1# RECORD 3777ce734c4a5c9380dfa227ede8503f4f25b44e2# TITLE: The impact of urban light rail transit on double-story terraced property values in the Greater Kuala Lumpur, Malaysia3# YEAR: 2021.0 VENUE: nan DOI: 10.22146/ijg.463024# AUTHORS: M. Dziauddin5# PAGES: 6 CANDIDATE TABLE PAGES (1-based): [3]6# ABSTRACT: Introducing a rail transit system into an urban region is expected to increase land values, and subsequently, residential property values. Despite this general belief, there has been limited research on the impact of proximity to urban light rail transit stations in developing countries setting. This study, therefore, investigates the impact of proximity to urban light rail transit on double-story terraced property values in the Greater Kuala Lumpur, Malaysia, using hedonic pricing model. The findings suggest urban light rail transit has a positive and significant impact on surrounding double-story terraced residential property values – estimated at a premium of 12.3% and 9.8% for properties located up to 1000 m and 1001 – 2000 m to the nearest light rail transit station respectively. The findings of the study are useful for investors and developers, namely to build more houses near urban rail transit infrastructure since they lead to significant appreciation in value.789===== PAGE 1 (score 18, numeric-paren 0, table-words 0, var-words 18) =====1011e impact of urban light rail transit on double-story terraced property 12values in the Greater Kuala Lumpur, Malaysia 13 14Mohd Faris Dziauddin 15Department of Geography, Faculty of Human Sciences, Universiti Pendidikan Sultan Idris, Malaysia 16Abstract. It is generally believed that a rail transit system plays a pivotal role in increasing land values, and 17subsequently, residential property values. Despite this general belief, there is considerably little empirical 18research on the impact of urban light rail transit systems on residential property values in developing 19countries setting. is study, therefore, investigates the impact of proximity to urban light rail transit on 20double-story terraced property values in the Greater Kuala Lumpur, Malaysia, using hedonic pricing model. 21e ĕndings suggest there is a statistically positive and signiĕcant impact of urban rail transit on 22surrounding double-story terraced residential property values – estimated at a premium of 12.3% and 9.8% 23for properties located up to 1000 m and 1001 – 2000 m to the nearest light rail transit station respectively. 24e ĕndings of the study are useful for investors and developers, namely to build more houses near urban 25rail transit infrastructure since they lead to signiĕcant appreciation in value. 261. Introduction 27Introducing urban light rail transit is expected to increase 28residential property values that are located within a short 29distance to the nearest station. is is due to the fact that 30residential properties that are located within a short distance 31to the nearest station enjoy easy access to work and 32amenities, 33greater 34mobility 35options 36and 37reduced 38transportation costs (Forouhar, 2016). Hence, in most cases, 39greater access to the nearest urban light rail transit stations is 40valued positively by property buyers and as a result, 41residential property values are signiĕcantly higher. Over the 42past 50 years or so, there have been numerous studies 43conducted to establish a positive relationship between urban 44light rail transit and residential property values. But, a limited 45of them have been conducted in the developing countries. 46Hence, the main objective of this study is to investigate the 47impact of proximity to urban light rail transit on double-48story terraced property values in Greater Kuala Lumpur, 49Malaysia, using hedonic pricing model. Since the common 50way to ĕnance public transport infrastructure is from the 51government (federal or local governments), either through 52tax revenues or borrowing, established the positive 53relationship between urban light rail transit and residential 54property values should certainly be considered as a 55mechanism to inĘuence the public decision and rationalize 56public investment in public transport infrastructure. 57From a theoretical point of view, the bid-rent theory 58established by Alonso (1964) and Muth (1969) provides 59better equipment to understand the relationship between 60urban light rail transit and residential property values. e 61bid-rent theory suggests that improvement in the transport 62system that increases accessibility to work, shopping and 63recreational sites which reduces transport costs is reĘected in 64higher land value along the corridor rather (particularly 65around stations) than land outside the corridor. As a result, 66we should expect a higher bid-rent surface at transit stations. 67One of the earliest studies to establish the positive 68relationship between urban rail transit and residential 69property values was carried in Philadelphia, New Jersey by 70Boyce, Allen, Mudge, Slater, & Isserman (1972). e results 71indicate that statistically signiĕcant positive impact of urban 72rail transit on property values in which for every minute of 73savings to the city centre, the residential property values 74increased by USD$149 to $200. Since then, a numerous 75studies have been carried out in many cities to establish a 76positive relationship between urban rail transit and 77residential property values: Amsterdam, Rotterdam, and 78Enschede (Debrezion, Pels, & Rietveld, 2011), Beijing (Feng, 79Li, and Zhao, 2011; Gu & Guo, 2008), Charlotte (Yan, 80Delmelle, & Duncan, 2012), Dublin (Mayor, Lyons, Duffy, & 81Tol, 2012), Ottawa (Hewitt & Hewitt, 2012), Houston (Pan, 822019), Montreal (Dubé, ériault, & Des Rosiers, 2013), 83Phoenix (Golub, Guhathakurta, & Sollapuram, 2012), 84Greater Kuala Lumpur (Dziauddin, Alvanides, & Powe, 852013), Minnesota (Cao & Lou, 2017), Tehran (Forouhar, 862016), and Warsaw (Trojanek & Gluszak, 2017). 87In the North America, Yan et al. (2012) discover 88residential properties located within 402 metres and 1,609 89metres from Charlotte’s light rail transit stations increased in 90value. In Ottawa, Hewitt and Hewitt (2012) report the O-91Train stations has a statistically signiĕcant impact on the 92 ISSN 2354-9114 (online), ISSN 0024-9521 (print) 93Indonesian Journal of Geography Vol.53 , No. 1, 2021 (61 – 66) 94DOI: http://dx.doi.org/10.22146/ijg.46302 website: https://jurnal.ugm.ac.id/ijg 95©2021 Faculty of Geography UGM and e Indonesian Geographers Association 96Received: 2019-07-07 97Accepted: 2021-03-10 98 99 100 101 102Keywords: 103Urban area; 104urban light rail transit; 105residential property values; 106hedonic pricing model; 107Greater Kuala Lumpur. 108 109 110Correspondent email: 111faris@fsk.upsi.edu.my 112RESEARCH ARTICLE 113©2021 by the authors. Licensee Indonesian Journal of Geography, Indonesia. 114is article is an open access article distributed under the terms and conditions of the Creative Commons 115Attribution(CC BY NC) licensehttps://creativecommons.org/licenses/by-nc/4.0/. 116117118===== PAGE 3 (score 30, numeric-paren 0, table-words 3, var-words 21) =====11912063 121 122Indonesian Journal of Geography, Vol. 53, No.1, 2021 : 61 – 66 123Two dichotomous urban rail transit variables are included in 124the ĕnal regression model and are the focus of this study: (1) 125residential properties located up to 1000 m to the nearest 126LRT station (DISTLRT1), and (2) residential properties 127located within 1001 – 2000 m to the nearest LRT station 128(DISTLRT2). 129Other 130locational 131characteristics 132include 133proximity to city centre (CBD), 134hospital (HOSP), 135recreational park (PARK), primary school (PRIMARYSCH), 136high-performing 137secondary 138school 139(HP_SECONDARYSCH), forest (FOREST), industrial area 140(INDUSTRIAL), the Central Side (CENTRAL) and the South 141Side (SOUTH) of Kuala Lumpur. 142Table 1 shows the summary statistics of dependent and 143exploratory variables used to estimate the impact of urban 144light rail transit on residential property values in this study. 145Based on a sample of about 1,640, following observation can 146be made: (1) property sales values range from MYR 100,000 147(USD 31,152, with the FOREX rate at MYR 3.21 or USD 1.00 148in 2010) to nearly MYR 3 million (USD 934,579); (2) the 149average property sales value is MYR 463,907 (USD 144,519); 150(3) the average property has a Ęoor size of around 139 square 151metre or 1,500 square feet; and (4) there are units with as 152low as 55 square metre or 592 square feet to as large as 372 153square metre or 4,000 square feet. 154In regression analysis, it is common to expect the 155presence of multicollinearity between the exploratory 156variables. Since the exploratory variables may be highly 157correlated, Pearson’s correlation coefficient and variance 158inĘation factors (VIFs) of each exploratory variable are 159employed in this study. According to Orford (1999) and 160Neter et al. (1985), a Pearson’s correlation coefficient value 161of 0.8 and above and a variance inĘation factor value of 10 162and above is generally considered harmful collinearity. In 163this study, pairs of exploratory variables that produce a 164correlation coefficient value of 0.8 or higher and variance 165inĘation factor value of 10 or higher are removed from the 166ĕnal model. Another major concern with regression analysis 167is the existence of heteroscedasticity. e existence of 168heteroscedasticity in the ĕnal data set is tested by using the 169Park test and its results suggest there is no heteroscedasticity 170in the error variance. 171To capture the impact of urban rail transit on double-172story terraced residential property values, several different 173functional forms were explored. e ĕnal form of hedonic 174pricing model equation takes the following form: 175 176lnYi = β0 + β1DISTLRT1i + β2DISTLRT2i + β3FLOORAREAi + 177β4LOTRAREAi + β5BEDROOMSi + β6FREEHOLDi + 178β7SHOPCENTREi 179+ 180β8HOSPITALi 181+ 182β9PARKi 183+ 184β10PRIMARYSCHi 185+ 186β11HP_SECONDARYSCHi 187+ 188β12FORESTi + β13CBDi + β14INDUSTRIALi + β15CENTRALi + 189β16SOUTHi + εi 190 (2) 191 192where lnYi is natural logarithm of the sale price of a property 193in Ringgit; DISTLRT1 is a dichotomous variable for 194residential properties located within 1000 m to the nearest 195LRT station; DISTLRT2 is a dichotomous variable for 196residential properties located within 1001 – 2000 m to the 197nearest LRT station; FLOORAREA is the Ęoor area of 198property in square feet; LOTAREA is the land area of 199property in square feet; BEDROOMS is a dichotomous 200variable for property with four or more bedrooms; 201FREEHOLD is a dichotomous variable for property with 202freehold holding status. SHOPCENTRE, HOSPITAL, PARK, 203PRIMARYSCH, HP_SECONDARYSCH, CBD, FOREST and 204INDUSTRIAL are respective proximity to the nearest shopping 205centre, hospital, recreational park, primary school, high 206performance secondary school, central business district, 207forest and industrial area which are all measured in metre. 208Finally, CENTRAL and SOUTH are dichotomous variables 209associated with the properties located in the central and 210southern sides of Greater Kuala Lumpur, respectively. β are 211parameters to be estimated, and ε is the normally distributed 212error term. 213In the 214semi-log functional form, 215the 216interpretation of the estimated coefficients β associates to 217their proportional (multiplied by 100). 218 219Units 220Mean 221S.D. 222PRICE 223Ln Price (dependent variable) 224MYR 225MYR 226463,906.61 22712.91 228280,676.81 2290.52 230DISTLRT1 231DISTLRT2 232Dichotomous variable (0 or 1) 233Dichotomous variable (0 or 1) 2340.20 2350.38 2360.40 2370.49 238FLOOSIZE 239LOTSIZE 240BEDS 241FREEHOLD 242Square feet 243Square feet 244Dichotomous variable (0 or 1) 245Dichotomous variable (0 or 1) 2461,549.74 2471,692.62 2480.39 2490.67 250407.69 251755.76 2520.49 2530.47 254SHOPCENTRE 255HOSP 256PARK 257PRIMARYSCH 258HP_SECONDARYSCH 259CBD 260FOREST 261INDUSTRIAL 262CENTRAL 263SOUTH 264Metre 265Metre 266Metre 267Metre 268Metre 269Metre 270Metre 271Metre 272Dichotomous variable (0 or 1) 273Dichotomous variable (0 or 1) 2741,596.29 2754,411.42 2762,532.41 2771,049.57 2783,876.53 2798,509.49 2801,919.91 2813,749.00 2820.30 2830.21 2841,003.98 2851,927.68 2861,230.34 2871,053.64 2882,050.43 2892,633.02 2901,144.00 2912,272.79 2920.46 2930.41 294Table 1. Descriptive statistics of dependent and independent variables 295296297===== PAGE 4 (score 27, numeric-paren 0, table-words 2, var-words 22) =====29829964 300 3013. Results and Discussion 302e results of the estimation of Equation (2) are 303presented in Table 2. e model ĕts the cross-sectional data 304well and explains 62% of the variation in double-story 305terraced residential property values. All independent 306variables that inĘuence double-story terraced residential 307property values are signiĕcant at the 0.05 level and have the 308expected positive or negative signs. e estimated coefficients 309reveal that light rail transit has a positive and signiĕcant 310impact on surrounding double-story terraced residential 311property values – estimated at a premium of 12.3% and 9.8% 312for properties located up to 1000 m (DISTLRT1), and 1001 – 3132000 m (DISTLRT2) to the nearest light rail transit station 314respectively. e results support the Alonso-Muth-Mills-315based hypothesis that residential properties around urban rail 316transit stations capitalise in value due to potential savings in 317transports costs. e impact of other structural and locational 318attributes variables is also estimated. e results show for 319every square foot increase in the Ęoor size (FLOORSIZE) and 320lot size (LOTAREA), the double-story terraced residential 321property values are estimated at a premium of 0.04% and 3220.02% respectively. As expected, the effect of number of 323bedrooms (BEDROOMS) on residential property values is 324also positive – double-story terraced residential property 325units with four or more bedrooms are to sell at a premium of 3264.2%. Residential property units with a freehold ownership 327rights (FREEHOLD) sell at a premium of 18.8%. 328In terms of locational attribute variables, the estimated 329coefficients suggest that double-story terraced residential 330property values increase if properties are located in close 331proximity to the nearest shopping centre (SHOPCENTRE), 332proximity to the nearest recreational park (PARK), proximity 333to 334the 335nearest 336high-performance 337secondary 338school 339(HP_SECONDARYSCH), and proximity to the nearest forest 340(FOREST) – estimated at a premium of 0.008%, 0.003%, 3410.007% and 0.003% for every metre closer to these locational 342attributes respectively. e estimated coefficient further 343denotes the proximity to the city centre (CBD) negatively 344impacts residential property values – residential property 345units decrease in value by about 0.002% for every metre away 346from the city centre, which conĕrms the monocentric urban 347structure theory for Greater Kuala Lumpur. Results from 348estimation also suggest residential property units located at 349the Central Side (CENTRAL) and the South Side (SOUTH) 350of Kuala Lumpur which have the largest share of middle-351income, 352upper 353middle-income 354and 355upper-income 356households, command a premium of 30.3% and 20.1% 357respectively over similar double-story terraced residential 358properties located in other parts of the Greater Kuala 359Lumpur. 360e results also indicate that, for every metre closer to the 361nearest 362hospital 363(HOSPITAL), 364primary 365school 366(PRIMARYSCH), and industrial area (INDUSTRIAL), the 367double-story terraced residential property values are sell at 368discount of 0.01%, 0.007%, and 0.003% respectively. 369Overall, this study contributes to the growing literature 370on the impact of urban light rail transit on residential 371property values in three ways, (1) it investigates the impact of 372urban light rail transit on double-story terraced residential 373property values in the developing country, i.e., Malaysia; (2) 374the impact of the Greater Kuala Lumpur’s urban light rail is 375investigated by utilising the 2009 and 2010 residential 376property database, which allows the effects to be captured 377aer a reasonable time span, i.e., 12 years aer the opening of 378the ĕrst urban light rail transit in the country; and (3) the 379impact of the Greater Kuala Lumpur’s urban light rail is 380measured by employing two distance intervals namely up to 3811000 m and 1001 – 2000 m of the nearest LRT station. 382Regarding (2), Gu and Guo (2008) found that the impact 383of Beijing subway line 13 on suburban residential property 384values two years aer the opening was no longer signiĕcant. 385Another study by He and Jing (2013) on Beijing subway line 3864 found that the impact lasted for four years. In addition, 387Dziauddin et al. (2013) study the Kelana Jaya LRT line in the 388Klang Valley, Malaysia based on the data in 2004 which is six 389years aer the opening suggested that the impact was 390marginal. e ĕnding of present study, however, suggest that 391double-story terraced residential property values appreciated 392largely 12 years aer the opening. is ĕnding reĘects the 393longer time that property buyers place to the availability of 394urban light rail transit in the area. In car loving country such 395as Malaysia it is not surprised to expect property buyers take 396longer time to respond positively towards the existence of 397urban light rail transit especially when Malaysian have 398witnessed the hikes in fuel and parking rates (particularly in 399the Kuala Lumpur city centre). As a study conducted by Chiu 400Chuen, Karim, and Yusoff (2014) in Greater Kuala Lumpur 401denotes that factors such as parking costs, fuel prices and car 402prices may considerably inĘuence the likelihood of 403commuters using a car as their main mode of transport. 404Regarding (3), the largest positive impact tends to occur 405within 1000 m of the nearest LRT station. is ĕnding is 406consistent with previous studies (Dubé et al., 2013; Dziauddin 407et al., 2013; Feng et al., 2011) suggesting that residential 408property values that are close to stations are higher than those 409located further away. For example, studies carried out by 410Dubé et al. (2013) in Montreal, Canada, and Feng et al. (2011) 411in Beijing, China found that the positive impact of rail transit 412on residential property values diminished with an increase of 413distance from the station. 414 4154. Conclusion 416is study uses a hedonic pricing model to investigate the 417impact of proximity to light rail transit on residential 418property values based on a sample of double-story terraced 419properties in the Greater Kuala Lumpur, Malaysia. e 420results show that light rail transit has a signiĕcant positive 421impact on proximate double-story terraced property values 422up to a 1000 m and 1001 – 2000 m radius to the nearest light 423rail transit station. e ĕt of the equation is found to be 424satisfactory as an adjusted R2 higher than 60% for cross-425sectional data. e impact is the highest when double-story 426terraced properties are located up to a 1000 m radius to the 427nearest light rail station. It can be concluded that urban light 428rail transit plays a signiĕcant role in determining residential 429property values. e results of this study can provide some 430new insights for investors and developers to fully consider of 431building more houses near urban rail transit infrastructure 432since there is a signiĕcant value appreciation. e results of 433this study may also help home buyers select reasonably 434locations when buying a new house so that it will ensure the 435investment appreciates in value over time. 436While the results are robust in investigating the impact of 437urban light rail transit on residential property values, a 438limitation of this study lies it its modelling approach and 439temporal effects. e main disadvantage of OLS is that it has 440THE IMPACT OF URBAN LIGHT RAIL 441Mohd Faris Dziauddin 442