# RECORD 39f8b5119ab63b41236467364d52b44c282cb98c # TITLE: # YEAR: VENUE: DOI: # AUTHORS: # PAGES: 17 CANDIDATE TABLE PAGES (1-based): [12, 13] # ABSTRACT: ===== PAGE 1 (score 14, numeric-paren 0, table-words 0, var-words 14) ===== Impact of rail transit station proximity to commercial property prices: utilizing big data in urban real estate Mohammed Ali Berawi1*, Perdana Miraj2,3  , Gunawan Saroji3 and Mustika Sari3 Introduction Investment in rail transit does not only offer benefits to the public interest, such as increasing physical activity [1], releasing traffic congestion [2, 3], reducing emission [4], and expanding consumer amenities [5], but it also enables property owners nearby tran- sit stations to capitalize their assets for greater benefits due to the land value uplift as the effect of accessibility improvement. The increases in land value indicate the posi- tive impact of rail transit to the local economy, people’s quality of life, and the overall property market. The capitalization of real estate market is a combination of structural characteristics, surrounding environment, physical attributes, and accessibility which created the market price of an area [6, 7]. Abstract  Urban transport investments have contributed to the exponential increase of value from land and properties around the built infrastructure. Although literature had shown evidence of value uplift from the residential property due to transit develop- ment, little is known about the impact on commercial property values. This paper aims to evaluate the impact of rail transit proximity to the commercial property market by taking into account the pre-operation of the LRT project in Jakarta, Indonesia. This study utilizes a big data approach to accelerate research data processing by employing data mining method, as well as geographical information system (GIS) and hedonic price modeling (HPM) to investigate the property prices to formulate empirical evidence for the research. The results show opposite evidence compared to previous studies which argued that accessibility may contribute to the property price when closer to the transit station. The findings indicate that proximity to rail transit has an insignificant impact on the commercial property price compared to other variables such as building size, number of rooms, location, and hospitals. A comparison of the findings between this research with other studies was discussed and recommendation for further research development was proposed. Keywords:  Commercial property, Hedonic price model, Jakarta indonesia, Proximity premium, Rail transit, Value capture Open Access © The Author(s) 2020. This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat​iveco​ mmons​.org/licen​ses/by/4.0/. RESEARCH Berawi et al. J Big Data (2020) 7:71 https://doi.org/10.1186/s40537-020-00348-z *Correspondence: maberawi@eng.ui.ac.id 1 Department of Civil Engineering, Universitas Indonesia, Depok 16424, Indonesia Full list of author information is available at the end of the article ===== PAGE 12 (score 30, numeric-paren 6, table-words 8, var-words 5) ===== Page 12 of 17 Berawi et al. J Big Data (2020) 7:71 the two variables and a value close to 0 indicates a weak relationship. On the other hand, positive ( +) and negative (-) signs suggest the direction of the relationship between the two variables. If the value is + (positive), then the two variables have a direct relation- ship. The significance of the variable can be seen from Sig. (2-tailed) when the value lower than 0.05. Among three variables (distance, number of rooms, and building size), as shown in Table 5, the variable of building size shows a correlation coefficient close to 1 and sig. A 2-tailed value lower than 0.05. A positive mark means that a larger building size con- tributes to a higher property price. In terms of location, three of the stations, namely Rasuna Said, Dukuh Atas, and Cawang, are highly correlated and contribute to the prop- erty price, respectively. Subsequently, the result suggests excluding CBD and university as both of them do not significantly contribute to the property price. The HPM equation uses SPSS software by taking into account the stepwise regres- sion method. This method evaluated the input and produced four equation models as the best alternative in the case study. The best equation selected based on the high- est R-squared close to 1. It will represent the impact of independent variables on the dependent variable. Based on the result in Table 6, model 4 is the best model due to the highest R-squared with 0.919 and suggests a lower error value with 7.80%, among other variables. Following the correlation test, another round of tests is performed, including multi- collinearity and heteroscedasticity tests. The multi-collinearity test aims to determine inter-correlation between independent variables in a regression model. The model will proceed for further analysis when any correlation between independent variables was not found. This test was taking into account tolerance and VIF score, which should be higher than 0.10 and less than 10, respectively. The result shows that all four models (Table 6) do not correlate with independent variables. However, the fourth model which Table 5  Correlation test between dependent and independent variables * Correlation is significant at the 0.05 level (2-tailed) ** Correlation is significant at the 0.01 level (2-tailed) Ln Price Sig. (2-tailed) N Ln Distance − 0.091 0.380 95 Ln NRooms − 0.046 0.659 95 Ln BSize 0.951** 0.000 95 Karet Kuningan 0.026 0.805 95 Dukuh Atas 0.212* 0.039 95 Rasuna Said − 0.216* 0.035 95 Cawang − 0.202* 0.049 95 Kuningan 0.135 0.191 95 Cikoko − 0.008 0.936 95 Ciliwung 0.113 0.276 95 Schools − 0.212* 0.039 95 University − 0.044 0.675 95 Hospital 0.212* 0.039 95 CBD 0.048 0.646 95 ===== PAGE 13 (score 22, numeric-paren 0, table-words 3, var-words 7) ===== Page 13 of 17 Berawi et al. J Big Data (2020) 7:71 includes four variables (Building Size, Hospital, Rasuna Said, and Number of Rooms) shows the highest score among others. Based on the findings, the formula regarding the effect of proximity to transit station on property values in the case study is presented as follows. where, Ln Price = Log Natural of property price (US$). Ln BSize = Log Natural building size (square meter). Ln NRooms = Log Natural number of rooms (unit). Hospital = Location of hospital within 1 km (dummy). RasunaSaid = Property location within 1 km from Rasuna Said (dummy). Overall, two variables that consist of building size and the number of rooms signifi- cantly contribute to the property price around the transit station. A larger building is suit- able for commercials and offices, which requires large space to perform daily activities. The findings showed that expanding the building size every square meter will increase the property price by US$ 6.45. These findings also found two variables consist of hospi- tal and Rasuna Said—a location of transit station generates negative signs, thus create an opposite correlation. Hospital existence in a 1 km radius of transit station is believed to decrease property price about 6.8% or equal to US$ 7.29 while the property located 1 km from Rasuna Said station is cheaper US$ 8.21 than properties beyond the radius. Discussion Previous studies showed insightful evidence to create a basis comparison for the pre- sent studies. Despite all these findings, the property market on each country has differ- ent characteristics and conditions, causing direct comparisons rather difficult. Table 1 (2) LnPrice = 14.288 + 0.524 Ln_BSize −0.068 Hospital −0.077 RasunaSaid + 0.032 Ln_NRooms + %error Table 6  Result of multi-collinearity and heteroscedasticity tests Model Unstandardized coefficients Standardized coefficients t Sig Collinearity statistics R-squared B Std. error Beta Tolerance VIF 1 (Constant) 14.185 0.119 119.283 0.000 0.904 Ln BSize 0.531 0.018 0.951 29.823 0.000 1.000 1.000 2 (Constant) 14.295 0.120 119.017 0.000 0.912 Ln BSize 0.524 0.017 0.940 30.408 0.000 0.984 1.017 Hospital − 0.085 0.029 − 0.092 − 2.965 0.004 0.984 1.017 3 (Constant) 14.324 0.118 121.557 0.000 0.916 Ln BSize 0.520 0.017 0.932 30.737 0.000 0.973 1.028 Hospital − 0.069 0.029 − 0.074 − 2.394 0.019 0.929 1.077 RasunaSaid − 0.074 0.031 − 0.074 − 2.370 0.020 0.928 1.078 4 (Constant) 14.288 0.117 122.273 0.000 0.919 Ln BSize 0.524 0.017 0.939 31.364 0.000 0.962 1.040 Hospital − 0.068 0.028 − 0.073 − 2.403 0.018 0.929 1.077 RasunaSaid − 0.077 0.031 − 0.077 − 2.510 0.014 0.926 1.080 Ln NRooms 0.032 0.015 0.062 2.111 0.038 0.986 1.014 ===== PAGE 14 (score 18, numeric-paren 0, table-words 0, var-words 18) ===== Page 14 of 17 Berawi et al. J Big Data (2020) 7:71 illustrates selected studies regarding the impact of rail station on commercial proper- ties worldwide. Unlike previous studies that showed premium accessibility to rail transit, this research found the accessibility attribute to the transit station has an insignificant impact on the property price. In contrast, structural and neighborhood attributes play a critical role in increasing the price of property nearby rail transit. There are some reasons the variable of distance to transit was out of the equation. First, data mining was conducted starting from September 2018 to March 2019, where the transit stations are yet to operate. At this stage, owners believed transit stations offer less added-value when the users still require additional modes of transportation (bus or motorbike ride-hailing) to reach their property. Accessibility within walking distance is argued to remain unattractive due to various reasons. Citizens of Jakarta tend to have a negative experience when accessing destinations through pedestrian paths since those paths are still lacking in width outside the arterial roads; usually only 1 m wide, lack- ing in amenities, and lacking in the number of lightings. Moreover, poor condition of walkway pavements, street vendors occupying the sidewalks, and the excessive exposure from vehicle noise and pollution are also among other reasons. Therefore, it is under- standable from the viewpoint of property owners to not taking into account better acces- sibility and mobility as they believe that do not significantly correlate to property prices. Second, location with strong branding plays a significant role in determining prop- erty prices despite the far location from the rail transit station. Rasuna Said, which was included in the formula, is located in the Kuningan area of South Jakarta that provides a balanced lifestyle between working and leisure. This area comprises a variety of office buildings, hotels, embassy offices, and a variety of tourist attractions, including shop- ping malls, art galleries, public open space, and many others that can be regarded as a place to relax or refresh after a day of work. However, Rasuna Said station in this project is argued to be located in a bad environment where there is a slum area nearby, bad odor from the polluted river, and it is crime-prone due to the incomplete street ameni- ties. Therefore, it is not a surprise when the property in Rasuna Said has a higher price though it is outside of 1 km walkability radius compared to those located with proximity from the rail station. Conclusion There is a huge potential to utilize a value capture approach to finance rail transit pro- jects by fully understanding the positive impact of accessibility improvement due to the existence of transit stations on property values. Although capitalization of property mar- ket around transit stations has been extensively investigated in developed countries such as in European Countries, the United States, and the East Asian region, research in the same topics in developing countries such as Indonesia remains limited. This research elaborates properties listed in several real estate marketplaces and takes into account seven LRT stations located in Jakarta, Indonesia, that are planned to fully operate in 2021 to fill the research gaps. Most studies use geographical information system (GIS) and hedonic price mod- eling (HPM) by taking into account the contractual price of property market. For instance, Barbara et al. [13] elaborated property sales data in Australia and New Zea- land from a property business company. On the other hand, Mohammad et al. [17]