1# RECORD e4f33aca0968c7c930129cd6f3fbf32507e5ac892# TITLE: THE EFFECTS OF A SHOPPING MALL ON HOUSING PRICES: A CASE STUDY IN HANGZHOU3# YEAR: 2018.0 VENUE: International Journal of Strategic Property Management DOI: 10.3846/ijspm.2019.63604# AUTHORS: Ling Zhang; Jiantao Zhou; E. Hui; Hai-Zhen Wen5# PAGES: 16 CANDIDATE TABLE PAGES (1-based): [7, 8, 9, 11, 14, 15, 16]6# ABSTRACT: There are few studies on the externalities of shopping malls affecting the housing market. This study aims to discuss two issues: (1) What is the intensity of the impact of a shopping mall? (2) When does the external influence of a shopping mall begin to reveal itself? The West Intime Shopping Mall in Hangzhou offers a unique situation to research the questions. By dividing the study area into nine blocks, using hedonic price theory, and the price gradient approach with housing price data from 2011 to 2015, we found that in the space dimension, the mall exerted a significantly positive effect on the housing prices of nearby blocks. With the increase in distance from the mall, the positive effect decreased. There were more significantly positive effects in blocks far away from the city center. In the time dimension, the effects of West Intime did not reveal themselves until the mall had started to operate and gradually matured over time, implying that the mall did not have the obvious expected impact on housing prices before the mall had begun operating.789===== PAGE 1 (score 11, numeric-paren 0, table-words 0, var-words 11) =====1011Copyright © 2019 The Author(s). Published by VGTU Press12This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unre-13stricted use, distribution, and reproduction in any medium, provided the original author and source are credited.14*Corresponding author. E-mail: wenhaizhen@263.net15International Journal of Strategic Property Management16ISSN: 1648-715X / eISSN: 1648-9179172019 Volume 23 Issue 1: 65–8018https://doi.org/10.3846/ijspm.2019.636019THE EFFECTS OF A SHOPPING MALL ON HOUSING PRICES: 20A CASE STUDY IN HANGZHOU21Ling ZHANG1, Jiantao ZHOU1, Eddie C. M. HUI2, Haizhen WEN1,*221 Center for Real Estate Studying, Department of Civil Engineering, Zhejiang University, China232 Department of Building and Real Estate, The Hong Kong Polytechnic University, Hung Hom, 24Kowloon, Hong Kong, China25Received 4 December 2017; accepted 30 August 201826Abstract. There are few studies on the externalities of shopping malls affecting the housing market. This study aims to 27discuss two issues: (1) What is the intensity of the impact of a shopping mall? (2) When does the external influence of a 28shopping mall begin to reveal itself? The West Intime Shopping Mall in Hangzhou offers a unique situation to research the 29questions. By dividing the study area into nine blocks, using hedonic price theory, and the price gradient approach with 30housing price data from 2011 to 2015, we found that in the space dimension, the mall exerted a significantly positive effect 31on the housing prices of nearby blocks. With the increase in distance from the mall, the positive effect decreased. There 32were more significantly positive effects in blocks far away from the city center. In the time dimension, the effects of West 33Intime did not reveal themselves until the mall had started to operate and gradually matured over time, implying that the 34mall did not have the obvious expected impact on housing prices before the mall had begun operating.35Keywords: shopping mall, housing price, price index, price gradient, hedonic models.36Introduction37With the suburbanization and economic growth occur-38ring since the 1960s in the United States, shopping centers 39and malls have been spreading throughout cities, first in 40the suburbs, then in the downtown areas (Carter, 2009). 41Shopping malls have dramatically reshaped the retailing 42industry, providing the impetus for regional economic de-43velopment and the improvement of living standards. Now, 44shopping malls are major centers of retail activity in the 45United States and around the world (Larsen, Shelton, & 46Wright, 2015). In recent years, alongside urban develop-47ment in China, shopping malls have grown quickly with 48urbanization, versatility, and convenience in transporta-49tion. Most malls deliver convenience to local residents and 50have strong influences on neighboring areas. Large-scale 51shopping malls have become attractive places to shop, vis-52it, and spend time. Most malls have also increased their 53leisure activity components (Fasli, Riza, & Erbilen, 2016).54There have been many types of studies focusing on 55the designs, functions, rents, and satisfactoriness of shop-56ping malls (Carter, 2009). Some studies have discussed the 57role of shopping centers in societies and urban spaces, but 58most have not covered the externalities of shopping cent-59ers. Shopping malls can deliver commercial convenience 60to the surrounding residences, indirectly increasing the 61values of these properties in the housing market. There-62fore, it makes sense to explore the effects of shopping 63malls on housing prices.64Compared to infrastructure, schools, and other public 65facilities, shopping malls are so complex that they may 66possess distinctive externalities. Sirpal (1994) and Des Ro-67siers, Lagana, Thériault, and Beaudoin (1996) discovered 68positive and negative effects at different distances from 69shopping malls. The externalities are directly related to the 70internalities, such as the scale, design and anchor tenants, 71of a mall. Moreover, the influence of a mall may change 72at different stages over time. Promotional advertising be-73fore a mall’s construction may fuel a rise in housing prices 74while noise from the construction works could worsen the 75living conditions of residents and lower the housing pric-76es. After a mall’s completion, its external influence usually 77differs with the maturity of its operation and the establish-78ment of word of mouth. Such externalities, particularly in 79terms of their scope and times of impacts on the housing 808182===== PAGE 7 (score 24, numeric-paren 0, table-words 3, var-words 8) =====8384International Journal of Strategic Property Management, 2019, 23(1): 65–80857186ral environment, living facilities, educational facilities, and 87sports facilities. In China, parents pay much attention to 88children’s education. The school district significantly influ-89ences housing prices (Feng & Lu, 2013; Wen, Y. Zhang, & 90L. Zhang, 2014b), so this paper introduces the quality of 91schools and the distances to schools to characterize the 92impact of education and to measure school quality in the 93same way as did the research of Wen et al. (2014b) on 94Hangzhou. Commercial facilities also have a certain de-95gree of impact on housing prices. Small commercial facili-96ties can be considered together with living facilities due to 97their small influence while the impact of shopping malls 98is examined separately.99The sample data are the housing transaction prices from 100the beginning of 2011 to the end of 2015, which were ob-101tained from a real estate agency in Hangzhou. Some rel-102evant information, such as living facilities, sports facilities, 103and bus route, was not provided by the agency but obtained 104through a field survey conducted in 2010, 2012 and 2014. 105Our institute has conducted follow-up surveys on the resi-106dential buildings in Hangzhou every two to three years since 1072008 to score the neighborhood and location characteristics 108of the community. A questionnaire has been designed to 109measure variables of interior environment, surrounding en-110vironment, sports facility, property management, quality of 111primary schools, and quality of junior high schools.112Mainly two aspects are considered in the choice of the 113research period. First, the real estate around West Intime 114developed later than that in the main city area. Although 115the average price of second-hand houses in the study area 116increased from 2008 to 2011, the housing transactions be-117fore 2011 of Zones 6, 7, 8 and 9 are absent, so the pre-118construction period cannot form a price gradient contrast. 119Second, data on the marketing activities of the mall before 120construction are scarce. In addition, there was little adver-121tising about West Intime before the construction. Therefore, 122we chose observations after 2011 for the sake of the equi-123librium of sample spatial distribution. To ensure the quality 124and quantity of the data, we have adopted the price data of 125the second-hand houses rather than the newly-built hous-126es. According to the statistical data, from 2011 to 2015 the 127transaction volume of newly-built houses was about 10,000 128units in study area, which is very close to the volume of the 129second-hand houses. However, these newly-built units are 130concentrated in 16 communities and mostly distributed in 131the area north of Zone 6. There are few newly-built hous-132es in Zones 1 to 5. The trading units of the second-hand 133houses are evenly distributed in 71 communities, which can 134better reflect the housing prices in the study area. To distin-135guish the changing impacts of the different stages, the data 136are divided into two periods: the construction period (P1: 137January 2011 – August 2013) and the operating period (P2: 138September 2013 – December 2015). The number of second-139hand transactions in this area was 838 in 2011, which dou-140bled in 2012 to 1,659, and continued to grow to 1,943 in 1412013. The increasing volume of second-hand transactions 142and the significant growth in the construction of newly- 143built houses north of West Intime reflect the signal effect 144of shopping malls on the housing market. The descriptive 145statistics of the variables are shown in Table 4.146Table 4. Descriptive statistics147Period 1148(January 2011 – August 2013, N = 3939)149Period 2150(September 2013 – December 2015, N = 5315)151Min152Max153Mean154Std.dev.155Min156Max157Mean158Std.dev.159Sale price (¥/m2)1608228.0016137623.1716220374.561634214.881648203.4516540851.2316620195.141674613.46168Size (m2)16923.82170264.6017190.0917231.10517323.71174330.9817594.7917631.259177Floor/Level No.1781179271806.211815.1031821183271846.441855.136186Total floors18741882918911.561907.44519141922919312.141947.516195Building’s age (year)19611972519811.421996.44620032012720212.692035.834204Distance to CBD (km)2054.3920610.792076.602081.6722094.3921010.792116.692121.626213Bus route21422153621620.152177.39021822193622019.572217.146222Distance to West Intime (km)2230.512246.502253.532261.2692270.512286.502293.482301.300231Distance to Hangzhou Incity (km)2320.462337.402343.002351.3462360.462377.402382.962391.277240Distance to Xicheng Square (km)2410.652427.802433.242441.9332450.652467.802473.272481.809249External environment250125152523.142530.670254125552563.242570.666258Inner environment259126052613.362620.861263126452653.532660.859267Living facilities268126952704.052711.251272127352743.962751.240276Sports facilities277127852792.662801.081281128252832.812841.092285Property management286128752883.032890.852290129152923.182930.796294Quality of primary school295129632972.022980.847299130033011.983020.862303Quality of junior high school304130523061.693070.462308130923101.643110.480312Distance to primary school (km)3130.103142.063150.593160.3693170.103182.063190.593200.366321Distance to junior high school (km)3220.193233.283241.423250.8383260.193273.283281.453290.841330331332===== PAGE 8 (score 103, numeric-paren 57, table-words 14, var-words 5) =====33333472335L. Zhang et al. The effects of a shopping mall on housing prices: a case study in Hangzhou3364. Analysis of results337The regression of the three models produced good results, 338which are discussed below.3394.1. The results of simple hedonic price modelling 340approaches341To estimate the impact of West Intime on the surrounding 342housing prices, we take advantage of the road distance to 343the mall to capture the price elasticity of the surround-344ing housing. The SPSS was used to estimate Equation (1) 345and analyze the housing characteristics. The results of the 346regression model are shown in Table 5.347Economic theory places few restrictions on the form 348of the hedonic price function (Hui et al., 2007; Feng & Lu, 3492013). By comparing the regression results of commonly 350adopted forms such as linear, semi-logarithmic (as shown 351in Table A2) and logarithmic, we found that logarithmic 352forms fit slightly better than the others. The correlation 353between independents was smaller and coefficient inter-354pretation was more significant. The adjusted R2 of Model 3551 is 0.642, which means that the independent variables 356could explain 64.2% of the dependent variable. Almost 357all the 21 variables are significant at the 1% level. To test 358for the spatial independence of the estimated error terms, 359referring to Fik et al. (2003), the residuals were regressed 360on the independent variables and location coordinates x, 361y and their polynomial expression. The coefficient esti-362mates of distance independent variables are all insignifi-363cant. While the location variable x2 is significant in Model 3641, there is no significant location variable in Model 2 or 365Model 3, in which the spatial dummy variables are in-366cluded (see Table 5). Due to the lack of location coordi-367nates for housing units, and in order to solve the possible 368spatial correlation problem, we estimated the coordinates 369for house units by random calculations with GIS, based 370on the coordinates and shapes of the communities. Then, 371we conducted spatial econometric models using the con-372structed spatial weights as a robustness test. The results are 373shown in Appendix 2.374The effect of the structure characteristics all obtain 375the expected results. The coefficient of Total floors is 376Table 5. Results of regression of the three models377Independent variables378Model 1379Model 2380Model 3381Coefficient382T – statistics383Coefficient384T – statistics385Coefficient386T – statistics387Constant38810.431***389231.55439010.947***391140.91239210.961***393141.590394Ln Size395–0.093***396–19.351397–0.112***398–24.940399–0.115***400–25.243401Ln Building’s age4020.017***4033.3264040.011***4055.1664060.012***4075.332408Ln Floor/Level No.4090.010***4104.000411–0.073***412–15.675413–0.074***414–15.837415Ln. Total floors416–0.033***417–7.127418–0.047***419–8.464420–0.046***421–8.473422Ln Distance to CBD423–0.235***424–17.770425–0.338***426–11.650427–0.342***428–11.665429Ln Distance to West Intime430–0.109***431–20.037432/433/434/435/436Ln Distance to Xicheng Square437–0.048***438–9.958439–0.039***440–4.025441–0.039***442–3.947443Bus route444–0.002***445–3.998446–0.004***447–7.507448–0.003***449–7.344450External environment4510.048***45211.4844530.070***45415.5414550.069***45615.134457Inner environment458–0.022***459–6.345460–0.031***461–7.815462–0.029***463–7.247464Living facilities465–0.007***466–3.089467–0.031***468–12.806469–0.031***470–12.824471Sports facilities4720.006**4732.5874740.022***4759.6954760.022***4779.451478Property management4790.084***48019.0284810.071***48216.6704830.072***48416.702485Quality of primary school4860.084***48734.3414880.063***48921.1084900.063***49120.962492Quality of junior high school4930.041***4945.7254950.0054960.3864970.0054980.340499Distance to primary school500–0.032***501–9.596502–0.028***503–7.140504–0.028***505–7.027506Distance to junior high school5070.033***50811.329509–0.047***510–10.440511–0.046***512–10.061513Y2514–0.071***515–12.109516/517/518–0.059***519–10.879520Y3521–0.007522–1.189523/524/5250.0045260.715527Y4528–0.055***529–8.861530/531/532–0.057***533–8.248534Y5535–0.059***536–9.797537/538/539–0.054***540–7.991541Adjusted R25420.6425430.7015440.696545Residuals regressed on location 546variables547x2 –0.002*** –2.111548x*y –0.001 –1.515549x2 0.000 –1.085550x*y 0.000 –1.272551x2 0.000 –0.952552x*y 0.000 –1.147553Notes: ***, **, and * represent significance at the 1%, 5% and 10% levels, respectively. The dependent variable is lnP. To test for the spatial independence 554of the estimated error terms, the residuals were regressed on the location variables, x, y coordinate and x2, y2, x*y.555556557===== PAGE 9 (score 65, numeric-paren 39, table-words 7, var-words 7) =====558559International Journal of Strategic Property Management, 2019, 23(1): 65–8056073561–0.033, which is significant and negative. This result is 562consistent with the reality that sub-high-rise buildings 563with lower heights are more expensive due to less shar-564ing of public areas and to more comfort. The coefficient 565of Distance to CBD is –0.235, which means that hous-566ing prices will decrease by 0.235% if the distance from 567a house to the city center increases by 1%, indicating 568that Wulin Square definitely has much influence on the 569housing market. The quality of primary and junior high 570schools nearby can significantly increase housing prices 571and Distance to primary school has significant negative 572effect. Contrary to expectations, the coefficient of Bus 573route is negative, this phenomenon may be attributed to 574the fact that increasing the number of bus routes may 575produce negative effects, such as noise and environmen-576tal pollution (Wen, Xiao, & Zhang, 2017). As for Neigh-577borhood Characteristics, neither Inner environment nor 578Living facilities is as expected, which may be because the 579sample is concentrated in a limited area. The variables 580measured in the same way did not perform well either in 581housing submarket around the Grand Canal (Wen et al., 5822017), but they were positively significant in house price 583regression with the sample of entire city (Wen et al., 5842014a). We should improve the measure of these vari-585ables in future studies.586The distances to the shopping malls are the main re-587search objects. West Intime and Hangzhou Incity are too 588close to each other, so there would be collinear problems 589if the two distance variables were both introduced into 590the model. Considering the smaller influence of Hang-591zhou Incity, Distance to Hangzhou Incity is not introduced 592into the model. The coefficients of Distance to West Intime 593and Distance to Xicheng Square are both significant and 594negative as were expected, but Distance to West Intime is 595a little larger with a value of –0.109. In the case of ex-596cluding the impacts of other characteristic variables, as 597the road distance to West Intime increases 1%, housing 598prices decrease by 0.109%. The result indicates that for the 599study area, West Intime has a significant impact on hous-600ing prices and is more influential than Xicheng Square. In 601this model, the influence of West Intime on housing prices 602is characterized initially and roughly. The following two 603models provide further analysis.6044.2. The results of improved hedonic price modeling 605approaches606Similarly, the least-square method was used to estimate 607Model 2. The adjusted R2 increases to 0.701, indicating 608that compared to Model 1, this model is more explanatory. 609Almost all the characteristic variables are significant at the 6101% level. The coefficient of Distance to Xicheng Square is 611also significant and similar to the result of Model 1. By 612controlling the influence of Xicheng Square, the impact of 613West Intime on housing prices can be captured by spatio-614temporal interactions, whose coefficients are shown in 615Table 6. There are 44 spatio-temporal interaction coeffi-616cients, of which 39 are significant.617In the space dimension, the housing price indices of 618the five zones in the southwestern region (Zones 1–5) are 619relatively higher than the four zones in the northwestern 620region (Zones 6–9).621With the purpose of observing the influence of West 622Intime on the different zones in more detail, we further 623calculated the price gradients among the various zones 624and observed the changes. This helped to analyze the 625temporal and spatial influences of West Intime. Zone 1 626is the farthest south, and so, is farthest away from West 627Intime and closest to the city center. The impact received 628from West Intime is relatively minimal. Therefore, Zone 6291 was set as the reference area and the price indices of 630the other zones were subtracted from the price gradient 631of each zone. For each zone, the price gradient of Zone i 632(i = 2, …, 9) is equal to the difference between the price 633indices of Zone i and Zone 1. The results are shown in 634Figures 2 and 3.635Figure 2 shows the price gradients for Zones 2–5 based 636on Zone 1. The results of 2011–2013 represent the price 637gradients of Period One while the results of 2014–2015 rep-638resent Period Two. The results are summarized as follows:6391) In Period One, the price gradients of the four zones 640were basically stabilized with few changes. Even in 6412012, which was the peak of the construction pe-642riod, the price gradient of Zone 5 was reduced by 6430.027. The noise generated by the project construc-644tion might have had a negative impact on housing 645prices (Henneberry, 1998).646Table 6. The value of the coefficient 647,i j648σ649 (Model 2)650Z1651Z2652Z3653Z4654Z5655Z6656Z7657Z8658Z9659Y1660/6610.042***6620.121***6630.167***6640.121***665–0.108***666–0.0396670.105***668–0.082***669Y2670–0.043***6710.0026720.069***6730.126***6740.051**675–0.209***676–0.197***6770.025678–0.148***679Y36800.0096810.078***6820.125***6830.200***6840.126***685–0.132***686–0.103***6870.068**688–0.063**689Y4690–0.048***6910.027**6920.099***6930.164***6940.145***695–0.141***696–0.131***697–0.058**698–0.070**699Y5700–0.071***7010.0067020.119***7030.159***7040.217***705–0.083***706–0.127***707–0.071***708–0.069**709Note: ***, **, and* represent significance at the 1%, 5% and 10% levels, respectively.710711712===== PAGE 10 (score 12, numeric-paren 0, table-words 2, var-words 7) =====71371474715L. Zhang et al. The effects of a shopping mall on housing prices: a case study in Hangzhou7162) In Period Two, the price growth rate of Zone 5 was 717very significant. Zones 3 and 4, which were slightly 718farther away, were also affected by the positive effect 719of the shopping mall, leading to the increase in the 720price gradients based on Zone 1. The price gradi-721ent of Zone 2 increased slightly compared with the 722price gradients of the other zones. The results show 723that a significant effect on the residential prices in 724the south area after the open of the mall and the 725“value-added” range decreased with the distance.726Zones 6–9 are located away from the city center and 727West Intime is the only large-scale commercial facility in 728the surrounding area. So, theoretically, these four zones are 729better able to reflect the spatio-temporal impact of West In-730time on the surrounding housing prices. Figure 3 shows the 731price gradients of the four zones based on Zone 1:7321) In Period One, the price gradients of the four zones 733fluctuated less. Zones 6 and 7 had a certain degree of 734price decline in 2012 and rebounding in 2013, largely 735due to the impact of the neighboring construction.7362) In Period Two, the price gradient of Zone 6 in-737creased significantly and the value was near zero in 7382015, showing that the price of Zone 6 had begun 739to catch up with Zone 1. Similarly to Zone 6, the 740price gradient of Zone 7 also showed a clear upward 741trend with a smaller rise. The changes in the price 742gradient of Zone 8 was confusing while there were 743still some visible positive changes in Zone 9.744In this model, the value-added effect can be clearly 745seen, especially in Zones 6–9, and the influence decreases 746from Zone 5 to the other zones on both sides, but this 747method of price gradient calculation still could not elimi-748nate the time fixed effect completely.7494.3. The results of extended price gradient 750approaches751The explanatory power of Model 3 is 69.6%, which is very 752close to that of Model 2. The confidence levels of the char-753acteristic variables are almost the same as before. Only one 754spatial-temporal interaction coefficient θ was not signifi-755cant. Therefore, the model can explain the existing data 756well. Then, the price gradient was calculated by λi + θi,p. 757The results are shown in Table 7.7580.0007590.0507600.1007610.1507620.2007630.2507640.3007650.3507662 0 1 17672 0 1 27682 0 1 37692 0 1 47702 0 1 5771σi,j-σ1,j772Year773Z2-Z1774Z3-Z1775Z4-Z1776Z5-Z1777Figure 2. Price gradients of Zones 2–5 based on Zone 1778–0.200779–0.1007800.0007810.1007820.20 07830.3007840.4007852 0 1 17862 0 1 27872 0 1 37882 0 1 47892 0 1 5790σi,j-σ1,j791Year792Z5-Z1793Z6-Z1794Z7-Z1795Z8-Z1796Z9-Z1797Figure 3. Price gradients of Zones 6–9 based on Zone 1798Note: As a reference, the price gradient for Zone 5 is also shown.799800801===== PAGE 11 (score 28, numeric-paren 15, table-words 1, var-words 6) =====802803International Journal of Strategic Property Management, 2019, 23(1): 65–8080475805Figure 4 exhibits the price gradients of nine zones in 806two periods. Zone 1 is still set as the benchmark. In the 807premise of eliminating the time fixed effect and assuming 808that the housing prices in Zone 1 had not changed during 809the construction and operating periods, the price gradi-810ents of the other zones were obtained based on Zone 1.811Compared to the previous two methods, this treatment 812was better able to remove the influence of other factors on 813the time dimension. The findings are as follows:8141) The price gradients of Period 2 are shown to be sig-815nificantly higher than those of Period 1 (except for 816Zone 8). In the area south of West Intime (from 817Z2–Z5), the price gradient differences between the 818two periods are 0.006, 0.043, 0.031, and 0.127, re-819spectively. Z3 and Z4 almost have the same value 820added effect caused by West Intime while Z2 has 821the smallest appreciation. Zone 5 has become the 822new peak in housing prices and the price increases 82313.5% in Period 2. The value-added effect can be 824found to decrease with distance, but it is not a com-825pletely linear relationship.8262) In the north area of West Intime (from Z6–Z9), the 827price gradient differences between the two periods 828are 0.090, 0.029, –0.093, and 0.057, respectively. 829Generally the price gradient difference between the 830two periods in each zone is negatively correlated 831with the distance from West Intime. The price gra-832dient difference of Zone 6 is the largest while Zone 8337 is the smallest except the negative appreciation in 834Zone 8. The time fixed effect has largely been elimi-835nated, so the price gradient difference is caused by 836the influence of West Intime.837The above three models can be good proof that West 838Intime has had a positive impact on the surrounding hous-839ing prices. Model 1 identifies the impact of West Intime 840on housing prices as being inversely proportional to the 841distance. The two price gradient models allow insight into 842the spatial price trend of West Intime’s surrounding area. 843We found that the impacts on the surrounding housing 844prices in Zones 5 and 6 were the largest. In contrast, the 845impacts on Zones 2 and 8 were relatively weak. In terms 846of time, examining the housing prices in the two periods 847shows that the increase in housing prices during the oper-848ating period was very large while during the construction 849period, housing prices changed little and even appeared 850to have experienced a certain degree of negative effects.851Conclusions852Using the price gradient method and hedonic price the-853ory, this study analyzed the influence of a shopping mall 854on the surrounding housing prices in terms of time and 855space. First, the results of the hedonic model show that 856West Intime has a significant impact on housing prices, 857which decays with distance. Then, in the price gradient 858method, location and time dummy variables were intro-859duced to capture changes in housing prices at different 860times in various locations. The results indicate that almost 861all the zones in the study area have had their prices influ-862enced to some extent if there was a shopping mall near-863by. Especially in the zones closely surrounding the mall, 864housing prices had obviously increased by more than 10 865percent upon the mall’s opening. The spatial and tempo-866ral interaction terms used in Model 2 and Model 3 can 867Table 7. The value of λi + θi,p (Model 3)868Z1869Z2870Z3871Z4872Z5873Z6874Z7875Z8876Z9877li878/8790.063***8800.120***8810.183***8820.112***883–0.137***884–0.105***8850.087***886–0.078***887θi,p888/8890.0068900.043***8910.031***8920.127***8930.090***8940.029***895–0.093***8960.057***897T1(λi + θi,p)8980.0008990.0639000.1209010.1839020.112903–0.137904–0.1059050.087906–0.078907T2(λi + θi,p)9080.0009090.0689100.1649110.2149120.239913–0.047914–0.076915–0.005916–0.022917Note: ***, **, and* represent significance at the 1%, 5% and 10% levels, respectively.918–0.20919–0.15920–0.10921–0.059220.009230.059240.109250.159260.209270.259280.30929Z1930Z2931Z3932Z4933Z5934Z6935Z7936Z8937Z9938λi+ θi,p939T1940T2941Figure 4. Price gradients of Zones 1–9942943944===== PAGE 12 (score 14, numeric-paren 0, table-words 0, var-words 13) =====94594676947L. Zhang et al. The effects of a shopping mall on housing prices: a case study in Hangzhou948effectively separate the impact of the shopping mall on 949the surrounding house prices during the construction and 950operation phases. Through the block division based on the 951main roads, we can clearly observe the reaction of residen-952tial groups at different locations relative to the shopping 953mall, which is rarely discussed in the previous literature.954To the best of our knowledge, this paper is one of the 955few studies concerning the impacts of a shopping mall at 956different periods of time. Before the completion of the 957construction, the expected effect of West Intime is very 958small. The surrounding housing prices were subject to 959minimal influence, which is different from the expectation 960effect found in some subway or transportation studies. The 961reason may be that the influence of a shopping mall large-962ly depends on its own anchor tenants and operational ca-963pacity. Before the completion of the construction of West 964Intime, people knew little about the mall, because it had 965not been promoted much. Therefore, there was no obvious 966influence on the housing price. Nevertheless, during the 967construction of the mall, the transaction of second-hand 968houses and the construction of new houses in the affect-969ed area increased significantly, showing the signal effect 970of the shopping mall. Upon completion in Period 2, the 971mall’s good operating conditions and reputation greatly 972enhanced its radiative influence in the area.973This paper provides some implications for city plan-974ners and governments. The model has evaluated the 975range of a mall’s influence on housing prices and offers a 976good reference for avoiding the imbalanced distribution 977of shopping malls in different districts and the scarcity 978of retail shops. Meanwhile, the effects of shopping malls 979on housing prices can be taken into consideration for 980land use planning by the government. Due to the vital 981influence of shopping malls on the surrounding housing 982prices, the introduction of malls and commercial facili-983ties is a good practice to leading the development of a 984new city area.985The limitations of this study include the short span 986of the data. We cannot study the impact of the shopping 987mall on housing prices before construction. Therefore, it 988is impossible to explore if the shopping mall had had an 989effect before its construction. The data only covers the two 990years after the opening of the mall, so we cannot observe 991changes in housing prices for any longer term. Neverthe-992less, due to the brand effect and good management, the 993mall performed well and entered a stable operating period 994during the first two years. Further research is needed to 995explore how the impact pattern differs at different scales 996and operational levels of shopping malls, and to observe 997the interaction effect of two or more shopping malls on 998housing market.999Acknowledgements1000The authors would like to thank the editor and the anon-1001ymous reviewers for the positive and constructive com-1002ments and suggestions.1003Funding1004This work was supported by the China Scholarship Coun-1005cil under Grant No. 201706325013; and Zhejiang Provin-1006cial Natural Science Foundation of China under Grant 1007No. LY18G030002.1008References1009Ahlfeldt, G. M., & Wendland, N. (2011). Fifty years of urban 1010accessibility: the impact of the urban railway network on the 1011land gradient in Berlin 1890–1936. Regional Science and Ur-1012ban Economics, 41(2), 77-88. 1013https://doi.org/10.1016/j.regsciurbeco.2010.10.0011014Alonso, W. (1964). Location and land use: toward a general theory 1015of land rent. Cambridge: Harvard University Press. 1016https://doi.org/10.4159/harvard.97806747308541017Berry, B. J., & Garrison, W. L. (1958). Recent developments of 1018central place theory. Papers in Regional Science, 4(1), 107-120. 1019https://doi.org/10.1111/j.1435-5597.1958.tb01625.x1020Brueckner, J. K. (1993). Inter-store externalities and space alloca-1021tion in shopping centers. The Journal of Real Estate Finance 1022and Economics, 7(1), 5-16. 1023https://doi.org/10.1007/BF010969321024Carter, C. (2009). Review articles: what we know about shopping 1025centers. Journal of Real Estate Literature, 17(2), 163-180.1026Carter, C. C., & Haloupek, W. J. (2002). Dispersion of stores of 1027the same type in shopping malls: theory and preliminary evi-1028dence. Journal of Property Research, 19(4), 291-311. 1029https://doi.org/10.1080/09599910220000135501030Chau, K. W., & Chin, T. L. (2003). A critical review of literature 1031on the hedonic price model. International Journal for Housing 1032Science and Its Applications, 27(2), 145-165.1033Chau, K. W., & Ng, F. F. (1998). The effects of improvement in public 1034transportation capacity on residential price gradient in Hong 1035Kong. Journal of Property Valuation and Investment, 16(4), 1036397-410. https://doi.org/10.1108/146357898102282041037Christaller, W. (1933). Die zentralen Orte in Süddeutschland: eine 1038ökonomisch-geographische Untersuchung über die Gesetzmäs-1039sigkeit der Verbreitung und Entwicklung der Siedlungen mit 1040städtischen Funktionen. University Microfilms.1041Clark, D. E., & Herrin, W. E. (2000). The impact of public school 1042attributes on home sale prices in California. Growth and 1043Change, 31(3), 385-407. 1044https://doi.org/10.1111/0017-4815.001341045Clark, W. A. (1968). Consumer travel patterns and the concept 1046of range 1. Annals of the Association of American Geogra-1047phers, 58(2), 386-396. 1048https://doi.org/10.1111/j.1467-8306.1968.tb00651.x1049Colwell, P. F., Gujral, S. S., & Coley, C. (1985). The impact of a 1050shopping center on the value of surrounding properties. Real 1051Estate Issues, 10(1), 35-39.1052Coulson, N. E. (1991). Really useful tests of the monocentric 1053model. Land Economics, 67(3), 299-307. 1054https://doi.org/10.2307/31464251055De Palma, A., Ginsburgh, V., Papageorgiou, Y. Y., & Thisse, J. F. 1056(1985). The principle of minimum differentiation holds under 1057sufficient heterogeneity. Econometrica: Journal of the Econo-1058metric Society, 767-781. https://doi.org/10.2307/19126531059Des Rosiers, F., Lagana, A., Thériault, M., & Beaudoin, M. (1996). 1060Shopping centres and house values: an empirical investiga-1061tion. Journal of Property Valuation and Investment, 14(4), 41-62. 1062https://doi.org/10.1108/14635789610153461106310641065===== PAGE 14 (score 86, numeric-paren 48, table-words 8, var-words 10) =====10661067781068L. Zhang et al. The effects of a shopping mall on housing prices: a case study in Hangzhou1069Wen, H., Zhang, Y., & Zhang, L. (2014b). Do educational facili-1070ties affect housing price? An empirical study in Hangzhou, 1071China. Habitat International, 42, 155-163. 1072https://doi.org/10.1016/j.habitatint.2013.12.0041073Yiu, C., & Tam, C. (2004). A review of recent empirical stud-1074ies on property price gradients. Journal of Real Estate Litera-1075ture, 12(3), 305-322.1076Appendix 11077With the purpose of testing the robustness of the em-1078pirical results, this study used the traditional partition-1079ing method with distance rings to verify whether the re-1080sults are consistent. The study area was divided into eight 1081zones (0−1 km, 1−1.5 km, 1.5−2 km, 2−2.5 km, 2.5−3 km, 10823−3.5 km, 3.5−4 km, and 4−5 km) according to their dis-1083tance to West Intime, as shown in Figure A1. Owing to 1084the lack of communities, Zones 1 and 8 span 1 km while 1085other zones span 0.5 km. In the east, we set a boundary 1086line along Xueyuan Road, which is located in the middle 1087line of West Intime and the CBD. The area to the east is 1088closer to the city center.1089The three models were applied again. The results are 1090shown in Table A1. The three models all behave well and 1091the adjusted R2 are above 0.55. Almost all independent 1092Yiu, C. Y., & Wong, S. K. (2005). The effects of expected trans-1093port improvements on housing prices. Urban Studies, 42(1), 1094113-125. https://doi.org/10.1080/00420980420003097201095Yu, T. H., Cho, S. H., & Kim, S. G. (2012). Assessing the resi-1096dential property tax revenue impact of a shopping center. The 1097Journal of Real Estate Finance and Economics, 45(3), 604-621. 1098https://doi.org/10.1007/s11146-010-9292-x1099Figure A1. Partition method of distance ring1100Table A1. Results of regression of the three models1101Independent variables1102Model 11103Model 21104Model 31105Coefficient1106T – statistics1107Coefficient1108Coefficient1109T – statistics1110Coefficient1111Constant111210.718***1113270.712111411.123***1115233.887111611.177***1117244.9441118Ln Size1119–0.090***1120–21.7111121–0.085***1122–20.9231123–0.085***1124–20.9761125Ln Floor/Level No.1126–0.033***1127–8.2781128–0.036***1129–9.0471130–0.036***1131–9.1201132Ln Total floors11330.009***11343.96111350.009***11364.56711370.010***11384.5991139Ln Building’s age1140–0.0011141–1.1621142–0.003***1143–3.5791144–0.004***1145–4.4261146Ln Distance to CBD1147–0.431***1148–39.4911149–0.538***1150–46.6941151–0.542***1152–46.8961153Ln Distance to West Intime1154–0.053***1155–15.4421156Ln Distance to Xicheng Square1157–0.016***1158–4.0771159–0.036***1160–7.4121161–0.038***1162–7.8481163Bus route1164–0.003***1165–9.9351166–0.004***1167–14.1691168–0.004***1169–14.1871170Surrounding environment11710.090***117225.05111730.051***117412.63511750.051***117612.4721177Inner environment1178–0.019***1179–6.5971180–0.0031181–1.1101182–0.0031183–1.1601184Living facilities1185–0.006**1186–2.11311870.00411881.34711890.00311901.1201191Sports facilities11920.035***119317.22111940.031***119515.40211960.030***119715.2251198Property management11990.063***120022.99612010.061***120222.29412030.062***120422.6151205Quality of primary school12060.054***120725.60212080.059***120925.86112100.059***121125.8301212Quality of junior high school12130.00312140.5131215–0.0091216–1.2621217–0.0141218–1.8721219Ln Distance to primary school1220–0.005**1221–2.33912220.009***12234.21912240.009***12254.2061226Ln Distance to junior high school12270.038***122816.2321229–0.0021230–0.8121231–0.0021232–0.7381233Y21234–0.063***1235–12.5701236/1237/1238–0.062***1239–13.0461240Y312410.00812421.5791243/1244/12450.01112462.3861247Y41248–0.039***1249–7.6331250/1251/1252–0.023***1253–3.3101254Y51255–0.048***1256–10.2951257/1258/1259–0.031***1260–4.6751261Adjusted R212620.55712630.59612640.5921265Note: ***, **, and* represent significance at the 1%, 5% and 10% levels, respectively.126612671268===== PAGE 15 (score 116, numeric-paren 94, table-words 4, var-words 3) =====12691270International Journal of Strategic Property Management, 2019, 23(1): 65–801271791272variables are significant and the values of the coefficients 1273are hardly different from the original partitioning meth-1274od.1275In Model 1, the coefficient of Distance to West Intime 1276is – 0.053, which shows the negative price gradient and 1277validates the influence of West Intime. In Model 2, house 1278prices in Zones 1 and 2 increased after 2014, while those 1279in other zones declined over two years, excepting for Zone 82. 1280In Model 3, as shown in Figure A2, there was a clear rise 1281in Zones 1 and 2 during Period 2, as prices in the other 1282regions fell. The results of the three models support the 1283robustness of the findings in the text.12842 The high house price in Zone 8 may be because most of the 1285samples are concentrated in the southeast of the ring, which 1286is closer to the city center. The uneven distribution of samples 1287and the difference in house prices in all directions may affect 1288the accuracy of the results. This is the main reason why we 1289have not chosen this division method in the text.1290Appendix 21291Using the location coordinates and the shape of the com-1292munities, as well as the number of samples therein, we 1293randomly generated the coordinates for each house unit 1294by GIS. Spatial weights could then be established in order 1295to conduct a spatial test and spatial regression. The re-1296gression results show that the spatial econometric model 1297SEM performs better (see Table A2 and A3). However, the 1298pattern of price gradients of Zones 1−9 with Model 2 and 1299Model 3 are very close to the results in the text. Due to 1300limited space, the results are not shown here and can be 1301provided upon request.1302 1303–0.251304–0.201305–0.151306–0.101307–0.0513080.0013090.051310Z11311Z21312Z31313Z41314Z51315Z61316Z71317Z81318λi + θi,p1319T11320T21321Figure A2. Price gradients of Zones 1–81322Table A2. Results of regression of different models1323Coefficient1324Model 11325Model 21326Model 31327Linear1328Linear – log1329Log – linear1330SAR1331SEM1332SEM1333SEM1334Constant133522205.649***133630108.698***13379.995***13382.972***133910.566***134010.837***134110.806***1342Size1343–22.814***1344–1740.935***13450.001***1346–0.096***1347–0.122***1348–0.122***1349–0.124***1350Floor/Level No.135120.424**1352190.989**1353–0.001***13540.011***13550.012***13560.012***13570.013***1358Total floors1359–32.672***1360–652.257***1361–0.002***1362–0.052***1363–0.102***1364–0.099***1365–0.100***1366Building’s age1367–30.326***1368235.921**1369–0.0011370–0.013***13710.028***1372–0.036***1373–0.030***1374Distance to CBD1375–789.131***1376–4095.293***1377–0.046***1378–0.046***1379–0.219***1380–0.291***1381–0.282***1382Distance to West Intime1383–305.480***1384–1893.405***1385–0.018***1386–0.037***1387–0.136***1388/1389/1390Distance to Xicheng Square1391–327.487***1392–959.592***1393–0.015***1394–0.008*1395–0.030*1396–0.0191397–0.0131398External environment1399718.103***1400730.509***14010.047***14020.015***14030.068***14040.076***14050.073***1406Inner environment1407–284.909***1408–366.970***1409–0.016***1410–0.0021411–0.0051412–0.018*1413–0.0141414Living facilities1415–181.152***1416–92.520*1417–0.013***1418–0.0031419–0.0041420–0.025***1421–0.026***1422Sports facilities1423300.827***1424131.780***14250.014***14260.004*14270.00814280.024***14290.023***1430Property management14311170.150***14321656.332***14330.061***14340.030***14350.081***14360.060***14370.061***1438Bus route1439–14.807*1440–18.432**1441–0.001***1442–0.001***1443–0.005***1444–0.005***1445–0.005***1446Distance to primary school1447–763.224***1448–548.397***1449–0.045***1450–0.014***1451–0.026**1452–0.019**1453–0.017*145414551456===== PAGE 16 (score 73, numeric-paren 54, table-words 4, var-words 3) =====14571458801459L. Zhang et al. The effects of a shopping mall on housing prices: a case study in Hangzhou1460Coefficient1461Model 11462Model 21463Model 31464Linear1465Linear – log1466Log – linear1467SAR1468SEM1469SEM1470SEM1471Distance to junior high 1472school1473835.816***1474613.014***14750.044***14760.008***14770.023**1478–0.047***1479–0.046***1480Quality of primary school14811534.241***14821790.623***14830.071***14840.026***14850.093***14860.067***14870.067***1488Quality of junior high school1489139.6381490186.55014910.038***14920.00814930.056**14940.03714950.0381496Y21497–1277.337***1498–1370.727***1499–0.067***1500–0.068***1501–0.074***1502/1503–0.064***1504Y3150560.7591506–84.84515070.0001508–0.0021509–0.015***1510/1511–0.0041512Y41513–825.525***1514–1015.551***1515–0.046***1516–0.046***1517–0.061***1518/1519–0.066***1520Y51521–745.043***1522–984.428***1523–0.048***1524–0.047***1525–0.067***1526/1527–0.066***1528W_lnprice1529/1530/1531/15320.755***1533/1534/1535/1536LAMBDA1537/1538/1539/1540/15410.803***15420.685***15430.681***1544Adjusted R215450.59115460.58715470.64415480.70915490.72215500.73415510.7281552Note: ***, **, and * represent significance at the 1%, 5% and 10% levels, respectively.1553Table A3. Diagnose test for OLS regression of Model 1−31554Model 11555Model 21556Model 31557Moran᾽s I (resudals)15580.223 (0.000)15590.118 (0.000)15600.114 (0.000)1561Lagrange Multiplier (lag)15625855.282 (0.000)15631512.888 (0.000)15641434.977 (0.000)1565Robust LM (lag)1566169.693 (0.000)15673.189 (0.074)15682.301 (0.129)1569Lagrange Multiplier (error)15708528.628 (0.000)15712386.997 (0.000)15722232.014 (0.000)1573Robust LM (error)15742843.040 (0.000)1575877.298 (0.000)1576799.337 (0.000)1577Note: the number in the parentheses are p value.1578End of Table A21579