total 358820 366568 368904 370263 373732 439906 431679 Half Mile 5178 5171 5125 5102 5188 8095 8081 Residential Total 89.0 % 88.9% 89.5% 89.7% 89.6% 86.6% 89.5% Half Mile 62.7% 62.6% 63.1% 63.1% 62.7% 63.5% 63.3% Commercial Total 3.6% 3.9% 3.9% 3.9% 3.9% 8.0% 4.8% Half Mile 20.3% 19.4% 20.0% 19.8% 19.8% 25.7% 24.2% Institutional Total 2.3% 1.7% 1.5% 1.5% 1.5% 1.2% 1.4% Half Mile 8.6% 8.3% 6.5% 6.4% 6.3% 3.8% 4.6% Industrial Total 1.2% 1.3% 1.2% 1.2% 1.3% 1.1% 1.1% Half Mile 3.8% 4.2% 4.1% 4.1% 4.1% 3.7% 3.6% Other Total 3.9% 4.2% 3.8% 3.7% 3.8% 3.1% 3.2% Half Mile 4.7% 5.6% 6.4% 6.6% 7.1% 3.3% 4.3%

PAGE 37

37 Lan d Value Change: All Land Use Types Two Mile Buffer To see the complete results of the land value regression models, see Table A 2 and Table A 3 . The model that captures the most significant changes in land value for all land use types use d the Two Mile dis tance metric (with a t statistic between 7.332 and 9.418 for the distance coefficient and an adjusted R square between .606 and .679) . In 2007, properties located within two mile s of SunRail stations we re worth approximately $ 103,726 more than those locate d more than two mile s away. Between 2008 and 2009, the premium for SunRail proximity fell by over $ 2 0,000. Whether this de crease in land values is due to SunRail or complicated changes in land values because of the recession cannot be determined with this methodology. However, this temporary de crease in the two one mile distance coefficient ended in 201 1when the premium had a high of $115,082 before settling back to $109,203 in 2013 . Additionally, when comparing the size of the coefficient for Phase 1 stati ons and all stations , it becomes obvious that property values in parcels located close to the Meadow Woods station (the only Phase 2 station in Orange County) diverged from the values of parcels near Phase 1 stations in 2009. This suggests that the imminen t opening of SunRail Phase 1 began to impact property values in 2009 ( Figure 4 1 ). The overall change in the value of the distance coefficient for properties located within the two mile buffer of SunRail stations between 2007 and 2013 was an increase of $5,477.

PAGE 38

38 Figure 4 1 . Two mile coefficient model, all land use t ypes . One Mile Buffer The models utilizing the One Mile coefficient were also statistically significant (with a t statistic between 4.337 and 7.622 for the distance coefficient and an adjusted R square between .606 and .679) . Between 2007 and 2013, the properties located within one mile of SunRail Phase 1 stations increased in val ue by a n average of $28,232 ( Figure 4 2).Unlike the two mile buffer, the one mile buffer coefficient was consistently greater for Phase 1 stations. However, like the two mile coefficient difference, the Phase 1 one mile coefficient was furthest from the Phase 2 coefficient in 2010 and 2011. Figure 4 2 . One mile coefficient model, all land use types. $60,000 $70,000 $80,000 $90,000 $100,000 $110,000 $120,000 2007 2008 2009 2010 2011 2012 2013 Phase 1 Phase 2 $60,000 $70,000 $80,000 $90,000 $100,000 $110,000 $120,000 2007 2008 2009 2010 2011 2012 2013 Phase 1 Phase 2

PAGE 39

39 Other Distance Measures Th e half mile and coefficient was not statistically significant in 2007 or 2008 (t statistic of 1.855 and 1.790, respectively) while the quarter mile coefficient was not statistically significant in 2008 (t statistic of 1.296). However, both generated signif icant results from 2009 to 2013, suggesting that any increase in property values within these two buffers wa s entirely unrelated to other pre existing factors as shown by the lack of significance in the first two years after SunRail was announced . Fig ure 4 3 . Half mile coefficient model, all land use types. More than the other buffers, the most striking feature of the half mile coefficient model is the peak in 2011. This could mean there was a jump in the demand for property within half a mile of SunR ail stations or that property within half a mile of stations did not decline in value as much as property further away. The latter explanation is more compelling because of general changes in the real estate market t hroughout the study period ( Figure 4 4). If that was all that was affecting this coefficient, the peak should have be en in 2012, not 2011. Regardless of the true cause, the overall change in the coefficient's magnitude of $8,475 wa s much smaller than that of the one mile model. $60,000 $80,000 $100,000 $120,000 $140,000 $160,000 $180,000 2007 2008 2009 2010 2011 2012 2013 Phase 1 Phase 2

PAGE 40

40 Figure 4 4 . O range County mean parcel just value , from Table A 4 . The quarter mile coefficient model show ed a strong upward trend from 2007 to 2013 with a total increase of $133,010 ( Figure 4 5) . However, the lack of statistical significance in 2008 lead the researcher to believe that it may be inappropriate to include 2007, since theoretically the coefficient should be significant for a continuous stretch if the effect was real. Starting in 2009, the increase in the coefficient's value wa s still a substantial $41,669. The linear distance coefficient was only statistically significant in 2011, 2012 and 2013. In 2011 the coefficient was 1.040, suggesting that for every linear foot a parcel's center is further from a SunRail station, the property value decline d by $1.04. In 2013, the coefficient rose to 1.332. A positive coefficient means that property values rise as the distance from a SunRail station increases. This positive distance coefficient was found in all of the other statistically significant linear distance mod els except for the industrial land use model. $220,000 $240,000 $260,000 $280,000 $300,000 $320,000 $340,000 $360,000 2007 2008 2009 2010 2011 2012 2013

PAGE 41

41 Figure 4 5 . Quarter mile coefficient model, all land use types. Note that the 2008 coefficient was not statistically significant, so t he trend line connects 2007 and 2009. Land Value Change: Residential Land Use The regressions for residential land uses found a substantial increase in the premium for proximity to SunRail over the study period in all four buffer models , with larger coefficients in the smaller buffers . This suggest ed that the proximity premium to SunRail ha d stronger effects at closer proximity. Two Mile Buffer The two mile buffer model remain ed relatively flat, with an increase of only $1,086 between 2007 and 2013. More interesting here wa s the reduction in the coefficient for the Phase 2 stati on model from 2009 to 2010 of $11,403. This observation wa s consistent with the divergence in the coefficients for the two mile buffer all land use model. Residential property located close to the Meadow Woods station did not experience an increase in valu e from SunRail until later than property located closer to other stations. Additionally, the property close to this station was located on a golf course that closed in 2007, so the housing market crash was likely to impact this area worse ( Shanklin, 2011). In 2011, the foreclosure rate in this neighborhood was 20% greater than the Orlando area as a whole ( Shanklin, 2011). $60,000 $110,000 $160,000 $210,000 $260,000 $310,000 $360,000 2007 2008 2009 2010 2011 2012 2013 Phase 1 Phase 2

PAGE 42

42 Figure 4 6 . Two mile coefficient model, residential land types. One Mile Buffer The mile coefficient remained relatively flat just und er $70,000 range but increased in 2008 and 2013 for a combined total increase of $1 0,793 ( Figure 4 7 ). Like the two mile buffer model, the difference between phase one and phase two was striking, especially considering how stable the Phase 1 coefficient wa s. Figure 4 7 . One mile coefficient model, residential land types. Half Mile Buffer The half mile coefficient for Phase 1 stations was statistically significant throughout and rose $18,705 . The Phase 2 station model was not statistically significant, pr ior to 2012. This model had much greater variation than the one and two $60,000 $70,000 $80,000 $90,000 $100,000 $110,000 2007 2008 2009 2010 2011 2012 2013 Phase 1 Phase 2 $30,000 $40,000 $50,000 $60,000 $70,000 $80,000 2007 2008 2009 2010 2011 2012 2013 Phase 1 Phase 2

PAGE 43

43 mile models and the t statistics were lower (between 4.190 and 12.100), so these findings have a greater chance of error. Figure 4 8 . Half mile coefficient model, residential land t ypes. Quarter Mile Buffer The 2007 , 2008 and 2009 quarter mile coefficient was not statistically significant at the 95% confidence interval, and between 2010 and 2013, the quarter mile coefficient fell $22,491 . Unlike the Phase 1 model, the Phase 2 model w as statistically significant every year except 2008. In 2007, residential properties within a quarter mile of future SunRail stations were worth $43,042 less than parcels located further away. By 2014, those parcels were worth more than further parcels by $86,765, an increase of $129,807 ( Figure 4 9) . Because of the relatively small number of residential parcels within a quarter mile of stations, this change could probably be attributed to a relatively small number of big new residential projects as opposed to slow changes at existing properties. $20,000 $25,000 $30,000 $35,000 $40,000 $45,000 $50,000 2007 2008 2009 2010 2011 2012 2013 Phase 1 Phase 2

PAGE 44

44 Figure 4 9 . Quarter mile coefficient model, residential land types. Linear Distance The linear distance coefficient was also statistically significant, with a $0. 91 increase in property values for each additional foot a property is from a SunRail station in 2007 to $ 1.24 increase in value f or the same metric in 2013 (T able A 2 ). This would suggest that proximity premiums to SunRail are declining overall. However, SunRail travels along the most densely populated co rridor in the region, and the distance variable r o se continuously as the distance from stations increases. So this change in the variable could also be interpreted to mean that the residential property values in the suburbs recovered in that time period. L and Value Change: Commercial Land Use Linear Distance All five distance coefficients were statistically insignificant for at least one year . However, the linear distance coefficient was statistically significant for every year except 201 1 . It declined from 2007 to 2012, bef ore rising sharply in 2013 ( Figure 4 10) . This could mean that SunRail was making commercial property close to stations more valuable from 2007 to 2012. However, there is no clear explanation for the change in 2013. Like the residential linear distance coefficient, the research er believes this says $50,000 $0 $50,000 $100,000 $150,000 $200,000 2007 2008 2009 2010 2011 2012 2013 Phase 1 Phase 2

PAGE 45

45 more about the performance of property values in the suburbs than it says about areas near SunRail. Figure 4 10 . Linear distance coefficient model, commercial land types. Half Mile Buffer Th e half mile buffer model wa s remarkable in that the Phase 1 and Phase 2 station datasets we re so similar for commercial land use. However, firm conclusions c ould not be drawn from this model because the 2013 data was statistically insignificant. If SunRail was an important driver of the change in commercial property values, the effect would get stronger as it g o t closer to the opening of the system. Figure 4 1 1 . Half mile coefficient model, commercial land types. $0 $2 $4 $6 $8 $10 $12 2007 2008 2009 2010 2011 2012 2013 Phase 2 $240,000 $290,000 $340,000 $390,000 $440,000 $490,000 $540,000 2007 2008 2009 2010 2011 2012 2013 Phase 1 Phase 2

PAGE 46

46 Quarter Mile Buffer The quarter mile coef ficient suggest ed an even stronger negative impact on commercial property values than the half mile model. Like that model, however, this model wa s not statistically significant for 2013, t herefore the likelihood that it accurately reflect ed the anticipato ry effect of SunRail is low. Figure 4 1 2 . Quarter mile coefficient model, commercial land types . Land Value Change: Institutional Land Use None of the five models found statistically significant coefficients for institutional land uses for more than one year of the study period . This was not particularly surprising because the percentage of institutional land uses near SunRail stations declined over the study period ( Table A 5) , suggesting that there was little additional institutional development. Land Value Change: Industrial Land Use The only model with any statistical significance for industrial land uses wa s the linear distance model. In 2007, for every foot an industrial parcel is further from a SunRail station the property value declined by $ 6.09 . In 2013, that decline dropped to $ 3.96 ( Figure 4 12) . $240,000 $340,000 $440,000 $540,000 $640,000 $740,000 2007 2008 2009 2010 2011 2012 2013 Phase 1 Phase 2

PAGE 47

47 Like the other linear distance models, this change was either a function of the suburban industrial land increasing in value or the industrial land near SunRail stations declining in value. This could b e a reasonable interpretation, because the zoning changes around stations ha d increased the number of potential neighbors who might not appreciate having industrial neighbors. This increase in nuisance liability could theoretically reduce the value of indu strial land. Additionally, because frequent passenger trains now run along the corridor, it may be more difficult to schedule freight deliveries by rail to the properties along the tracks. However, none of the buffer based models had any sort of statistica l significance, so it is far more likely that this result was a result of industrial land use changes in the suburbs. Figure 4 1 3 . Linear distance coefficient model, industrial land types . Building Permits b y Jurisdiction Overall trends in building construction along the SunRail line w ere determined by calculating the total number and total value of permits issued within the quarter mile, half mile, one m ile and two mile buffers around stations . These were compared to the total number and value o f permits within each jurisdiction. $8 $7 $6 $5 $4 $3 $2 $1 $0 2007 2008 2009 2010 2011 2012 2013 Phase 2

PAGE 48

48 Maitland Within the City of Ma itland, overall building permit activity declined from 1 , 738 permits issued in 2007 to 1 , 608 permits issued in 2013 ( Table C 2 ). However, the total number of permits issued for the half, one and two mile buffers increased throughout the study period. This is what one would expect if SunRail was stimulating development near the station. Despite a general reduction in the volume of permits issued in the city, the areas near the SunRail station gradually began t o represent a higher percentage of overall activity ( Figure 4 14). Figu re 4 14 . Building permits issued by City of Maitland over study period . The value of the building permits issued in Maitland also declined over the study period. However, the share of the permit value located near the SunRail station increased ( Figure 4 1 5) while the value of the permits located within the quarter and half mile buffers increased in value substantially. Within the h alf mile buffer, it was apparent that permits increased in both value and number while the types of permits also changed. The share of residential permits increased by 14.3% from 2007 to 2013 while the residential share of the value of permits increased 18 .5%. This suggests that residential developments closer to SunRail 0 250 500 750 1000 1250 1500 1750 2000 2007 2008 2009 2010 2011 2012 2013 Total Two Mile One Mile Half Mile Quarter Mile

PAGE 49

49 were more valuable or larger than those outside the half mile buffer and that the makeup of the neighborhoods near the stations was becoming more residential in character. Figure 4 15 . Percent of total building permit value issued by City of Maitland for each distance measure over study period. Table 4 2. Perm its issued within a half mile of SunRail stations compared to all permits issued by the City of Maitland. Number of permits Year Total Percent residential Percent nonresidential Percent of total 2007 105 48.6% 51.4% 5.9 % 2008 102 58.8% 41.2% 6.6 % 20 0 9 127 53.5% 46.5% 8.8 % 2010 113 49.6% 50.4% 7.4 % 2011 158 51.3% 48.7% 10 .0 % 2012 130 53.8% 46.2% 8.7 % 2013 151 62.9% 37.1% 9.4 % Value of permits 2007 $972,429.62 36.5% 63.5% 1.6 % 2008 $903,951.00 53.6% 46.4% 1.3 % 2009 $1,421,767.84 50.3% 49.7% 5.9 % 2010 $2,524,051.22 37.4% 62.6% 8.2 % 2011 $1,653,279.31 59.4% 40.6% 7.7 % 2012 $1,092,892.00 22.9% 77.1% 5.6 % 2013 $1,962,674.27 55.0% 45.0% 7.9 % 0% 20% 40% 60% 80% 100% 2007 2008 2009 2010 2011 2012 2013 Total Two Mile One Mile Half Mile Quarter Mile

PAGE 50

50 Winter Park Within the City of Winter Park, overall building permit activity declined from 6,199 permits issued in 2007 to 5,336 permits issued in 2013. However, like in Maitland, the total number of permits issued for the half, one and two mile buffers increased throughout the study period. This is what one would expect if SunRail was stimulat ing development near the station. Despite a general reduction in the volume of permits issued in the city, starting in 2009 the areas near the SunRail station gradually increased the pace of development ( Figure 4 16). Figure 4 16. Building permits issue d by City of Winter Park over study period. The value of the building permits issued in Winter Park also declined over the study period. However, the share of the permit value located within the quarter half and one mile buffers began to increase in 2011 ( Figure 4 17). Within the half mile buffer, it was apparent that permits increased in both value and number starting in 201 2 af ter declining from 2 007 to 2011 ( Table 4 3 ). The share of permit value falling within the half mile buffer also fluctuated until it began to rise in 2012. The share of residential permits declined by 6.9% from 2007 to 2013 while the residential share of th e value of permits fluctuated without a particular pattern. 0 1000 2000 3000 4000 5000 6000 7000 2007 2008 2009 2010 2011 2012 2013 Total Two Mile One Mile Half Mile Quarter Mile

PAGE 51

51 Figure 4 17. Percent of total b uilding permit value issued by City of Winter Park for each distance measure over study period. Table 4 3. Permits issued within a half mile of SunRail stations compared to all permit s issued by the City of Winter Park. Number of permits Year Total Percent residential Percent nonresidential Percent of total 2007 803 79.3% 20.7% 13.0% 2008 549 72.7% 27.3% 11.0% 2009 662 75.4% 24.6% 15.6% 2010 630 74.8% 25.2% 14.3% 2011 576 68. 8% 31.3% 12.6% 2012 711 68.5% 31.5% 15.4% 2013 902 72.4% 27.6% 16.2% Value of permits 2007 $16,728,214.00 74.6% 25.4% 13.3% 2008 $8,232,511.00 59.2% 40.8% 9.3% 2009 $9,449,177.00 67.0% 33.0% 13.4% 2010 $9,390,259.00 73.1% 26.9% 4.7% 2011 $2 2,370,385.00 73.9% 26.1% 11.9% 2012 $55,159,761.00 80.8% 19.2% 21.7% 2013 $89,847,973.00 52.2% 47.8% 20.5% This lack of strong trends in the Winter Park data was somewhat expected because Winter Park is fairly built out with relatively high value residential and commercial developments near the SunRail station . There may be fewer opportunities 0% 20% 40% 60% 80% 100% 2007 2008 2009 2010 2011 2012 2013 Total Two Mile One Mile Half Mile Quarter Mile

PAGE 52

52 for re development in Winter Park than in other places along SunRail because of Winter Park's pre existing affluence. Orlando Within the City of Orlando , overall building permit activity declined from 1,571 permits issued in 2007 to 788 permits issued in 2013. H owever, unlike Maitland and Winter Park, the total number of permits issued for all distances from the rail stations decreased throughout the study period ( Figure 4 18). This is not what one would expect if SunRail was stimulating development near the stations. However, the four Orlando stations were meant to be destinations instead of origins, so perhaps the ability of a destination station to generate re development is limited by the strength of the overall economy. After all, a commuter ra il system such as this is designed to increase the accessibility of downtown, but if the economy is contracting like it did for much of the study period, one would not expect the increase in accessibility to be important in a time when overall congestion i s decreasing due to the weak economy. Figure 4 18. Building permits issued by City of Orlando over study period. The value of the building permits issued in Orlando also declined over the study period. However, the share of the permit value located with in the quarter half and one mile buffe rs peaked in 2008 and 2011 ( Figure 4 1 9 ). The fluctuating results can be 0 500 1000 1500 2000 2007 2008 2009 2010 2011 2012 2013 Total Two Mile One Mile Half Mile Quarter Mile

PAGE 53

53 explained by the average high value of a relatively small number of permits close to the stations . Figure 4 19. Percent of total building permit value issued by City of Orlando for each distance measure over study period. Ta ble 4 4. Permits issued within a half mile of SunRail stations compared to all permits issued by the City of Orlando. Number of permits Year Total Percent residential Percent nonresidential Percent of total 2007 126 3.2% 96.8% 8.0% 2008 119 12.6% 8 7.4% 12.0% 2009 51 7.8% 92.2% 10.4% 2010 47 10.6% 89.4% 9.9% 2011 34 5.9% 94.1% 7.5% 2012 41 4.9% 95.1% 6.0% 2013 50 12.0% 88.0% 6.3% Value of permits 2007 $25,084,939.00 2.0% 98.0% 4.4% 2008 $205,261,718.00 10.2% 89.8% 29.5% 2009 $50,515, 973.00 0.9% 99.1% 16.6% 2010 $44,059,882.00 1.7% 98.3% 24.0% 2011 $37,940,520.00 1.7% 98.3% 17.0% 2012 $49,496,978.00 2.1% 97.9% 14.4% 2013 $75,258,615.00 35.4% 64.6% 23.4% Within the half mile buffer, permits decreased in both value and number until 2012 when th ey began to increase modest ly ( Table 4 4 ). The share of permit value 0% 20% 40% 60% 80% 100% 2007 2008 2009 2010 2011 2012 2013 Total Two Mile One Mile Half Mile Quarter Mile

PAGE 54

54 falling within the half mile buffer fluctuated throughout the study analysis period. The share of the value of residential permits re mained between 1.7% and 2.1% except for in 2008 and 2013 which had much larger shares of 10.2% and 35.4%, respectively. This lack of strong trends in the Orlando data fits into the issue related to destination stations and accessibility described above. Orange County Building permit data from unincorporated Orange County did not include data from 2007 to 2009 and it did not include permit value data. With only four years permits, it is difficult to establish a trend. Countywide, there was an increase in the number of permits from 2010 to 2012, fol lowed by a decline in 2013 ( Figure 4 20). The permits ranged from 70.4% residential in 2011 to 76.5% residential in 2012. The number of permits issued within the buffers varied greatly, but generally it could be said that there were very few permits at the quarter to half mile buffers. This is not what one would expect if SunRail was stimulating development near the stations. However, of the two stations in unincorporated Orange County, the Sand Lake Road station is a park and ride located in a predominately industrial area whi le M eadow W oods is the Phase 2 station in Orange County. Figure 4 20. Building permits issued by Orange County from 2010 to 2013. 0 100 200 300 400 500 2007 2008 2009 2010 2011 2012 2013 Two Mile One Mile Half Mile Quarter Mile

PAGE 55

55 Table 4 5. Permits issued within a half mile of SunRail stations compared to all permits issued by Orange County. Year Total Percent residential Percent nonresidential Percent of total 2010 0 0.0% 0.0% 0.0% 2011 40 60.0% 40.0% 1.1% 2012 0 0.0% 0.0% 0.0% 2013 3 100.0% 0.0% 0.1% Building Permits by Station The SunRail Transit Oriented Development (TOD) Workshop Sk etchbook ( Olore, 2011) identifie d five TOD typologies that were recommended for SunRail station areas to address their existing conditions and expected growth. Orange County stations fall into four of these typologies: Downtown, Urban Center , Village Cente r and Neighborhood Center. The researcher identified one station in each typology to analyze in greater detail. Complete station level data for all eight stations is located in Table C 2 . Downtown (Church Street S tation) Olore (2011) define d the Downtown t ypology as having high density, mixed uses with a compact pedestrian oriented environment, an active defined center, limited structured parking and urban parks and open space. The Church Street station is located on South Street in downtown Orlando and is a short walking distance to the offices and attractions downtown. It was the second busiest station in Orange County with 2,638 passengers on the first week of paid service ( Fluker, 2014). As seen in Figure 4 21, the number of building permits issued near Church Street station declined between 2008 and 2011. The percent of total permits issued within a quarter mile declined from 0.9% in 2007 to 0.2% in 2011. As discussed above, this may be due to the fact that downtown Orlando was already built out and that the station is a destination for commuters in an area traditionally used as a destination for commuters.

PAGE 56

56 Unlike some of the other stations, the area around this station wa s already well suited for commuter rail. Figure 4 21. Building permits issued nea r Church Street s tation over study period . A substantial portion of all permits issued during the study period were made for projects located within two miles of the Church Street station. In 2008 and 2011, 23.6% and 24.9% of the value of projects permitte d were issued near this station. However, most of this value was located in the one and two mile buffers (outside of the conventional half mile circle walking distance). Figure 4 22. Building permit value distribution issued near Church Street s tat ion over study period. However, within the half mile buffer, over $153 million worth of investments were permitted in this area ( Table 4 6). Much of the development downtown was 0 50 100 150 200 250 300 2007 2008 2009 2010 2011 2012 2013 Two Mile One Mile Half Mile Quarter Mile 0% 5% 10% 15% 20% 25% 30% 2007 2008 2009 2010 2011 2012 2013 Two Mile One Mile Half Mile Quarter Mile

PAGE 57

57 nonresidential. In 2008, there was a single $15 million residential building permit issued. Other than that, there were no residential permits issued until 2012 and 2013 which both had a single residential permit. Table 4 6. Permits issued within a half mile of Church Street station compared to all permits issued from 2007 to 2013. Number of permits Year Total Percent residential Percent nonresidential Percent of total 2007 88 0.0% 100.0 % 0.9% 2008 69 1.4% 98.6% 0.9% 2009 31 0.0% 100.0% 0.5% 2010 29 0.0% 100.0% 0.3% 2011 24 0.0% 100.0% 0.2% 2012 27 3.7% 96.3% 0.2% 2013 26 3.8% 96.2% 0.2% Value of permits 2007 $17,980,306.00 0.0% 100.0% 2.4% 2008 $34,568,860.00 46.3% 53.7% 4.0% 2009 $21,575,317.00 0.0% 100.0% 5.4% 2010 $7,752,345.00 0.0% 100.0% 1.9% 2011 $33,178,850.00 0.0% 100.0% 7.7% 2012 $21,925,532.00 4.6% 95.4% 3.5% 2013 $16,078,790.00 0.0% 100.0% 2.1% Urban Center ( Florida Hospital Health Village S tation ) The Urban Center typology was defined as having high density (predominately residential), mixed uses with a compact pedestrian oriented environ ment, an active defined center, limited structured parking and urban parks and open space (Olore, 2011) . The main difference between the Urban Center and Downtown typologies wa s the lower density of the Urban Center. The Florida Hospital Health Village sta tion is located between two parking garages on the Florida Hospital campus and is a short walking distance to the museums in Loch Haven Park . This station is currently less residential than the typology suggests, with very little high density residential. It was one

PAGE 58

58 of the less busy stations in Orange County with 937 passengers on the first week of paid service (Fluker, 2014). As seen in Figure 4 23 , the number of building permits issued near this station declined between 200 7 and 201 0 . The vast majority of permits were issued for projects in the one or two mile buffers. The percent of total permits issued within a half mile declined from 0.3% to 0.2%, with a high of 0.7% in 2008 . One possible cause for the relatively low number of permits is that much of th e land within walking distance of the station is either a part of the Florida Hospital, Loch Haven Park, or medical offices. Like the area around Church Street station, this area is already built out and that the station is a destination for commuters in a n area traditionally used as a destination for commuters. The majority of the area outside of the institutional land uses within walking distance of the station are single family homes. According to the head of Strategic Property Development with Florida H ospital's parent company, they are in the midst of planning to develop some higher density residential uses on site in the near future. However, the main reason the hospital supported SunRail was to reduce parking demand on their landlocked site. Because o f SunRail, the Florida Hospital is building 1,600 fewer parking spaces in their current expansion efforts than they would have otherwise (J. Barry, personal communication, September 9, 2013) .

PAGE 59

59 Figure 4 23. Building permits issued near Florida Hospital He alth Village station over study period . An increasing portion of all permits issued during the study period were made for projects located within two miles of the station. In 2008 and 2013, 24.8% and 32.8% of the value of projects permitted were issued wer e near this station. In 2008 16.4% of the permitted building value was located within a q uarter mile of the station ( Figure 4 24). Figure 4 24. Building permit value distribution issued near Florida Hospital Health Village station over study period . W ithin the half mile buffer, over $ 228 million worth of investments w ere permitted in this area ( Table 4 7). Much of the permits were residential, but in most years the nonresidential permits accounted for over 90% of the value of the projects . 0 500 1000 1500 2000 2500 2007 2008 2009 2010 2011 2012 2013 Two Mile One Mile Half Mile Quarter Mile 0% 5% 10% 15% 20% 25% 30% 35% 2007 2008 2009 2010 2011 2012 2013 Two Mile One Mile Half Mile Quarter Mile

PAGE 60

60 Table 4 7. Permits issued within a half mile of Florida Hospital Health Village station compared to all permits issued from 2007 to 2013. Number of permits Year Total Percent residential Percent nonresidential Percent of total 2007 32 59.4% 40.6% 0.3% 2008 55 5 8.2% 41.8% 0.7% 2009 17 41.2% 58.8% 0.3% 2010 19 57.9% 42.1% 0.2% 2011 21 57.1% 42.9% 0.2% 2012 22 59.1% 40.9% 0.2% 2013 22 50.0% 50.0% 0.2% Value of permits 2007 $1,405,411.00 37.8% 62.2% 0.2% 2008 $147,083,419.00 3.9% 96.1% 17.2% 2009 $2 6,693,879.00 1.8% 98.2% 6.7% 2010 $14,494,837.00 9.3% 90.7% 3.5% 2011 $4,868,241.00 27.5% 72.5% 1.1% 2012 $26,535,298.00 0.5% 99.5% 4.3% 2013 $7,293,995.00 11.1% 88.9% 0.9% Village Center (Maitland S tation) The Village Center typology ha d medium dens ity (predominately residential, higher densities within a quarter mile of stations with a gradual shift to lower densities), mixed uses integrating residential and local serving retail, a compact pedestrian oriented environment, an active defined center, l imited managed parking, including on street parking and urban parks and open space (Olore, 2011). The Maitland station is located on North Orlando Avenue (US Route 17 92) , one of the major six lane north south arterials through Maitland . The station curren tly is the site of a park and ride lot and is located across the street from a car dealership. It was one of the least busy stations in Orange County with 892 passengers on the first week of paid service (Fluker, 2014).

PAGE 61

61 As seen in Figure 4 2 5 , the number o f building permits issued near this station rose gradually starting in 2010 . The vast majority of permits were issued for projects in the one or two mile buffers. The percent of total permits issued within a half mile rose gradually from 0. 1 % to 0. 3 %, with a high of 0. 6 % in 20 10 . This suggest ed that the pace of development near the station wa s increasing and that the areas near the station we re receiving a greater portion of the developments than before. This makes sense, because the area around the station currently has little in common with the Village Center typology, so there wa s more potential here than other mo r e developed station areas. A large share of permits were located in the one and two mile buffer area. This could indicate that the station's pa rk and ride lot makes the station's influence area wider because passengers arrive by car. Figure 4 25. Building permits issued near Maitland station over study period . An increasing portion of all permits issued during the study period were made for p rojects located within two miles of the station. In 2010, 33.1% of the value of projects permitted wer e issued were near this station ( Figure 4 2 6 ). 0 500 1000 1500 2000 2500 2007 2008 2009 2010 2011 2012 2013 Two Mile One Mile Half Mile Quarter Mile

PAGE 62

62 Figure 4 2 6 . Building permit value distribution issued near Florida Hospital Health Village station ove r study period . Within the half mile buffer, over $ 10 million worth of investments were permitted in t his area from 2007 to 2013 ( Table 4 8). The mix of residential to nonresidential permits generally fluctuated around an even split, suggesting a relative consistency in the types of projects being built in the station area. However, the increase in absolute number of permits suggest ed that the area is growing in part because of SunRail. Table 4 8. Permits issued within a half mile of Maitland station compa red to all permits issued from 2007 to 2013. Number of permits Year Total Percent residential Percent nonresidential Percent of total 2007 107 49.5% 50.5% 1.1% 2008 111 62.2% 37.8% 1.5% 2009 127 53.5% 46.5% 2.1% 2010 115 50.4% 49.6% 1.2% 2011 159 51.6% 48.4% 1.5% 2012 134 55.2% 44.8% 1.2% 2013 154 63.6% 36.4% 1.5% Value of permits 2007 $972,429.62 36.5% 63.5% 0.1% 2008 $1,154,446.00 72.3% 27.7% 0.1% 2009 $1,421,767.84 50.3% 49.7% 0.4% 2010 $2,524,051.22 37.4% 62.6% 0.6% 2011 $1,653, 279.31 59.4% 40.6% 0.4% 2012 $1,155,892.00 27.1% 72.9% 0.2% 2013 $1,974,674.27 55.3% 44.7% 0.3% 0% 5% 10% 15% 20% 25% 30% 35% 2007 2008 2009 2010 2011 2012 2013 Two Mile One Mile Half Mile Quarter Mile

PAGE 63

63 Neighborhood Center (Meadow Woods S tation ) The Neighborhood Center typology is defined as having a low density primarily residential uses with a compac t pedestrian friendly environment, on street parking and urban parks and open space (Olore, 2011). The Meadow Woods station is a SunRail Phase 2 station located on Fairway Woods Boulevard on the former site of a small strip mall. The neighborhood around th e station was built in the late 1980's and consists primarily of single family houses. The neighborhood was built around a golf course that closed in 2007 and is now owned by a church (Shanklin, 2011). Figure 4 27 . Building permits issued nea r Meadow Woods station over study period . As seen in Figure 4 2 7 , there was a large number of p ermits issued within the one and two mile buffers in 2011. However, in each year data was available, the number of permits issued within walking distance was negligible. One possible cause for the relatively low number of permits is that the residential pr operty near the station is built along a closed golf course, so there is uncertainty about the future characteristics of the neighborhood. One plan to redevelop the golf course involved turning the 176 acre site into a New Urbanist style town center in the midst of the existing neighborhood 0 50 100 150 200 250 300 350 400 450 2007 2008 2009 2010 2011 2012 2013 Two Mile One Mile Half Mile Quarter Mile

PAGE 64

64 (Shanklin, 2011) . However, the researcher could not find any indication that the plans to redevelop the golf course were making any progress on this project as of this writing. Table 4 9. Permits issued within a half mile of Meadow Woods station compared to all permits issued from 2010 to 2013. Year Total Percent residential Percent nonresidential Percent of total 2010 0 NA NA 0.0% 20 11 27 88.9% 11.1% 0.3% 2012 0 NA NA 0.0% 2013 3 100.0% 0.0% 0.0% This study investigated the anticipatory effects of SunRail on land use, p arcel just value, number and value of building permits in each jurisdiction, and the number and value of building permits at each station area typ e in Orange County. The analysis of land use within a half mile of stations found minimal changes in the share of property in each land use category. The hedonic regressions of the parcel data found sta tistically significant changes i n the value of residential property located near SunRail stations. All jurisdictions within Orange County saw a reduction in the number of annual permits between 2007 and 2013. Howeve r, Maitland and Winter Park saw an increase in the share of permits and permit value located near their stations. The Village Center type station ( M aitland S tation ) was the only station out of the four stat ion types that saw an increase in the number and value of permits near the stations. The three other station types analyzed did not have an increase in the number or value of permits , probably because of a combination of the recession and physical site constraints near stati ons.

PAGE 65

65 CHAPTER 5 DISCUSSION The results described above must be understood within the context of Orlando's recent transportation history. In the late 1990s, a light rail system was studi ed and federal money was received , but it was cancelled when the Orange County Commission voted not to support the project in 1999 (Krueger, 2001). Because of this, when Orange, Osceola, Volusia, and Seminole counties and the City of Orlando entered into a n interlocal agreement with the FDOT to create SunRail , developers had reason to be skeptical about the likelihood that it would actually be built. Phase 1 of this system, identical to what opened in May 2014, was originally scheduled to open in 2010 (Ham burg & Pino, 2007). From the time that the original interlocal agreements were made in 2007 until Governor Scott approved the spending on the project in July 2011, numerous approval delays made the future of SunRail less than certain (Tracy & Deslatte, 201 1b). In addition to the uncertainty surrounding the creation of SunRail's between 2007 and 2011, that period also saw a severe recession that s ignificantly impacted residential property values. Therefore, it is no surprise that even the strongest residenti al property value models presented above showed little growth in the proximity indicators until after 2011. The idea that the impact of SunRail shouldn't have been felt significantly until after 2011 seems to hold up relatively well in the permit data as w ell. In particular, the shares of permit value within a mile of stations in Maitland and within a half mile in Winter Park seems to have risen substantially after 2011. However, large increases in shares of building permit value were not seen in Orlando . This could also be a result of

PAGE 66

66 the station areas in Orlando being more densely developed than the other jurisdictions. But t he effect here may be less robust than the results in those two cities may suggest . With confidence, it would be difficult to draw any strong conclusions about SunRail's impact on property values or building permit activity near stations. However, the results from the permit data for residential property cannot be dismissed out of hand. They seem to show at least a modest positive effect of commuter rail proximity in this study. The residential one mile buffer model saw a much more modest premium over t he half mile model suggesting that all else being equal, the anticipatory effects of SunRail on residential land uses decreases as distance from the station increases. This is probably a function of the zoning changes allowing higher densities near station s for TOD. Likewise, there was a definite trend in the permit data towards having a larger share of permits being issued near stations. This may be attributable to other investments made in these areas or to changes in zoning to allow for greater density. Many of the effects to did appear to have occurred mirrored the observations of E. H. Spengler in 1930 on the effects of new rail transit opening in New York City. In particular, the changing distribution of the location of building permit values due to S unRail matches Spengler's observation that new transit shifts value rather than creating much new value, and that previously developed areas tend to have smaller changes in value (as cited in Damm et al., 1980, p. 317). A potential barrier to SunRail's abi lity to raise property values is the "back door problem", described by Loukaitou Sideris and Banerjee (2000) as a situation where a transit route is located along the backs of buildings and through nondescript industrial areas lacking easy access to high d ensity residential areas or adequate park and ride

PAGE 67

67 parking capacity. (Loukaitou Sideris & Banerjee, 2000). In this respect, SunRail's prospects may be better than other transit services on former freight right of ways because many of these barriers are no t present at most SunRail stations due to pro active land use policy targeting station areas for higher density development and pedestrian infrastructure improvements (City of Orlando, 2012, Orange County Community, Environmental & Development Services Pla nning Division, 2012, City of Maitland, 2010) . It was outside of the scope of this thesis to determine if these changes in property value or building permit distributions were more dependent upon zoning than SunRail. Due to the large capital costs associat ed with building commuter rail systems, if upzoning wa s all that wa s necessary to spur this type of economic development, it would be important to know by how much . However, if the changes in zoning were made politically possible only because of the presen ce of SunRail, it would be wrong to attribute the changes to zoning alone.

PAGE 68

68 CHAPTER 6 CONCLUSION The purpose of this study was to determine the extent of the impacts of SunRail on land use, property values , and building permit activity from 2007 to 2013 . Orange County has a car dependent growth pattern typical for the southeast whose lessons may apply to other places with similar growth patterns . This study found a statistically significant relationship between the proximity of residential land uses and SunRail stations with the buffer distance mode ls. The half mile buffer distance models show ed that property values near stations were over $18,000 higher than they would be without SunRail . The one mile buffer model saw a much more modest premium over the same time period suggesting that all else bein g equal, the anticipatory effects of SunRail on residential land uses decreases as distance from the station increases. This was probably a function of the zoning changes allowing higher densities near stations for TOD . T he reduction in effect size as dist ance increase d suggests that the market assigns value to being within walking distance from the stations. Building permit data at a municipal level showed that between 2007 and 2013 the share of development located close to stations increased. In Maitland, the station area saw an overall increase in permit activity despite an overall decline in the number of permits issued citywide. Winter Park and Orlando both saw smaller increases in the number and value of permits probably because they were developed to begin with. Permit activities in station areas depended on the context of the station. Downtown and Urban Center stations generally did not see a substantial increase in the number or value of permits over time. However, they remained areas of high investm ent

PAGE 69

69 throughout the study period. This is likely because these areas were already developed in ways that take advantage of the new commuter rail service. Village center stations saw a general increase in the number and value of permits while there was not e nough data to draw firm conclusions about the Neighborhood Center station. Further Research The other effects of proximity to SunRail are les s easy to draw conclusions from. I n particular the linear distance models probably show ed the changes in property v alues in the suburbs more than any direct effect of SunRail. This problem might be remedied by restricting the distance in which models are included in the study. Originally, the researcher included the entire county in the control group and found similar results. Apparently, limiting the control group to properties within the Orange County Urban Service Area was still too broad. Future research might investigate GIS based methods to select parcels based on the existing built density of an area matching th at of SunRail station areas. Additionally, these models could probably be improved if they included more details about the parcels, like number of restrooms or bedrooms in residential buildings . A potential area for future refinement of this study would be to include the built square footage of each land use type instead of the number of parcels. This was not done for this study because of known inconsistencies with the way built square footage is reported in this dataset for timeshares or condo hotels. Thi s land use change could also be improved if it included changes within a land use classification. For instance, if a multifamily dwelling was built on the site of a single family house, the methodology used here would not reflect that change.

PAGE 70

70 Additionally, n ot all SunRail stations in Orange County are TODs. The Maitland and Sand Lake Road stations opened as park and ride stations with low density suburban developments surrounding them. It is conceivable that such stations would be less likely to increase pr operty values nearby because the station would primarily benefit people who arrive by car from a larger catchment area. Limitations T he property and building permit value data w ere not adjusted for inflation throughout the study period . This was not taken into consideration because of the relatively short period of the study and because of the low inflation levels between 2007 and 2013. However, this could potentially account for a significant share of the statistically significant impacts reported in this study. Overall, this research shows that the anticipatory effects of SunRail we re predominantly related to increasing the value of residential properties closer than one mile of stations. Additionally, the areas near SunRail have maintained their original mix of land uses except for a reduction in institutional land uses. Further research should address the issues above and expand the demographic analysis once data becomes available. Without these additions, it would still appear that SunRail has had a posi tive effect on economic development near station areas for residential land uses.

PAGE 71

71 APPENDIX A PARCEL DATA Table A 1 . Land Use Code Classifications. DOR Code Description DOR Code Description Residential land uses 1003 Vacant Multi Family (10 Units Or 300 Multi Family 10+ Units More) 301 Apartment Low Income Housing 1004 Vacant Condo Site Tax Credit 2801 Manufactured Home Park 310 Modern Apartment Complex 3905 Hotel Extended Stay 311 Student Housing 7400 Retirement Community 315 High Rise Apar tment 7800 Rest Home 400 Condominium Residential 1 Vacant Residential 401 Condominium Single Family 4 Vacant Condo Residential 19 Vacant Home Owners Association 450 Condominium Manufactured Home 20 Mfr Home With Sticker 471 Residential Condo Cls 1 100 Single Family 472 Residential Condo Cls 2 101 Single Family 473 Residential Condo Cls 3 102 Single Family Class II 474 Residential Condo Cls 4 103 Single Family Class III 475 Residential Condo Cls 5 104 Single Family Class IV 494 Condominium Single Family 105 Single Family Class V Residential Class 2 119 Improved Home Owner Association 500 Cooperatives 120 Townhouse 550 Cooperatives Manufactured Home 121 Class II Townhouse 600 Retirement Homes 130 Single Family Residential Lake 610 Assiste d Living Front 800 Multi Family 131 Single Family Residential Canal 801 Multi Family 1 Unit Front 802 Multi Family 2 Units 135 Single Family Residential Lake 803 Multi Family 3 Units View 804 Multi Family 4 Units 140 Single Family Residentia l Golf 805 Multi Family 5 9 Units

PAGE 72

72 Table A 1 . Continued DOR Code Description DOR Code Description Residential land uses continued 154 Townhomes Class II 814 Quadraplex 175 Rooming House 821 Class II Duplex 1 Unit 181 1 Unit Of Duplex 822 Class II Duplex 182 1 Unit Of Class 2 Duplex 823 Class II Triplex 194 Single Family 824 Class II Quadraplex 195 Single Family Class 3 830 Multi Family 196 Single Family Class 4 890 Multi Family 197 Single Family Class 5 891 Multi Family Class II 1 Unit 200 Manufactured Home 892 Multi Family Class II 2 Units 201 Manufactured Home 893 Multi Family Class II 3 Units 202 Manufactured Home 894 Multi Family Class II 4 Units 210 Manufactured Home 895 Multi Family Class II 5 9 Units 299 Manufactured H ome Community 900 Rooming House Commercial land uses 1000 Vacant Commercial 3100 Drive In/Open Stadium 1019 Vacant Commercial Association 3200 Theater/Auditorium 1100 Stores, 1 Story 3300 Nightclub/Bars 1101 Condo Retail I 3400 Recreational/Meetin g 1102 Condo Retail II 3500 Tourist Attraction 1103 Condo Retail III 3501 T.A. Sound Stage 1110 Convenience Store 3502 T.A. Stadium 1119 Improved Commercial Association 3503 T.A. Theater 1200 Store/Office/Converted Residential 3504 T.A. Ridehousing 1 210 Store/Office/Res Class 2 3505 Tourist Attraction 1220 Store/Office/Res Class 3 3506 Tourist Attraction 1300 Department Store 3507 Tourist Attraction 1400 Supermarket 3508 Tourist Attraction 1500 Regional Shopping 3509 T.A. Cubic 1600 Community Sho pping 3510 Tourist Attraction 1700 Office Buildings 3511 Tourist Attraction 1701 Condo Prof Bldg 3513 T.A. Theater M.K. 1702 Modular Office 3514 T.A. Ridehousing M.K. 1703 Condo Office I 3515 T.A. Restaurant M.K. 1704 Condo Office II 3517 T.A. Retail M.K. 1705 Condo Office III 3520 Tourist Attraction 1706 Cond Office Medical I 3525 Tourist Attraction 1707 Cond Office Medical II 3575 Tourist Attraction 1710 Cond Off Prof I 3700 Race Tracks 1711 Cond Off Prof II 3800 Golf Course

PAGE 73

73 Table A 1 . Continue d DOR Code Description DOR Code Description Commercial land uses continued 1712 Cond Off Prof III 3900 Motel 1715 Condo Office 2 3 Stories I 3901 Condo Hotel I 1716 Condo Office 2 3 Stories II 3902 Condo Hotel II 1717 Condo Office 2 3 Stories I II 3903 Condo Hotel III 1800 Multi Story Office 2 3 Stories 3904 Condo Hotel IV 1801 High Rise Condo 4+ Stories 3910 Hotel Limited Services 1802 Office 4 8 Stories 3915 Select Service Hotel 1803 Office High Rise 9+ 3920 Hotel Full Service 1900 Profess ional Building 3925 Hotel Luxury 1910 Professional Child Care Center 3930 Convention Center 2100 Restaurants/Cafe 3940 Undeclared Timeshare 2101 Condo Restaurant 7720 Country Club 2200 Restaurant Chain 9011 Lease Retail 2300 Financial Building/Bank 90 17 Lease Office 2400 Insurance Company 9610 Movie Studio 2500 Flex Space 410 Condominium Professional Building 2504 Condo Flex Space I 411 Condominium Office Building Retail 2505 Condo Flex Space II 412 Condominium Office Building 2506 Condo Flex Spac e III 417 Condominium Office Building 2 Or 2510 Telecom/Data Center More Stories 2600 Service Station 419 Condominium Professional 2700 Vehicle Sale Showroom Building (Architectural Design) 2710 Vehicle Service Building 420 Condominium Medical Buil ding 2720 Tire Dealer 421 Condominium Restaurant 2730 Lube Facility 425 Condominium Flexible Space 2740 Vehicle Repair 430 Condominium Time Share 2900 Wholesale Outlet 439 Condominium Hotel/Motel 3000 Florist/Greenhouse Industrial land uses 40 00 Vacant Industrial 4806 Condo Warehouse II 4100 Light Manufacturing 4810 Distribution Warehouse 4110 Class A Manufaturing 4820 Mini Warehouse 4200 Heavy Manufacturing 4830 Truck Terminal 4210 Class A Heavy Industry 4840 Sales Warehouses 4300 Lumber Yards 4900 Open Storage 4400 Packing Plants 8920 Utility, Gas, Electricity, 4500 Bottlers Communications, Water & Sewer 4600 Food Processing 9100 Utility 4610 Food Processing Freezer 9110 Communication Tower Sites

PAGE 74

74 Table A 1 . Continued DOR Code De scription DOR Code Description Industrial land uses continued 4700 Mineral Processing 9810 Railroad Termial/Station/Yard 4800 Warehousing Centrally Assessed 4801 Condo Warehouse Distribution I I 440 Condominium Warehouse 4802 Condo Warehouse Dis tribution II (Distribution) 4805 Condo Warehouse I 448 Condominium Warehouse Institutional land uses 2000 Airports, Commercial 8300 School 2010 Transit Terminals 8400 College 7000 Vacant Institutional 8500 Hospital 7100 Religious 8600 County (O ther Than Public Schools, 7200 School Private Colleges,Hospitals) Including 7300 Hospital Private Non Municip Govt 7301 Hospital Private 8620 Utility, Gas, Electricity, 7500 Charitable Communications, Water & Sewer 7700 Lodge/Union Hall 870 0 State (Other Than Military,Forests, 7900 Cultural Pks,Rec Areas,Hosp,Colleges) 8100 Military 8800 Federal 8286 County Owned 8900 Municipal (Other Than Parks, Rec 8287 State Owned Areas, Colleges, Hospitals) 8288 Federal Owned 8910 Airport 8289 M unicipal Owned Code and titles from Property (DOR) Use Codes. (2010). Orange County Property Appraiser . Retrieved June 24, 2013, from http://www.ocpafl.org/Searches/Lookups.aspx/Code/PropertyUse

PAGE 75

75 Table A 2 . Summary of all Phase 1 regression models. Year 2007 2008 2009 2010 2011 2012 2013 Distance Coefficient Model, All Land Use Types Distance Coefficient 0.476 * 0.510 * 0.350 * 0.593 * 1.040 1.233 1.332 t statistic 1.185 1.320 0.866 1.481 2.887 5.006 3.551 Model Adjusted R Square 0.606 0.634 0.642 0.646 0.679 0.611 0.650 Quarter Mile Coefficient Model, All Land Use Types, Phase 1 Stations Quarter Mile Coefficient 191377.487 86280.812 * 282718.218 2348 77.401 278066.931 309619.435 324387.553 t statistic 2.812 1.296 4.012 3.323 4.406 9.537 6.653 Model Adjusted R Square 0.606 0.634 0.642 0.646 0.679 0.611 0.65 Half Mile Coefficient Model, All Land Use Types, Phase 1 Stations Half Mile Coefficient 61813.557 * 58054.510 * 118231.546 131475.457 173227.950 134471.763 126707.027 t statistic 1.855 1.790 3.518 3.922 5.766 7.597 4.693 Model Adjusted R Square 0.606 0.634 0.642 0.646 0.679 0.611 0.650 One Mile Coefficient Model, All Land Use Types, Phase 1 Stations One Mile Coefficient 81066.512 85245.762 86482.844 91061.656 110063.430 85747.671 109298.088 t statistic 4.337 4.696 4.611 4.884 6.561 7.622 6.3 82 Model Adjusted R Square 0.606 0.634 0.642 0.646 0.679 0.611 0.650 Two Mile Coefficient Model, All Land Use Types, Phase 1 Stations Two Mile Coefficient 103726.868 99292.233 89221.082 98101.270 115082.201 83245.927 109203.643 t statistic 7.396 7.332 6.372 7.043 9.158 9.418 8.119 Model Adjusted R Square 0.606 0.634 0.642 0.646 0.679 0.611 0.650 Distance Coefficient Model, Residential Land Use Types Distance Coefficien t 0.972 1.149 0.964 0.971 1.100 1.221 1.240 t statistic 16.609 20.411 16.816 18.902 23.534 26.847 25.774 Model Adjusted R Square 0.780 0.825 0.837 0.840 0.838 0.824 0.833 * indicates statistically insignificant coefficient at t he 95% confidence level.

PAGE 76

76 Table A 2 . Continued Year 2007 2008 2009 2010 2011 2012 2013 Quarter Mile Coefficient Model, Residential Land Use Types, Phase 1 Stations Quarter Mile Coeffi cient 14965.537 * 10104.149 * 12466.640 * 140933.410 49177.547 99288.684 118442.240 t statistic 0.936 0.617 0.739 9.186 3.560 12.264 14.157 Model Adjusted R Square 0.779 0.825 0.837 0.840 0.837 0.823 0.833 Half Mile Coefficient Model, Residential Land Use Types, Phase 1 Stations Half Mile Coefficient 26988.295 29916.427 23343.238 36347.907 30159.116 45193.416 45693.352 t statistic 4.827 5.500 4.190 7.254 6.624 12.100 11.491 Model Adjusted R Square 0.779 0.825 0.837 0.840 0.837 0.823 0.833 One Mile Coefficient Model, Residential Land Use Types, Phase 1 Stations One Mile Coefficient 63290.511 67761.635 68185.837 69218.937 68964.673 69184.011 74263.611 t sta tistic 23.648 26.083 25.653 29.051 31.787 33.279 33.612 Model Adjusted R Square 0.779 0.825 0.837 0.840 0.837 0.823 0.834 Two Mile Coefficient Model, Residential Land Use Types, Phase 1 Stations Two Mile Coefficient 9 3881.253 94726.892 100834.320 100790.093 97390.415 95643.041 94968.050 t statistic 48.547 50.830 52.952 59.097 62.589 60.466 56.590 Model Adjusted R Square 0.781 0.827 0.838 0.842 0.840 0.825 0.835 Distance Coefficient Model, C ommercial Land Use Types Distance Coefficient 9.811 10.493 9.976 9.057 6.387 * 4.969 10.438 t statistic 3.128 3.415 3.318 3.015 2.212 3.168 2.895 Model Adjusted R Square 0.749 0.785 0.810 0.813 0.756 0.754 0.642 Quarter Mile Coefficient Model, Commercial Land Use Types, Phase 1 Stations Quarter Mile Coefficient 463411.961 * 681041.282 697147.976 590018.718 574745.635 296149.412 306406.789 * t statistic 2.538 3.645 3.773 3.154 3.192 3.256 1.600 Model Adjusted R Square 0.749 0.785 0.810 0.813 0.756 0.754 0.642 * indicates statistically insignificant coefficient at the 95% confidence level.

PAGE 77

77 Table A 2 . Continued Year 2007 200 8 2009 2010 2011 2012 2013 Half Mile Coefficient Model, Commercial Land Use Types, Phase 1 Stations Half Mile Coefficient 366408.305 487136.434 510926.158 478756.915 428012.962 248832.291 314222.188 * t statistic 2 .866 3.738 3.968 3.695 3.435 3.775 2.254 Model Adjusted R Square 0.749 0.785 0.810 0.813 0.756 0.754 0.642 One Mile Coefficient Model, Commercial Land Use Types, Phase 1 Stations One Mile Coefficient 259640.252 * 345642 .695 349891.085 305798.746 263390.483 * 209880.866 288097.367 * t statistic 2.355 3.117 3.192 2.778 2.481 3.338 2.158 Model Adjusted R Square 0.749 0.785 0.810 0.813 0.756 0.754 0.642 Two Mile Coefficient Model, Commercial Land Us e Types, Phase 1 Stations Two Mile Coefficient 12591.548 * 27499.838 * 38392.214 * 12890.748 * 27525.428 * 67047.839 * 135665.027 * t statistic 0.121 0.267 0.380 0.127 0.280 1.124 1.073 Model Adjusted R Square 0.749 0.785 0.810 0.813 0.756 0.754 0.642 Distance Coefficient Model, Institutional Land Use Types Distance Coefficient 16.197 * 21.665 * 18.900 * 16.868 * 16.551 * 3.520 * 16.754 * t statistic 2.266 2.365 1.495 1.415 1.377 0.309 0.911 Model Adjusted R Square 0.909 0.911 0.877 0.887 0.881 0.748 0.763 Quarter Mile Coefficient Model, Institutional Land Use Types, Phase 1 Stations Quarter Mile Coefficient 778217.785 * 657720.145 * 1925516.915 * 1789043.328 * 1322105.574 * 3048256.997 2448170.109 * t statistic 1.380 1.104 1.859 1.833 1.351 3.329 1.731 Model Adjusted R Square 0.909 0.911 0.877 0.887 0.881 0.748 0.764 Half Mile Coefficient Model, Institutiona l Land Use Types, Phase 1 Stations Half Mile Coefficient 456033.466 * 384133.357 * 1214531.476 * 1433421.875 * 1344442.377 * 1571856.506 851335.845 * t statistic 1.330 0.953 2.089 2.596 2.405 2.930 1.023 Model Adjusted R Square 0.909 0.911 0.877 0.887 0.881 0.748 0.763 * indicates statistically insignificant coefficient at the 95% confidence level.

PAGE 78

78 Table A 2 . Continued Year 2007 2008 2009 2010 2011 20 12 2013 One Mile Coefficient Model, Institutional Land Use Types, Phase 1 Stations One Mile Coefficient 128577.687 * 86760.761 * 167590.043 * 230787.162 * 200244.629 273248.080 * 172929.438 * t statistic 0.439 0.253 0.352 0.515 0.439 0.627 0.249 Model Adjusted R Square 0.909 0.911 0.876 0.887 0.881 0.748 0.763 Two Mile Coefficient Model, Institutional Land Use Types Two Mile Coefficient 13451.715 * 59747.417 * 253480.540 * 1289 63.261 * 145289.338 * 83960.788 * 57265.750 * t statistic 0.051 0.183 0.549 0.296 0.329 0.200 0.086 Model Adjusted R Square 0.909 0.911 0.876 0.887 0.881 0.748 0.763 Distance Coefficient Model, Industrial Land Use Types Distance Coefficient 6.094 5.899 5.072 3.911 3.806 4.664 3.964 t statistic 4.727 4.316 4.003 3.425 4.218 5.643 5.222 Model Adjusted R Square 0.807 0.823 0.862 0.864 0.882 0.859 0.865 Quarter Mile Coefficient Model, Industrial Land Use Types, Phase 1 Stations Quarter Mile Coefficient 130401.218 * 228355.900 * 174963.434 * 150372.209 * 96790.284 * 37929.261 * 50594.987 * t statistic 1.256 1.974 1.613 1.498 1.227 0.71 3 1.025 Model Adjusted R Square 0.806 0.822 0.862 0.864 0.881 0.858 0.864 Half Mile Coefficient Model, Industrial Land Use Types, Phase 1 Stations Half Mile Coefficient 51527.124 * 101181.197 * 83715.451 * 43025.887 * 1 5953.186 * 62284.547 * 79823.238 * t statistic 0.749 1.405 1.224 0.687 0.321 1.529 2.110 Model Adjusted R Square 0.806 0.822 0.862 0.864 0.881 0.858 0.864 One Mile Coefficient Model, Industrial Land Use Types, Phase 1 Stations One Mile Coefficient 57414.549 * 1788.096 * 8147.360 * 26209.191 * 15215.915 * 80157.467 * 79674.360 * t statistic 1.168 0.034 0.167 0.590 0.433 2.534 2.728 Model Adjusted R Square 0.806 0.822 0.862 0.864 0.881 0.858 0.865 * indicates statistically insignificant coefficient at the 95% confidence level.

PAGE 79

79 Table A 2 . Continued Year 2007 2008 2009 2010 2011 2012 2013 Two Mile Coefficient Model, Industrial Land Use Types, Phase 1 Stations Two Mile Coefficient 116922.377 * 73031.046 * 96792.543 * 121358.343 93821.292 123819.783 135238.955 t statistic 2.606 1.523 2.172 3.000 2.937 4.240 5.021 Mod el Adjusted R Square 0.807 0.822 0.862 0.864 0.882 0.859 0.865 * indicates statistically insignificant coefficient at the 95% confidence level.

PAGE 80

80 Table A 3 . Summary of all Phase 2 regression models. Year 2007 2008 2009 20 10 2011 2012 2013 Quarter Mile Coefficient Model, All Land Use Types, Phase 2 Stations Distance Coefficient 154776.572 66919.211 * 223905.916 181604.913 217338.924 275719.250 293160.173 t statistic 2.471 1.094 3. 470 2.810 3.758 8.860 6.265 Model Adjusted R Square 0.707 0.634 0.642 0.646 0.679 0.611 0.650 Half Mile Coefficient Model, All Land Use Types, Phase 2 Stations Quarter Mile Coefficient 47911.872 * 47409.355 * 92602.32 5 101024.833 136203.365 116088.370 110506.297 t statistic 1.552 1.578 2.976 3.257 4.892 6.856 4.281 Model Adjusted R Square 0.606 0.634 0.642 0.646 0.679 0.611 0.650 One Mile Coefficient Model, All Land Use Types, Phase 2 Station s Half Mile Coefficient 71543.633 78095.564 71652.707 71180.761 87927.931 71278.078 97041.394 t statistic 3.968 4.471 3.968 3.965 5.441 6.492 5.806 Model Adjusted R Square 0.606 0.634 0.642 0.646 0.679 0.611 0.650 Two Mile Coefficient Model, All Land Use Types, Phase 2 Stations One Mile Coefficient 103810.845 101861.895 81424.784 81258.872 100294.523 69777.870 98171.402 t statistic 7.112 7.234 5.577 5.599 7.659 7.658 7. 080 Model Adjusted R Square 0.606 0.634 0.642 0.646 0.679 0.611 0.650 Quarter Mile Coefficient Model, Residential Land Use Types, Phase 2 Stations Two Mile Coefficient 43042.438 32282.619 * 36835.118 33446.430 19794.21 4 * 68255.956 86764.948 t statistic 3.363 2.519 2.798 2.808 1.834 9.173 11.229 Model Adjusted R Square 0.779 0.825 0.837 0.840 0.837 0.823 0.833 Half Mile Coefficient Model, Residential Land Use Types, Phase 2 Stations Distance Coefficient 724.382 * 3598.717 * 6311.599 * 1986.214 * 5178.794 * 24997.048 26489.119 t statistic 0.147 0.751 1.286 0.450 1.290 7.190 7.158 Model Adjusted R Square 0.779 0.825 0.837 0.840 0.837 0.823 0.833 * indicates statistically insignificant coefficient at the 95% confidence level.

PAGE 81

81 Table A 3 . Continued Year 2007 2008 2009 2010 2011 2012 2013 One Mile Coefficient Model, Residential Land Use Types, P hase 2 Stations Quarter Mile Coefficient 43290.259 47433.505 42878.053 37344.946 39310.902 50778.260 57010.918 t statistic 16.940 19.176 16.945 16.456 19.011 25.205 26.636 Model Adjusted R Square 0.780 0.825 0.837 0 .840 0.838 0.824 0.833 Two Mile Coefficient Model, Residential Land Use Types, Phase 2 Stations Half Mile Coefficient 82067.582 84092.022 82559.529 71156.232 72537.055 79285.741 80052.467 t statistic 40.465 43.053 41.3 60 39.755 44.416 48.381 46.097 Model Adjusted R Square 0.781 0.826 0.838 0.841 0.838 0.825 0.834 Quarter Mile Coefficient Model, Commercial Land Use Types, Phase 2 Stations One Mile Coefficient 459096.907 * 670355.641 68 3785.124 579403.052 562412.996 293868.141 307638.442 * t statistic 2.531 3.617 3.734 3.126 3.152 3.243 1.612 Model Adjusted R Square 0.749 0.785 0.810 0.813 0.756 0.754 0.642 Half Mile Coefficient Model, Commercial Land Use Types, Phase 2 Stations Two Mile Coefficient 367772.534 485188.364 508613.963 477746.173 423721.799 349499.560 319367.846 * t statistic 2.883 3.735 3.963 3.700 3.411 3.784 2.291 Model Adjusted R Square 0.749 0.785 0.810 0. 813 0.756 0.754 0.642 One Mile Coefficient Model, Commercial Land Use Types, Phase 2 Stations Distance Coefficient 264603.585 * 346973.076 350493.654 309673.235 262142.609 * 213850.057 301134.981 * t statistic 2.392 3.119 3 .183 2.802 2.458 3.386 2.245 Model Adjusted R Square 0.749 0.785 0.810 0.813 0.756 0.754 0.642 Two Mile Coefficient Model, Commercial Land Use Types, Phase 2 Stations Quarter Mile Coefficient 10204.356 * 47130.331 * 56 727.723 * 16713.162 * 162.724 * 61869.193 * 139623.039 * t statistic 0.097 0.452 0.551 0.162 0.002 1.270 1.092 Model Adjusted R Square 0.749 0.785 0.810 0.813 0.756 0.754 0.642 * indicates statistically insignificant coefficien t at the 95% confidence level.

PAGE 82

82 Table A 3 . Continued Year 2007 2008 2009 2010 2011 2012 2013 Quarter Mile Coefficient Model, Institutional Land Use Types, Phase 2 Stations Half Mile Co efficient 772414.894 * 657720.145 * 1925516.915 * 1789043.328 * 1322105.574 * 2969435.057 2365016.451 * t statistic 1.375 1.104 1.859 1.833 1.351 3.277 1.695 Model Adjusted R Square 0.909 0.911 0.877 0.887 0.881 0.748 0.764 Half Mile Coefficient Model, Institutional Land Use Types, Phase 2 Stations One Mile Coefficient 489043.057 * 419086.971 * 1196224.131 * 1417342.092 * 1330942.837 * 1592065.798 806084.592 t statistic 1.433 1.041 2.061 2.572 2.386 2.977 0.973 Model Adjusted R Square 0.909 0.911 0.877 0.887 0.881 0.748 0.763 One Mile Coefficient Model, Institutional Land Use Types, Phase 2 Stations Two Mile Coefficient 162326.834 * 119522.146 * 124944.484 * 184711.7 31 * 157408.501 * 289973.182 * 127971.118 * t statistic 0.554 0.347 0.261 0.411 0.344 0.661 0.183 Model Adjusted R Square 0.909 0.911 0.876 0.887 0.881 0.748 0.763 Two Mile Coefficient Model, Institutional Land Use Types Distance Coefficient 31507.827 * 95000.137 * 389464.283 * 263409.846 * 277074.166 * 71045.871 * 128256.353 * t statistic 0.115 0.282 0.816 0.585 0.606 0.163 0.184 Model Adjusted R Square 0.909 0.911 0.876 0.887 0.881 0.748 0.764 Quarter Mile Coefficient Model, Industrial Land Use Types, Phase 2 Stations Quarter Mile Coefficient 130401.218 * 228355.900 * 174963.434 * 150372.209 * 96790.284 * 37929.261 * 50594.987 * t statistic 1.256 1 .974 1.613 1.498 1.227 0.713 1.025 Model Adjusted R Square 0.806 0.822 0.862 0.864 0.881 0.858 0.864 Half Mile Coefficient Model, Industrial Land Use Types, Phase 2 Stations Half Mile Coefficient 51527.124 * 101181 .197 * 83715.451 * 43025.887 * 15953.186 * 62284.547 * 79823.238 * t statistic 0.749 1.405 1.224 0.687 0.321 1.529 2.110 Model Adjusted R Square 0.806 0.822 0.862 0.864 0.881 0.858 0.864 * indicates statistically insignificant coe fficient at the 95% confidence level.

PAGE 83

83 Table A 3 . Continued Year 2007 2008 2009 2010 2011 2012 2013 One Mile Coefficient Model, Industrial Land Use Types, Phase 2 Stations One Mile Coefficient 59831.621 * 499.862 * 30402.575 * 4639.281 * 3805.057 * 66533.841 * 64971.795 * t statistic 1.220 0.010 0.626 0.105 0.109 2.110 2.232 Model Adjusted R Square 0.806 0.822 0.862 0.864 0.881 0.858 0.865 Two Mile Coef ficient Model, Industrial Land Use Types, Phase 2 Stations Two Mile Coefficient 167644.265 168061.431 113373.559 * 119603.447 100639.571 119828.736 117585.839 t statistic 3.774 3.440 2.489 2.893 3.077 4.031 4.295 Model Adjusted R Square 0.807 0.823 0.862 0.864 0.882 0.859 0.865 * indicates statistically insignificant coefficient at the 95% confidence level.

PAGE 84

84 Table A 4 . Orange County parcel mean just value, all land uses Year Mean Just Value Number of Parcels Standard Deviation 2007 $312,276.24 358820 3427201.68 2008 $344,949.40 366568 3484422.69 2009 $337,415.02 368904 3598650.11 2010 $282,537.13 370263 3547286.72 2011 $252,353.12 373732 3342377.08 2 012 $225,453.13 439906 2443412.89 2013 $242,937.38 431679 3790686.22 Table A 5. Land use types as a percent of total parcels within a half mile of stations . Year 2007 2008 2009 2010 2011 2012 2013 All land use types Total 358820 366568 368904 370263 373732 439906 431679 Half Mile 5178 5171 5125 5102 5188 8095 8081 Residential land uses Total 88.97% 88.92% 89.53% 89.67% 89.59% 86.55% 89.49% Half Mile 62.69% 62.56% 63.06% 63.07% 62.70% 63.53% 63.32% Commercial land uses Total 3.62% 3.90% 3.92% 3.88% 3.91% 8.03% 4.81% Half Mile 20.26% 19.40% 19.98% 19.82% 19.83% 25.66% 24.16% Institutional land uses Total 2.30% 1.68% 1.49% 1.51% 1.49% 1.21% 1.36% Half Mile 8.59% 8.26% 6.48% 6.37% 6.32% 3.84% 4.57% Industrial land uses Total 1.20% 1.26% 1.23% 1.24% 1.25% 1.10% 1.14% Half Mile 3.75% 4.22% 4.12% 4.10% 4.05% 3.66% 3.61% Other land uses Total 3.91% 4.24% 3.84% 3.70% 3.77% 3.11% 3.21% Half Mile 4.71% 5.57% 6.36% 6.64% 7.09% 3.31% 4.34%

PAGE 85

85 APPENDIX B JURISDICTION LEVEL PERMIT DATA Table B 1. Summary of building permit data by jurisdiction. Maitland Permits Year 2007 2008 2009 2010 2011 2012 2013 Qu arter Mile Number of Permits 62 50 65 43 68 44 56 % Residential 51.6% 56.0% 49.2% 53.5% 50.0% 45.5% 57.1% % Other 48.4% 44.0% 50.8% 46.5% 50.0% 54.5% 42.9% % of Total in Jurisdiction 3.5% 3.2% 4.5% 2.8% 4.3% 2.9% 3.5% To tal Value of Permits $328,037 $517,087 $888,436 $996,878 $287,020 $247,832 $950,284 % Residential 54.5% 52.4% 47.0% 76.4% 64.9% 13.6% 56.2% % Other 45.5% 47.6% 53.0% 23.6% 35.1% 86.4% 43.8% % of Total in Jurisdiction 0.6% 0.7% 3.7% 3.2% 1.3% 1. 3% 3.8% Half Mile Number of Permits 105 102 127 113 158 130 151 % Residential 48.6% 58.8% 53.5% 49.6% 51.3% 53.8% 62.9% % Other 51.4% 41.2% 46.5% 50.4% 48.7% 46.2% 37.1% % of Total in Jurisdiction 5.9% 6.6% 8.8% 7.4% 10.0 % 8.7% 9.4% Total Value of Permits $972,430 $903,951 $1,421,768 $2,524,051 $1,653,279 $1,092,892 $1,962,674 % Residential 36.5% 53.6% 50.3% 37.4% 59.4% 22.9% 55.0% % Other 63.5% 46.4% 49.7% 62.6% 40.6% 77.1% 45.0% % of Total in Jurisdiction 1.6% 1.3% 5.9% 8.2% 7.7% 5.6% 7.9% One Mile Number of Permits 468 560 528 522 650 675 763 % Residential 51.9% 58.4% 50.2% 58.6% 52.2% 54.1% 56.2% % Other 48.1% 41.6% 49.8% 41.4% 47.8% 45.9% 43.8% % of Total in Jurisdicti on 26.2% 36.2% 36.7% 34.1% 41.2% 45.1% 47.5% Total Value of Permits $19,334,082 $13,832,411 $7,881,384 $9,184,329 $7,014,849 $9,398,038 $12,322,852 % Residential 76.2% 59.5% 33.3% 42.0% 48.2% 33.4% 44.3% % Other 23.8% 40.5% 66.7% 58.0% 51.8% 66 .6% 55.7% % of Total in Jurisdiction 32.6% 19.8% 32.8% 29.8% 32.6% 48.3% 49.6%

PAGE 86

86 Table B 1. Continued Maitland Permits Year 2007 2008 2009 2010 2011 2012 2013 Two Mile Number of Permits 1125 1173 1080 1158 1304 1391 1381 % Residential 50.8% 57.3% 52.6% 56.3% 51.0% 53.3% 56.1% % Other 49.2% 42.7% 47.4% 43.7% 49.0% 46.7% 43.9% % of Total in Jurisdiction 63.1% 75.9% 75.2% 75.7% 82.6% 93.0% 85.9% Total Value of Permits $41,069,505 $43 ,051,925 $16,189,558 $24,953,673 $18,697,809 $17,882,973 $22,622,765 % Residential 68.5% 72.3% 32.1% 45.7% 38.5% 39.3% 48.6% % Other 31.5% 27.7% 67.9% 54.3% 61.5% 60.7% 51.4% % of Total in Jurisdiction 69.3% 61.5% 67.4% 80.9% 86.8% 91.8% 91.1 % Total in Jurisdiction Number of Permits 1783 1546 1437 1530 1578 1496 1608 % Residential 50.8% 57.6% 52.5% 55.2% 50.8% 53.7% 56.8% % Other 49.2% 42.4% 47.5% 44.8% 49.2% 46.3% 43.2% % of Total in Jurisdiction 100.0% 100 .0% 100.0% 100.0% 100.0% 100.0% 100.0% Total Value of Permits $59,224,700 $69,993,812 $24,006,617 $30,857,213 $21,544,320 $19,471,988 $24,841,515 % Residential 71.1% 48.2% 29.8% 42.0% 38.3% 42.8% 49.8% % Other 28.9% 51.8% 70.2% 58.0% 61.7% 57.2 % 50.2% % of Total in Jurisdiction 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Winter Park Permits Quarter Mile Number of Permits 357 235 342 283 194 258 209 % Residential 80.1% 75.3% 75.4% 76.0% 66.0% 66.7% 141.1% % Other 19.9% 24.7% 24.6% 24.0% 34.0% 33.3% 41.1% % of Total in Jurisdiction 5.8% 4.7% 8.0% 6.4% 4.2% 5.6% 3.8% Total Value of Permits $4,981,551 $5,021,486 $4,548,399 $5,067,468 $10,006,686 $14,882,889 $49,362,365 % Residenti al 92.9% 44.5% 67.7% 69.3% 62.1% 73.8% 41.9% % Other 7.1% 55.5% 32.3% 30.7% 37.9% 26.2% 58.1% % of Total in Jurisdiction 4.0% 5.6% 6.5% 2.5% 5.3% 5.8% 11.3%

PAGE 87

87 Table B 1. Continued Winter Park Permits Year 2007 2008 200 9 2010 2011 2012 2013 Half Mile Number of Permits 803 549 662 630 576 711 902 % Residential 79.3% 72.7% 75.4% 74.8% 68.8% 68.5% 72.4% % Other 20.7% 27.3% 24.6% 25.2% 31.3% 31.5% 27.6% % of Total in Jurisdiction 13.0% 11. 0% 15.6% 14.3% 12.6% 15.4% 16.2% Total Value of Permits $16,728,214 $8,232,511 $9,449,177 $9,390,259 $22,370,385 $55,159,761 $89,847,973 % Residential 74.6% 59.2% 67.0% 73.1% 73.9% 80.8% 52.2% % Other 25.4% 40.8% 33.0% 26.9% 26.1% 19.2% 47.8% % of Total in Jurisdiction 13.3% 9.3% 13.4% 4.7% 11.9% 21.7% 20.5% One Mile Number of Permits 2621 1628 1680 1756 1887 2032 2643 % Residential 78.6% 70.7% 73.7% 72.1% 69.1% 69.3% 74.0% % Other 21.4% 29.3% 26.3% 27.9% 30.9% 30.7% 26.0% % of Total in Jurisdiction 42.3% 32.7% 39.5% 39.7% 41.2% 44.0% 47.5% Total Value of Permits $51,618,000 $25,665,778 $25,075,465 $36,022,099 $72,006,221 $118,647,099 $211,527,040 % Residential 72.0% 72.3% 73.6% 70.9% 80.1% 72.4% 65. 3% % Other 28.0% 27.7% 26.4% 29.1% 19.9% 27.6% 34.7% % of Total in Jurisdiction 41.1% 28.8% 35.7% 18.0% 38.4% 46.6% 48.3% Two Mile Number of Permits 6011 4538 4021 4069 4220 4394 5336 % Residential 78.7% 70.2% 73.7% 71.2% 68.7% 68.9% 74.3% % Other 21.3% 29.8% 26.3% 28.8% 31.3% 31.1% 25.7% % of Total in Jurisdiction 97.0% 91.2% 94.5% 92.1% 92.2% 95.1% 95.9% Total Value of Permits $115,975,873 $78,987,590 $68,588,297 $184,530,952 $157,727,606 $248,127,499 $426,98 0,736 % Residential 76.6% 66.9% 67.2% 29.7% 83.7% 68.2% 67.7% % Other 23.4% 33.1% 32.8% 70.3% 16.3% 31.8% 32.3% % of Total in Jurisdiction 92.3% 88.8% 97.6% 92.1% 84.2% 97.5% 97.6%

PAGE 88

88 Table B 1. Continued Winter Park Permits Year 2007 2008 2009 2010 2011 2012 2013 Total in Jurisdiction Number of Permits 6199 4974 4257 4419 4577 4618 5563 % Residential 78.6% 70.1% 73.9% 71.0% 68.6% 68.9% 74.1% % Other 21.4% 29.9% 26.1% 29.0% 31.4% 31. 1% 25.9% % of Total in Jurisdiction 1 1 1 1 1 1 1 Total Value of Permits $125,683,664 $88,999,266 $70,293,614 $200,460,213 $187,336,672 $254,529,130 $437,615,748 % Residential 72.9% 67.6% 66.6% 28.6% 84.7% 68.3% 67.5% % Other 27.1% 32.4% 33 .4% 71.4% 15.3% 31.7% 32.5% % of Total in Jurisdiction 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Orlando Permits Quarter Mile Number of Permits 85 38 29 23 21 24 23 % Residential 0.0% 2.6% 0.0% 0.0% 0 .0% 4.2% 4.3% % Other 100.0% 97.4% 100.0% 100.0% 100.0% 95.8% 95.7% % of Total in Jurisdiction 5.4% 3.8% 5.9% 4.8% 4.6% 3.5% 2.9% Total Value of Permits $11,792,564 $160,099,068 $44,919,195 $14,352,100 $29,458,473 $40,331,177 $35,501,459 % Residential 0.0% 1.8% 0.0% 0.0% 0.0% 2.5% 0.0% % Other 100.0% 98.2% 100.0% 100.0% 100.0% 97.5% 100.0% % of Total in Jurisdiction 2.1% 23.0% 14.7% 7.8% 13.2% 11.7% 11.0% Half Mile Number of Permits 126 119 51 47 34 41 50 % Residential 3.2% 12.6% 7.8% 10.6% 5.9% 4.9% 12.0% % Other 96.8% 87.4% 92.2% 89.4% 94.1% 95.1% 88.0% % of Total in Jurisdiction 8.0% 12.0% 10.4% 9.9% 7.5% 6.0% 6.3% Total Value of Permits $25,084,939 $205,261,718 $50,515,973 $44,059,882 $37,940 ,520 $49,496,978 $75,258,615 % Residential 2.0% 10.2% 0.9% 1.7% 1.7% 2.1% 35.4% % Other 98.0% 89.8% 99.1% 98.3% 98.3% 97.9% 64.6% % of Total in Jurisdiction 4.4% 29.5% 16.6% 24.0% 17.0% 14.4% 23.4%

PAGE 89

89 Table B 1. Continued Orlando Permi ts Year 2007 2008 2009 2010 2011 2012 2013 One Mile Number of Permits 219 251 131 87 63 67 99 % Residential 32.4% 26.7% 36.6% 20.7% 19.0% 17.9% 28.3% % Other 67.6% 73.3% 63.4% 79.3% 81.0% 82.1% 71.7% % of Total in Jurisdiction 13.9% 25.4% 26.7% 18.3% 13.9% 9.8% 12.6% Total Value of Permits $52,236,514 $329,104,953 $69,374,106 $53,129,573 $107,424,042 $53,406,701 $91,344,363 % Residential 36.2% 10.5% 23.6% 9.2% 3.3% 7.0% 38.4% % Other 63.8% 89.5% 76.4% 90.8% 96.7% 93.0% 61.6% % of Total in Jurisdiction 9.2% 47.3% 22.7% 28.9% 48.0% 15.5% 28.4% Two Mile Number of Permits 317 343 166 125 109 104 175 % Residential 36.0% 33.2% 38.0% 24.0% 27.5% 25.0% 25.7% % Oth er 64.0% 66.8% 62.0% 76.0% 72.5% 75.0% 74.3% % of Total in Jurisdiction 20.2% 34.7% 33.8% 26.3% 24.0% 15.1% 22.2% Total Value of Permits $84,590,196 $369,445,479 $83,195,840 $60,953,087 $120,470,059 $62,252,421 $108,344,550 % Residential 46.2% 13.5% 26.3% 14.6% 7.1% 12.3% 36.9% % Other 53.8% 86.5% 73.7% 85.4% 92.9% 87.7% 63.1% % of Total in Jurisdiction 15.0% 53.1% 27.3% 33.2% 53.9% 18.1% 33.7% Total in Jurisdiction Number of Permits 1571 989 491 475 454 687 788 % Residential 56.1% 45.0% 44.8% 45.9% 54.4% 64.3% 54.2% % Other 43.9% 55.0% 55.2% 54.1% 45.6% 35.7% 45.8% % of Total in Jurisdiction 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Total Value of Permits $565,512,909 $695,946,159 $305,086,96 4 $183,579,657 $223,703,017 $343,864,785 $321,839,508 % Residential 55.4% 31.2% 30.6% 30.6% 39.6% 47.2% 52.6% % Other 44.6% 68.8% 69.4% 69.4% 60.4% 52.8% 47.4% % of Total in Jurisdiction 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

PAGE 90

90 Tabl e B 1. Continued Year 2007 2008 2009 2010 2011 2012 2013 Orange County Permits Quarter Mile Number of Permits 0 6 0 1 % Residential 0.0% 100.0% 0.0% 100.0% % Other 0.0% 0.0% 0.0% 0.0% % of Total in Jurisdiction 0.0% 0.2% 0.0% 0.0% Half Mile Number of Permits 0 40 0 3 % Residential 0.0% 60.0% 0.0% 100.0% % Other 0.0% 40.0% 0.0% 0.0% % of Total in Jurisdiction 0.0% 1.1% 0.0% 0.1% O ne Mile Number of Permits 35 117 2 69 % Residential 37.1% 54.7% 100.0% 98.6% % Other 62.9% 45.3% 0.0% 1.4% % of Total in Jurisdiction 1.0% 3.1% 0.0% 2.8% Two Mile Number of Permits 285 54 2 67 232 % Residential 65.3% 76.8% 79.1% 77.2% % Other 34.7% 23.2% 20.9% 22.8% % of Total in Jurisdiction 8.2% 14.3% 1.7% 9.3% Total in Jurisdiction Number of Permits 3464 3800 4059 2506 % Residentia l 74.6% 70.4% 76.5% 76.4% % Other 25.4% 29.6% 23.5% 23.6% % of Total in Jurisdiction 100.0% 100.0% 100.0% 100.0%

PAGE 91

91 APPENDIX C STATION LEVEL DATA Table C 1 . Total SunRail boardings by station for the week of May 19, 2014 . Station Boardings Maitland 892 Winter Park 3707 Florida Hospital Health Village 937 Lynx Central 1406 Church Street 2638 Orlando Health/Amtrak 666 Sand Lake Road 2035 Meadow Woods NA Fluker, A. (2014, May 28). See which SunRail stations drew th e biggest numbers in Week 1 Orlando Business Journal. Orlando Business Journal . Retrieved June 8, 2014, from http://www.bizjournals.com/orlando/blog/2014/05/see which sunrail stations drew the biggest.html?ana=twt

PAGE 92

92 Table C 2 . Summary of building pe rmit data by station. Maitland Station Permits Year 2007 2008 2009 2010 2011 2012 2013 Quarter Mile Number of Permits 64 59 65 45 69 48 59 % Residential 53.1% 62.7% 49.2% 55.6% 50.7% 50.0% 59.3% % Other 46.9% 37.3% 50.8% 44.4% 49.3% 50.0% 40.7% % of Total 1% 1% 1% 0% 1% 0% 1% Total Value of Permits $328,037 $767,582 $888,436 $996,878 $287,020 $310,832 $962,284 % Residential 54.5% 67.9% 47.0% 76.4% 64.9% 31.1% 56.7% % Other 45.5% 32.1% 53. 0% 23.6% 35.1% 68.9% 43.3% % of Total 0.0% 0.1% 0.2% 0.2% 0.1% 0.1% 0.1% Half Mile Number of Permits 107 111 127 115 159 134 154 % Residential 49.5% 62.2% 53.5% 50.4% 51.6% 55.2% 63.6% % Other 50.5% 37.8% 46.5% 49.6% 48.4 % 44.8% 36.4% % of Total 1% 1% 2% 1% 2% 1% 1% Total Value of Permits $972,430 $1,154,446 $1,421,768 $2,524,051 $1,653,279 $1,155,892 $1,974,674 % Residential 36.5% 72.3% 50.3% 37.4% 59.4% 27.1% 55.3% % Other 63.5% 27.7% 49.7% 62.6% 40.6% 72 .9% 44.7% % of Total 0.1% 0.1% 0.4% 0.6% 0.4% 0.2% 0.3% One Mile Number of Permits 462 583 529 520 648 681 765 % Residential 52.2% 59.0% 50.3% 58.8% 52.2% 54.5% 56.3% % Other 47.8% 41.0% 49.7% 41.2% 47.8% 45.5% 43.7% % of Total 4.8% 7.8% 8.6% 5.3% 6.2% 6.3% 7.3% Total Value of Permits $19,325,449 $14,337,534 $7,887,384 $9,149,329 $7,005,819 $9,474,855 $12,324,067 % Residential 76.2% 60.8% 33.4% 41.7% 48.2% 33.9% 44.3% % Other 23.8% 39.2% 66.6% 58.3% 51.8 % 66.1% 55.7% % of Total 2.6% 1.7% 2.0% 2.2% 1.6% 1.5% 1.6%

PAGE 93

93 Table C 2 . Continued Maitland Station Permits Year 2007 2008 2009 2010 2011 2012 2013 Two Mile Number of Permits 2070 1949 1687 187 5 1965 2016 2155 % Residential 62.9% 62.1% 60.2% 61.0% 57.6% 58.3% 62.6% % Other 37.1% 37.9% 39.8% 39.0% 42.4% 41.7% 37.4% % of Total 21.7% 26.0% 27.3% 19.0% 18.9% 18.6% 20.6% Total Value of Permits $61,447,419 $55,781,679 $22,874,429 $137, 150,840 $47,822,120 $78,937,078 $114,921,226 % Residential 73.7% 75.1% 47.1% 14.3% 69.7% 38.1% 56.6% % Other 26.3% 24.9% 52.9% 85.7% 30.3% 61.9% 43.4% % of Total 8.2% 6.5% 5.7% 33.1% 11.1% 12.8% 14.7% Winter Park Station Permits Quarter Mile Number of Permits 355 226 342 281 193 254 406 % Residential 80.0% 74.3% 75.4% 75.8% 65.8% 66.1% 71.9% % Other 20.0% 25.7% 24.6% 24.2% 34.2% 33.9% 28.1% % of Total 3.7% 3.0% 5.5% 2.8% 1.9% 2.3% 3.9% Tot al Value of Permits $4,981,551 $4,770,991 $4,548,399 $5,067,468 $10,006,686 $14,819,889 $49,350,365 % Residential 92.9% 41.6% 67.7% 69.3% 62.1% 73.7% 41.8% % Other 7.1% 58.4% 32.3% 30.7% 37.9% 26.3% 58.2% % of Total 0.7% 0.6% 1.1% 1.2% 2.3% 2.4 % 6.3% Half Mile Number of Permits 779 512 655 621 563 692 885 % Residential 79.6% 72.5% 75.6% 74.6% 68.2% 68.1% 72.2% % Other 20.4% 27.5% 24.4% 25.4% 31.8% 31.9% 27.8% % of Total 8.2% 6.8% 10.6% 6.3% 5.4% 6.4% 8.5% T otal Value of Permits $16,687,367 $7,195,594 $9,389,854 $8,786,459 $21,486,364 $55,012,656 $88,911,978 % Residential 74.6% 54.0% 66.8% 71.2% 72.8% 80.7% 51.8% % Other 25.4% 46.0% 33.2% 28.8% 27.2% 19.3% 48.2% % of Total 2.2% 0.8% 2.4% 2.1% 5.0% 8.9% 11.3%

PAGE 94

94 Table C 2 . Continued Winter Park Station Permits Year 2007 2008 2009 2010 2011 2012 2013 One Mile Number of Permits 2241 1418 1492 1543 1598 1787 2398 % Residential 78.8% 70. 7% 74.1% 71.5% 69.0% 69.1% 73.9% % Other 21.2% 29.3% 25.9% 28.5% 31.0% 30.9% 26.1% % of Total 23.5% 18.9% 24.1% 15.6% 15.4% 16.5% 22.9% Total Value of Permits $42,649,616 $22,390,741 $23,708,935 $31,005,852 $61,648,160 $110,916,227 $178,572,363 % Residential 75.8% 69.1% 75.5% 70.2% 80.5% 72.0% 60.9% % Other 24.2% 30.9% 24.5% 29.8% 19.5% 28.0% 39.1% % of Total 5.7% 2.6% 5.9% 7.5% 14.3% 18.0% 22.8% Two Mile Number of Permits 6148 4725 4176 4285 4590 4776 5644 % Residential 77.3% 69.5% 72.1% 70.4% 67.5% 67.6% 73.3% % Other 22.7% 30.5% 27.9% 29.6% 32.5% 32.4% 26.7% % of Total 64.4% 62.9% 67.5% 43.3% 44.1% 44.0% 53.9% Total Value of Permits $127,366,585 $98,579,069 $71,167,783 $188,610,971 $161,567,06 6 $247,695,752 $432,090,146 % Residential 77.3% 63.8% 65.8% 30.1% 82.8% 68.0% 67.6% % Other 22.7% 36.2% 34.2% 69.9% 17.2% 32.0% 32.4% % of Total 17.0% 11.5% 17.8% 45.5% 37.3% 40.1% 55.1% Florida Hospital Health Village Station Permits Quarter Mile Number of Permits 2 11 5 5 6 5 7 % Residential 0.0% 9.1% 0.0% 0.0% 0.0% 0.0% 0.0% % Other 100.0% 90.9% 100.0% 100.0% 100.0% 100.0% 100.0% % of Total 0.0% 0.1% 0.1% 0.1% 0.1% 0.0% 0.1% Total Valu e of Permits $41,350 $140,011,614 $25,722,728 $12,971,000 $3,387,300 $18,538,297 $6,365,000 % Residential 0.0% 2.0% 0.0% 0.0% 0.0% 0.0% 0.0% % Other 100.0% 98.0% 100.0% 100.0% 100.0% 100.0% 100.0% % of Total 0.0% 16.4% 6.4% 3.1% 0.8% 3.0% 0.8 %

PAGE 95

95 Table C 2 . Continued Florida Hospital Health Village Station Permits Year 2007 2008 2009 2010 2011 2012 2013 Half Mile Number of Permits 32 55 17 19 21 22 22 % Residential 59.4% 58.2% 4 1.2% 57.9% 57.1% 59.1% 50.0% % Other 40.6% 41.8% 58.8% 42.1% 42.9% 40.9% 50.0% % of Total 0.3% 0.7% 0.3% 0.2% 0.2% 0.2% 0.2% Total Value of Permits $1,405,411 $147,083,419 $26,693,879 $14,494,837 $4,868,241 $26,535,298 $7,293,995 % Resident ial 37.8% 3.9% 1.8% 9.3% 27.5% 0.5% 11.1% % Other 62.2% 96.1% 98.2% 90.7% 72.5% 99.5% 88.9% % of Total 0.2% 17.2% 6.7% 3.5% 1.1% 4.3% 0.9% One Mile Number of Permits 428 240 203 241 304 255 269 % Residential 76.4% 67.9% 6 8.0% 72.2% 66.8% 67.8% 70.3% % Other 23.6% 32.1% 32.0% 27.8% 33.2% 32.2% 29.7% % of Total 4.5% 3.2% 3.3% 2.4% 2.9% 2.3% 2.6% Total Value of Permits $21,641,048 $155,941,455 $29,352,186 $21,270,271 $15,104,811 $35,913,154 $45,474,959 % Resid ential 60.1% 8.7% 6.6% 31.6% 56.3% 20.5% 72.2% % Other 39.9% 91.3% 93.4% 68.4% 43.7% 79.5% 27.8% % of Total 2.9% 18.2% 7.3% 5.1% 3.5% 5.8% 5.8% Two Mile Number of Permits 2293 1591 1535 1515 1659 1748 2323 % Residential 7 6.8% 72.0% 74.2% 71.9% 67.6% 68.4% 72.9% % Other 23.2% 28.0% 25.8% 28.1% 32.4% 31.6% 27.1% % of Total 24.0% 21.2% 24.8% 15.3% 15.9% 16.1% 22.2% Total Value of Permits $73,934,273 $212,103,836 $54,967,238 $46,623,316 $66,238,265 $109,949,289 $25 7,274,765 % Residential 69.1% 16.3% 33.1% 51.0% 72.6% 55.0% 63.0% % Other 30.9% 83.7% 66.9% 49.0% 27.4% 45.0% 37.0% % of Total 9.9% 24.8% 13.8% 11.2% 15.3% 17.8% 32.8%

PAGE 96

96 Table C 2 . Continued Year 2007 2008 2009 2010 2011 2012 2013 Lynx Central Station Permits Quarter Mile Number of Permits 2 1 2 1 0 0 2 % Residential 0.0% 0.0% 0.0% 0.0% NA NA 0.0% % Other 100.0% 100.0% 100.0% 100.0% NA NA 100.0% % of Total 0.0% 0.0% 0. 0% 0.0% 0.0% 0.0% 0.0% Total Value of Permits $162,325 $700,000 $700,000 $350,000 $0 $0 $273,000 % Residential 0.0% 0.0% 0.0% 0.0% NA NA 0.0% % Other 100.0% 100.0% 100.0% 100.0% NA NA 100.0% % of Total 0.0% 0.1% 0.2% 0.1% 0.0% 0.0% 0.0% Half Mile Number of Permits 26 58 13 12 7 13 19 % Residential 0.0% 1.7% 7.7% 0.0% 14.3% 0.0% 26.3% % Other 100.0% 98.3% 92.3% 100.0% 85.7% 100.0% 73.7% % of Total 0.3% 0.8% 0.2% 0.1% 0.1% 0.1% 0.2% Total Value of Permits $6,186,378 $20,486,561 $17,772,407 $1,027,795 $7,801,727 $1,316,680 $38,126,856 % Residential 0.0% 0.0% 0.1% 0.0% 2.2% 0.0% 69.9% % Other 100.0% 100.0% 99.9% 100.0% 97.8% 100.0% 30.1% % of Total 0.8% 2.4% 4.4% 0.2% 1.8% 0.2% 4.9% One Mile Number of Permits 130 157 50 43 42 38 55 % Residential 8.5% 11.5% 8.0% 4.7% 9.5% 10.5% 21.8% % Other 91.5% 88.5% 92.0% 95.3% 90.5% 89.5% 78.2% % of Total 1.4% 2.1% 0.8% 0.4% 0.4% 0.3% 0.5% Total Value of Permits $25,649,9 36 $125,283,845 $23,850,729 $8,908,611 $99,956,244 $23,183,334 $60,136,045 % Residential 13.3% 15.6% 4.5% 5.5% 0.7% 7.7% 47.7% % Other 86.7% 84.4% 95.5% 94.5% 99.3% 92.3% 52.3% % of Total 3.4% 14.7% 6.0% 2.1% 23.1% 3.8% 7.7%

PAGE 97

97 Table C 2 . Cont inued Year 2007 2008 2009 2010 2011 2012 2013 Lynx Central Station Permits Two Mile Number of Permits 241 243 133 91 98 91 123 % Residential 33.2% 27.6% 44.4% 28.6% 26.5% 100.0% 28.5% % Other 66.8% 72.4% 55.6% 71.4% 73.5% 0.0% 71.5% % of Total 2.5% 3.2% 2.2% 0.9% 0.9% 0.8% 1.2% Total Value of Permits $64,116,293 $289,518,438 $78,103,099 $50,711,750 $112,510,164 $58,316,541 $97,434,343 % Residential 44.6% 11.9% 26.8% 10.9% 6.4 % 12.1% 37.1% % Other 55.4% 88.1% 73.2% 89.1% 93.6% 87.9% 62.9% % of Total 8.5% 33.9% 19.6% 12.2% 26.0% 9.4% 12.4% Church Street Station Permits Quarter Mile Number of Permits 78 25 21 17 14 17 13 % Resi dential 0.0% 0.0% 0.0% 0.0% 0.0% 5.9% 7.7% % Other 100.0% 100.0% 100.0% 100.0% 100.0% 94.1% 92.3% % of Total 0.8% 0.3% 0.3% 0.2% 0.1% 0.2% 0.1% Total Value of Permits $11,487,889 $11,387,454 $18,067,467 $1,031,100 $26,068,723 $20,899,542 $14,23 1,590 % Residential 0.0% 0.0% 0.0% 0.0% 0.0% 4.8% 0.0% % Other 100.0% 100.0% 100.0% 100.0% 100.0% 95.2% 100.0% % of Total 1.5% 1.3% 4.5% 0.2% 6.0% 3.4% 1.8% Half Mile Number of Permits 88 69 31 29 24 27 26 % Residenti al 0.0% 1.4% 0.0% 0.0% 0.0% 3.7% 3.8% % Other 100.0% 98.6% 100.0% 100.0% 100.0% 96.3% 96.2% % of Total 0.9% 0.9% 0.5% 0.3% 0.2% 0.2% 0.2% Total Value of Permits $17,980,306 $34,568,860 $21,575,317 $7,752,345 $33,178,850 $21,925,532 $16,078,790 % Residential 0.0% 46.3% 0.0% 0.0% 0.0% 4.6% 0.0% % Other 100.0% 53.7% 100.0% 100.0% 100.0% 95.4% 100.0% % of Total 2.4% 4.0% 5.4% 1.9% 7.7% 3.5% 2.1%

PAGE 98

98 Table C 2 . Continued Year 2007 2008 2009 2010 2011 2012 2013 Chur ch Street Station Permits One Mile Number of Permits 126 149 80 46 42 39 42 % Residential 9.5% 12.8% 47.5% 8.7% 9.5% 5.1% 16.7% % Other 90.5% 87.2% 52.5% 91.3% 90.5% 94.9% 83.3% % of Total 1.3% 2.0% 1.3% 0. 5% 0.4% 0.4% 0.4% Total Value of Permits $24,292,629 $113,187,741 $37,546,084 $31,035,828 $99,509,505 $23,142,984 $34,601,589 % Residential 11.3% 16.6% 36.5% 1.7% 0.4% 4.6% 4.7% % Other 88.7% 83.4% 63.5% 98.3% 99.6% 95.4% 95.3% % of Total 3 .2% 13.2% 9.4% 7.5% 23.0% 3.7% 4.4% Two Mile Number of Permits 221 260 141 133 88 89 119 % Residential 33.0% 27.3% 36.9% 45.9% 26.1% 29.2% 25.2% % Other 67.0% 72.7% 63.1% 54.1% 73.9% 70.8% 74.8% % of Total 2.3% 3.5% 2.3% 1.3% 0.8% 0.8% 1.1% Total Value of Permits $49,773,754 $201,757,626 $53,634,070 $41,242,265 $107,869,228 $31,049,021 $90,402,419 % Residential 41.2% 16.6% 36.0% 9.3% 5.6% 19.1% 37.6% % Other 58.8% 83.4% 64.0% 90.7% 94.4% 80.9% 62.4% % of To tal 6.6% 23.6% 13.4% 9.9% 24.9% 5.0% 11.5% Orlando Health/Amtrak Station Permits Quarter Mile Number of Permits 3 1 1 0 1 2 1 % Residential 0.0% 0.0% 0.0% #DIV/0! 0.0% 0.0% 0.0% % Other 100.0% 100.0% 100.0% #D IV/0! 100.0% 100.0% 100.0% % of Total 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% Total Value of Permits $10,100 $8,000,000 $429,000 $0 $2,450 $893,338 $14,631,869 % Residential 0.0% 0.0% 0.0% #DIV/0! 0.0% 0.0% 0.0% % Other 100.0% 100.0% 100.0% #DIV /0! 100.0% 100.0% 100.0% % of Total 0.0% 0.9% 0.1% 0.0% 0.0% 0.1% 1.9%

PAGE 99

99 Table C 2 . Continued Year 2007 2008 2009 2010 2011 2012 2013 Orlando Health/Amtrak Station Permits Half Mile Number o f Permits 9 3 5 3 1 4 4 % Residential 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% % Other 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% % of Total 0.1% 0.0% 0.1% 0.0% 0.0% 0.0% 0.0% Total Value of Permits $3,111,108 $8,373,375 $1,804,000 $22,050, 500 $2,450 $944,253 $15,292,169 % Residential 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% % Other 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% % of Total 0.4% 1.0% 0.5% 5.3% 0.0% 0.2% 1.9% One Mile Number of Permits 126 66 87 55 28 27 33 % Residential 23.8% 24.2% 43.7% 30.9% 21.4% 11.1% 27.3% % Other 76.2% 75.8% 56.3% 69.1% 78.6% 88.9% 72.7% % of Total 1.3% 0.9% 1.4% 0.6% 0.3% 0.2% 0.3% Total Value of Permits $28,205,436 $61,512,452 $21,511,731 $34,923,038 $28,7 19,701 $22,668,921 $33,743,411 % Residential 26.2% 5.2% 64.6% 4.1% 8.1% 7.0% 8.3% % Other 73.8% 94.8% 35.4% 95.9% 91.9% 93.0% 91.7% % of Total 3.8% 7.2% 5.4% 8.4% 6.6% 3.7% 4.3% Two Mile Number of Permits 215 235 123 229 66 81 106 % Residential 27.9% 23.4% 35.8% 48.0% 19.7% 18.5% 23.6% % Other 72.1% 76.6% 64.2% 52.0% 80.3% 81.5% 76.4% % of Total 2.3% 3.1% 2.0% 2.3% 0.6% 0.7% 1.0% Total Value of Permits $45,259,337 $181,875,205 $46,602,491 $39,839,280 $104,5 38,003 $26,936,479 $85,876,348 % Residential 36.3% 15.2% 33.1% 7.2% 3.4% 9.2% 37.4% % Other 63.7% 84.8% 66.9% 92.8% 96.6% 90.8% 62.6% % of Total 6.0% 21.3% 11.7% 9.6% 24.2% 4.4% 10.9%

PAGE 100

100 Table C 2 . Continued Year 2007 2008 2009 2010 2011 2012 2013 Sand Lake Road Station Permits Quarter Mile Number of Permits 0 0 0 0 % Residential NA NA NA NA % Other NA NA NA NA % of Total 0.0% 0.0% 0.0% 0.0% Half Mile Number of Permits 0 11 0 0 % Residential NA 0.0% NA NA % Other NA 100.0% NA NA % of Total 0.0% 0.1% 0.0% 0.0% One Mile Number of Permits 0 36 2 3 % Residential NA 5.6% 100.0% 66.7% % Other NA 94.4% 0.0% 33.3% % of Total 0.0% 0.3% 0.0% 0.0% Two Mile Number of Permits 22 127 16 28 % Residential 72.7% 33.9% 87.5% 71.4% % Other 27.3% 66.1% 12.5% 28.6% % of Total 0.2% 1.2% 0.1% 0.3 % Meadow Woods Station Permits Quarter Mile Number of Permits 0 6 0 1 % Residential NA 100.0% NA 100.0% % Other NA 0.0% NA 0.0% % of Total 0.0% 0.1% 0.0% 0.0%

PAGE 101

101 Table C 2 . Continued Year 2007 2008 2009 2010 2011 2012 2013 Meadow Woods Station Permits Half Mile Number of Permits 0 27 0 3 % Residential NA 88.9% NA 100.0% % Other NA 11.1% NA 0.0% % of Total 0. 0% 0.3% 0.0% 0.0% One Mile Number of Permits 15 78 0 66 % Residential 0.0% 78.2% NA 100.0% % Other 100.0% 21.8% NA 0.0% % of Total 0.2% 0.7% 0.0% 0.6% Two Mile Number of Permits 29 38 5 30 205 % Residential 0.0% 88.8% 100.0% 77.6% % Other 100.0% 11.2% 0.0% 22.4% % of Total 0.3% 3.7% 0.3% 2.0% All Permit Data Number of Permits 9553 7509 6185 9888 10409 10860 10465 % Residential 69.7% 64.2% 66.6% 68.6% 65.9% 69.4% 70.5% % Other 30.3% 35.8% 33.4% 31.4% 34.1% 30.6% 29.5% % of Total 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Total Value of Permits $750,421,273 $854,939,237 $399,387,195 $414,897,083 $432,584,009 $617,865,903 $7 84,296,771 % Residential 59.6% 36.3% 36.9% 30.5% 59.0% 55.8% 60.8% % Other 40.4% 63.7% 63.1% 69.5% 41.0% 44.2% 39.2% % of Total 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

PAGE 102

102 LIST OF REFERENCES Alonso, W. (1964). The historic and the structural theories of urban form: their implications for urban renewal. Land Economics, 40(2), 227 231. Retrieved June 11, 2014, from http://www.jstor.org/stable/3144355 Badoe, D. A., & Miller, E. J. (200 0). Transportation land use interaction: empirical findings in North America, and their implications for modeling. Transportation Research Part D: Transport and Environment , 5 (4), 235 263. Retrieved September 25, 2013 from http:// dx.doi.org/10.1016/S1361 9209(99)00036 X Bae, C. H. C., Jun, M. J., & Park, H. (2003). The impact of Seoul's subway Line 5 on residential property values. Transport Policy , 10 (2), 85 94. Retrieved September 25, 2013 from http:// dx.doi.org/10.1016/S0967 070X(02)00048 3 Baum Snow, N., & Kahn, M. E. (2000). The effects of new public projects to expand urban rail transit. Journal of Public Economics, 77(2), 241 263 . Retrieved September 25, 2013 from http:// dx.doi.org/10.1016/S0047 2727(99)00085 7 Bellinger, W. K. (2006). The economic valuation of train horn noise: A US case study. Transportation Research Part D: Transport and Environment, 11(4), 310 314. Retrieved May 25, 2014 from http:// dx.doi.org/ 10.1016/j.trd.2006.06.002 Billings, S. B. (2011). Estimating the value of a new transit option. R egional Science and Urban Economics, 41(6), 525 536. Retrieved September 25, 2013, from http://dx.doi.org/10.1016/j.regsciurbeco.2011.03.013 Bowes, D. R., & Ihlanfeldt, K. R. (2001). Identi fying the impacts of rail transit stations on residential property values. Journal of Urban Economics , 50 (1), 1 25. Retrieved September 25, 2014 from http:// dx.doi.org/10.1006/juec.2001.2214 Brons, M., Nijkamp, P., Pels, E., & Rietveld, P. (2003). Railroad noise: economic valuation and policy. Transportation Research Part D: Transport and Environment, 8(3), 169 184. Retrieved May 27, 2014 from http:// dx.doi.org/10.1016/S1361 9209(02)00048 2 Celik, H. M., & Yankaya, U. (2006). The impact of rail transit investment on the residential property values in developing countries: The case of Izmir Subway, Turkey. Property Management, 24(4), 369 382. R etrieved May 29, 2014, from http://dx.doi.org/10.1108/02637470610671604 Cervero, R., & Duncan, M. (2002). Transit's value added effects: light and commuter rail services and commercial land values. Transportation Research Record: Journal of the Transportation Research Board , 1805(1), 8 15. Retrieved on March 6, 2014, from http://dx.doi.org/10.3141/1805 02

PAGE 103

103 Cervero, R., & Kang, C. D. (2011). Bus rapid tra nsit impacts on land uses and land values in Seoul, Korea. Transport Policy , 18 (1), 102 116. Retrieved September 25, 2013 from http:// dx.doi.org/10.1016/j.tranpol.2010.06.005 Cervero, R., & Land is, J. (1993). Assessing the impacts of urban rail transit on local real estate markets using quasi experimental comparisons. Transportation Research Part A: Policy and Practice , 27 (1), 13 22.Retrieved September 25, 2013 from http:// dx.doi.org/10.1016/0965 8564(93)90013 B Cervero, R., & Landis, J. (1997). Twenty years of the Bay Area Rapid Transit system: Land use and development impacts. Transportation Research Part A: Policy and Practice, 31(4), 309 333. Retrieved March 3, 2014, from http://dx.doi.org/10.1016/S0965 8564(96)00027 4 City of Orlando. (2012). Growth Management Plan . Retrieved from http://www.cityoforlando.net/city planning/comprehensive plan/ City of Maitland. (2010). City of Maitland 2030 Comprehensive Development Plan. Retrieved from http://www.itsmymaitland.com/myJSSImages/file/CDP2030.pdf Damm, D., Lerman, S. R., Lerner Lam, E., & Young, J. (1980). Response of urban real estate values in anticipation of the Washington Metro. Journal of Transport Economics and Policy, 14(3), 315 336. Retrieved June 11, 2014, from http://www.jstor.org/stable/20052588 Debrezion, G., Pels, E., & Rietveld, P. (2007 ). The impact of railway stations on residential and commercial property value: a me ta analysis. The Journal of Real Estate Finance and Economics , 35 (2), 161 180. Retrieved September 25, 2013 from http:// dx.doi.org/10.1007/s11146 007 9032 z Federal Railroad Administration (FRA). (n .d.). Horn Noise FAQ. United States Department of Transportation Federal Railroad Administration. Retrieved May 20, 2014, from http://www.fra.dot.gov/Page/P0599 Fluker, A. (2014, May 28). See which SunRail st ations drew the biggest numbers in Week 1. Orlando Business Journal . Retrieved June 8, 2014, from http://www.bizjournals.com/orlando/blog/ 2014/05/see which sunrail stations drew the biggest.html?ana=twt Garrett, T. A. (2004). Light Rail Transit in America: Policy Issues and Prospects for Economic Development. St. Louis: Federal Reserve Bank of St. Louis. Retrieved June 1, 2014 from http://www.stlouisfed.org/community_development/assets/pdf/light_rail.pdf Grimes, A., & Young, C. (2010). Anticipatory effects of rail upgrades: Auckland's Western Line. Welli ngton, N.Z.: Motu Economic and Public Policy Research. Retrieved March 4, 2014 from http://dx.doi.org/10.2139/ssrn.1684406

PAGE 104

104 Guerra, E., Cervero, R., & Tischler, D. (2011). The half mile circle: does it re present transit station catchments?. Retrieved September 25, 2013 from http://escholarship.org/uc/item/0d84c2f4 Hamburg, J., & Pino, M. (2007, July 24). Orlando on board for rail -Osceola postpones v ote.. The Orlando Sentinel , Retrieved July 7, 2014, from http://www.highbeam.com/doc/1G1 166744644.html?refid=bibme_hf Henneberry, J. (1998). Transport investment and house prices. Journal of Property Valuation and Investment, 16(2), 144 158. Retrieved on March 4, 2014 from http://dx.doi.org/10.1108/14635789810212913 Knaap, G. J., Ding, C., & Hopkins, L. D. (2001). Do plans m atter?: the effects of light rail plans on land values in station areas. Journal of Planning Education and Research, 21(1), 32 39. Retrieved June 4, 2014, from http://dx.doi.org/10.1177/0739456X0102 100103 Krueger, J. (2001, October 30). Orange votes to pay for share of light rail study. Orlando Business Journal . Retrieved July 17, 2014, from http://www.bizjournals.com/or lando/stories/2001/10/29/daily21.html Landis, J., Guhathakurta, S. & Zhang, M. (1994). Capitalization of transit investments into single family home prices: a comparative analysis of five California rail transit systems. University of California Transport ation Center . UC Berkeley: University of California Transportation Center. Retrieved June 24, 2014 from http://escholarship.org/uc/item/80f3p5n1 Loukaitou Sideris, A. (2010). A new found popularity for transit oriented developments? Lessons from Southern California. Journal of Urban Design , 15 (1), 49 68. Retrieved September 5, 2013 from http:// dx.doi.org/10.1080/13574800903429399 Loukaitou Sideri s, A., & Banerjee, T. (2000). The Blue Line blues: Why the vision of transit village may not materialize despite impressive growth in transit ridership. Journal of Urban Design , 5 (2), 101 125. Retrieved September 25, 2013 from http:// dx.doi.org/10.1080/713683964 Mcmillen, D. P., & Mcdonald, J. (2004). Reaction of House Prices to a New Rapid Transit Line: Chicago's Midway Line, 1983 1999. Real Estate Economics, 32(3), 463 486. Retrieved March 4, 2014, from http://dx.doi.org/10.1111/j.1080 8620.2004.00099.x Olore, T. (2011). SunRail Transit Oriented Development Workshop Sketchbook Update Section 2. Orlando, FL: Florida Department of Transportation . Retrieved on June 2, 2014 from http://business.sunrail.com/welcome/page/transitorienteddevelopment

PAGE 105

105 Orange County Community, Environmental & Development Services Planning Di vision. (2012). Orange County, FL Comprehensive plan 2010 2030: Destination 2030 . Retrieved September 25, 2013 from ht tp://www.orangecountyfl.net/Portals/0/resource%20library/planning%20 %20development/Comprehensive%20Plan%20GOPS%202030.pdf Panda Consulting. (2012) Florida Parcel Data 2012 . Florida Geographic Data Library. Retrieved May 27, 2014 from http://www.fgdl.org/ Petheram, S. J., Nelson, A. C., Miller, M., & Ewing, R. (2013). Use of the real estate market to establish light rail station catchment areas. Transportation Research Record: Journal of the Transportation Research Boar d, 2357( 1), 95 99. Retrieved March 4, 2014, from http://dx.doi.org/10.3141/2357 11 Project Documents. (2013). SunRail A Better Way to Go. Retrieved November 1, 2013, from http://corporate.sunrail.com/welcome/page/projectdocuments Shanklin, M. (2011, May 14). Hispanic church is unlikely champion of SunRail. Orlando Sentinel. Retrieved June 9, 2014, from http://articles.orlandosentinel.com/2011 05 14/business/os sunrail church mortgage 20110514_1_sunrail rail line church leaders Shanklin, M. (2013, A ugust 8). SunRail slow to spark development. SunRail slow to spark development. Retrieved September 8, 2013, from http://articles.orlandosentinel.com/2013 08 08/business/os sunrail development 20130808_1_sunrail rida development corp downtown orlando Tracy, D. (2009, February 24). Central Florida commuter rail: Who pays if somebody gets hurt?. Orlando Sentine l . Retrieved J une 7, 2014, from http://articles.orlandosentinel.com/2009 02 24/news/sunrail24_1_csx sovereign immunity liability for accidents Tracy, D. (2010, December 8). SunRail back on track after Amtrak backs down. Orlando Sentinel . Retrieved July 3, 2014, from http://articles.orlandosentinel.com/2010 12 08/business/os amtrak sunrail 20101208_1_sunrail supporters central florida commuter train amtrak officials Tracy, D. (2013, November 23). Federal money for SunRail i n jeopardy. Orlando Sentinel . Retrieved November 25, 2013, from http://articles.orlandosentinel.com/2013 11 23/news/os sunr ail money worried 20131123_1_sunrail federal money attkisson Tracy, D. (2014, May 20). SunRail paid ridership below expectations. Orlando Sentinel. Retrieved June 5, 2014, from http://articles.orlandosentinel.com/2014 05 20/news/os sunrail day one riders 20140520_1_sunrail officials maitland station ridership

PAGE 106

106 Tracy, D., & Deslatte, A. (2011 a , January 28). Rick Scott freezes four contracts for SunRail. Orlando Sentinel . Retrieved July 3, 2014, from http:// articles.orlandosentinel.com/2011 01 28/news/os sunrail contracts delayed 20110128_1_sunrail supporters jacksonville train high speed train Tracy, D., & Deslatte, A. T. (2011 b , July 1). Central Florida's SunRail project is back on track. Orlando Sentinel . Retrieved July 6, 2014, from http://articles.orlandosentinel.com/2011 07 01/news/os sunra il scott decision 20110701_1_commuter train project cost overruns transportation alternative U.S. Census Bureau. (2014). State and County QuickFacts. Orange County QuickFacts from the US Census Bureau. Retrieved June 11, 2014, from http://quickfacts.census.gov/qfd/states/12/12095.html U.S. Department of Transportation (USDOT)., Federal Transit Administration (FTA)., & Florida Department of Transportation (FDOT). (2008). 1. Central Florida Commuter Rail Transit Supplemental Environmental Impact Statement (pp. 1 7). Retrieved on June 11, 2014 from http://business.sunrail.com/uploads/allprojectdocs/435.pdf Value Adjustmen t Board. (n.d.). Orange County Comptroller. Retrieved June 6, 2014, from http://www.occompt.com/index.php/vab/vab Weinstein, B. L., & Clover, T. L. (2002). An assement of the DART LRT on taxable proper ty valuations and transit oriented development. Dallas, TX: Dallas Area Rapid Transit. Retrieved June 24, 2014 from http://www.valleymetro.org/ images/uploads/lightrail_publications/2002_DART_LR T__Property_Values.pdf

PAGE 107

107 BIOGRAPHICAL SKETCH Benjamin Lytle was born in Melbourne, FL in 1988. He graduated from the University at Buffalo in 2011 with a Bachelor of Arts in e nvironmental d esign with a min or in a rchitecture. In 2012, he entered the University of Florida's Graduate School to study u rban and r egional p lanning. His interests are focused on transportation planning and the interactions between transportation and land use. He is currently employe d by AECOM in Orlando, FL as a planner.