Shaping Suburbia: a Comparison of State-Led and Market-Led Suburbs In

Home , Bundang, Bundang High School, Seongnam, Yongin

Original Article Shaping suburbia: A comparison of state-led and market- led suburbs in Seoul Metropolitan Area, South Korea

Hyungkyoo Kima and Seung-Nam Kimb,* aDepartment of Urban Design and Planning, Hongik University, 94 Wausan-ro, Mapo-gu, Seoul, 121-791, South Korea. E-mail: [email protected] bArchitecture and Urban Research Institute, B-301 Acrotower Ofﬁce, 230 Simin-daero, Dongan-gu, Anyang-si, Gyeonggi-do 431-908, South Korea. E-mail: [email protected] *Corresponding author.

Abstract The Seoul Metropolitan Area (SMA), South Korea, is a region where two distinct types of suburbs exist today, new towns developed by the public sector and suburban communities by the private. This article comparatively analyzes Bundang, a state-led new town, and Suji, a market-led community, in the SMA in terms of two key foci. The ﬁrst is urban form characteristics, which include land use pattern, street pattern and circulation, public open space, built form pattern, and control. The other is residents’ livability and satisfaction, which include public transit service and use and neighborhood satisfaction. Findings suggest that Bundang outperforms Suji in all aspects, demonstrating the beneﬁts and effectiveness of planning in shaping suburbia, while Suji shows many negative signs of sprawl. URBAN DESIGN International advance online publication, 28 October 2015; doi:10.1057/udi.2015.19

Keywords: new town; suburb; Seoul Metropolitan Area; urban form; livability; neighborhood satisfaction

Introduction increase in the late twentieth century, migration to suburbs for more affordable housing options and Suburbanization is an international phenomenon healthier living environments was another conse- (Forsyth, 2014). Since the ﬁrst modern suburbs quence along with the densiﬁcation and compac- were born at the urban fringes in the United tion of cities (Phelps and Wu, 2011). Kingdom (UK), they have prospered in many parts In the process of shaping suburbia, a clear of the world and have even become the dominant distinction exists in the involvement of the public landscape in countries like the United States (US) and private sectors. In the UK, it was the public (Jackson, 1985; Fishman, 1987; Hayden, 2004; sector that played the leading role. The 1940 Bruegmann, 2005). A major driving force behind Barlow Report, drafted as a Royal Commission, this growth was the desire for a more livable and the 1944 Greater London Plan by Sir Patrick environment away from the ills of overcrowded Abercrombie supported planned decentralization cities. In the UK, John Nash designed Park Village, and development of new towns to mitigate the side and Ebenezer Howard envisioned the garden city. effects of London’s soaring population and hous- In the US, Frederick Law Olmsted and Calvert ing shortage (Gibberd et al, 1980; Hall, 2002). The Vaux argued for healthy suburban living as in 1945 New Town Committee and the 1946 New Riverside, and Clarence Stein and Henry Wright Towns Act provided the basis for the construction designed Radburn. Asian cities were not exempt of 21 new towns, including 8 in the London from experiencing extensive suburbanization. Metropolitan Area, under the supervision of the Especially in countries that went through rapid New Town development corporations (Oxford economic growth, urbanization and population Brookes University, 2006).

In the US, the role of the public sector has among urbanists around the world that ideas and been relatively indirect (Bruegmann, 2005). The views of suburban growth should be shared inter- development of Greenbelt cities (Greenbelt, Green- nationally (Forsyth, 2014), this study attempts to dale and Greenhills) led by Rexford Tugwell dur- expand the knowledge on suburbs and provide ing the New Deal era was the only case of direct alternatives to overcome the ill consequences of federal intervention in the country’s suburban suburban sprawl that many cities and regions development history. Rather, it was the private around the world face today. sector that played a bigger role in shaping the American suburbs, including milestones that range from Riverside, streetcar suburbs, Rad- Suburbanization in the SMA burn and Levittown to edge cities (Garreau, 1991), boomburbs (Lang and LeFurgy, 2007) and metro- South Korea and its suburbanization have been burbs (Knox, 2008). Notable planned communities relatively little introduced to international plan- such as Reston, Irvine, Columbia and Woodlands ners and designers. After the Korean War in the as well as new urbanist projects that provided 1950s and before becoming a member of the an alternative model to suburban sprawl (Heid, Organisation for Economic Co-operation and 2004; Forsyth, 2005) were also led by private Development (OECD) in 1996, the country went developers. State and local governments generally through one of the most rapid economic develop- relied on implementing policies such as growth ments in the world. The active role of the public management programs and subdivision layout sector seen in the UK, Denmark, Germany, the standards to guide the private sector to reach a Netherlands and Japan (Alterman, 2001) has also ’ more desirable state of regional development been evident in South Korea. The country s land (Southworth and Parthasarathy, 1996; Hayden, use policy is characterized by the dominance of a 2004; Chapin, 2012). strong top-down governance whereby the state The Seoul Metropolitan Area (SMA) is the long- establishes major planning policies and frequently time center of economy, politics and culture of the intervenes in urban and regional issues (Bae and Korean peninsula. Composed of Seoul Special Sellers, 2007; Lee and Shin, 2012). ’ City, Incheon Metropolitan City and Gyeonggi One consequence of South Korea s rapid eco- Province, it accommodates 22.5 million residents nomic development was a sharp increase in popu- and is the largest metropolitan area by population lation which was mostly absorbed by the capital in the country and the ﬁfth largest in the world city, Seoul (Meier, 1970; Hill and Kim, 2000). after the Tokyo, Jakarta, Delhi and Manila metro- Between 1960 and 1980, the population of Seoul, politan areas (Demographia, 2015). In addition, the which covers only 0.6 per cent of the land area of SMA is a region where extensive suburbanization the country, rose by 242 per cent, while that of the occurred in the last several decades but also where nation increased by 52 per cent (Figure 1). The both public and private sectors played a vital role, soaring population resulted in serious congestion, leaving two distinctive types of suburban develop- housing shortages and an extreme surge in hous- ment patterns behind. The state-led suburbs, ing and land prices in Seoul (Kim and Han, 2012). usually in the form of large-scale master-planned new towns, were developed by the public sector as 30% a national policy measure since the 1970s with Seoul Incheon and Gyeonggi Combined the goal of supplying affordable, quality housing.

The market-led suburbs, developed by the private 20% sector especially since the 1990s, are usually in the form of small- to mid-sized communities. Whether the public or private sector should take 10% more control in shaping suburbia, or even in urban development in general, remains a challenging question. However, a comparative analysis of the 0% 1960 1966 1970 1975 1980 1985 1990 1995 2000 2005 2010 SMA’s state-led new town and market-led community in this study, with respect to their urban Figure 1: Population shares of Seoul and Incheon and ’ Gyeonggi combined, 1960–2010. form characteristics and the residents livability Note: 1965 Population Census was postponed to 1966 due to and satisfaction, may provide guidance for plan- national budget deﬁcit. ners and designers. As consensus is growing Source: Korea Statistical Information Service (http://kosis.kr).

Figure 2: Locations of major state-led new towns in the SMA.

The state’s reaction was to intervene by supplying the Two Million-Home Construction Plan to miti- a large number of quality, affordable housing gate chronic housing shortage and soaring housing in the SMA, especially outside Seoul in Incheon prices. Under this plan, the state-run public devel- and Gyeonggi in the form of large-scale new opers master-planned and constructed five new towns (Lee and Oh, 2012). Around 1970, the Korea towns, the SMA’s so-called first-generation new National Housing Corporation and the Korea towns – Bundang, Ilsan, Pyeongchon, Sanbon and Land Corporation were established by the state as Joongdong – each accommodating up to about public developers to undertake planning and con- 390 000 new residents (Figure 2). They were able struction of the new towns.1 Local governments to tame the overcrowding of Seoul and accelerated and regional public corporations and agencies population growth in Incheon and Gyeonggi since participated in the process. the 1990s (Figure 1). In the 2000s, to promote In the 1970s, the first state-led new towns in the further stability of the region’s housing market, SMA emerged when Ansan, an industrial city, and the state additionally built 10 second-generation Gwacheon, an administrative city, were built south new towns – Pangyo, Dongtan I and II, Gwang- of Seoul. In 1980, the state enacted the Housing Site gyo, Godeok, Hangang, Geomdan, Unjeong, Development Promotion Act that provided a legal Yangju and Wirye – each accommodating up to framework for the public sector to acquire, develop about 286 000 residents.2 and manage land for constructing large-scale new However in the mid-1990s, the deregulated, towns. Circa 1990, development of state-led new marketized political economy and voices critical towns reached its zenith when the state adopted of large-scale state-led projects gave room for

© 2015 Macmillan Publishers Ltd. 1357-5317 URBAN DESIGN International 1–20 3 Kim and Kim private developers to participate in the suburban medical, educational and institutional purposes development processes (Lee and Shin, 2012). The (Korea Land Corporation, 1997). Its master plan slackened national planning regulations facilitated and area-wide urban design plan were prepared residential and commercial development on agri- by the Korea Research Institute for Human Settle- cultural lands (Korea Land Corporation, 2000). ments, a state-run think tank of national land While the private developers had been no more policy. Elements of the master plan included land than general construction contractors or develop- use, housing, circulation and public transit, public ment partners in building the first-generation new open space, civic facilities, and infrastructure. The towns, they started to construct their own small- to urban design plan provided regulations on build- mid-sized residential communities as the market ing form and placement, street design and circula- opened up. Local critics refer these communities to tion, open space, civic facilities, and infrastructure. as ‘parasitic towns’ (Kwon, 2003; Jun et al, 2013) for Although some were later modified, these ele- adjoining existing state-led new towns and free- ments and regulations were strictly enforced riding their infrastructure. throughout Bundang. Other first-generation new towns and the second-generation new towns, as well as those in other metropolitan areas of the Case Context country, reapplied many of the planning and design principles adopted in Bundang. For comparative analysis, we selected Bundang to Suji, located directly south of Bundang, is one represent the state-led new town and Suji the of the best-known market-led suburban commu- market-led community. The two suburbs are nities in the SMA. With a name that literally located immediately adjacent to each other in means ‘water and branch’ in Korean, it was southeast of Seoul in Gyeonggi (Figure 3). Local originally an area of agricultural fields, villages planners and decision makers consider them as and creeks. In the mid-1990s, Suji received enor- comparable urban areas for their similarities in mous attention in the region’s property market geographical location, distance to Seoul, develop- for its low land prices and proximity to Bundang ment times, population size and socioeconomic and Seoul. These new residential communities status of the residents. As of 2014, Bundang attracted many home buyers, who were looking accommodates a total population of 405 814 in for affordable housing in the region, thus increas- 1960 Ha, and Suji 340 420 in 4210 Ha (Table 1). ing Suji’spopulationbymorethansixtimesonly Henceforth, we use the term Bundang to refer to within 10 years between 1990 and 2000.3 Direct the Bundang New Town located in Bundang Dis- state-led planning occurred in parts of Suji, allo- trict of Seongnam City, and Suji to the Suji District cating land for commercial or civic use.4 How- of Yongin City. ever, it was the private developers who shaped Bundang is the most representative state-led most of Suji’s landscape, converting the area into new town in South Korea (Urban Design Institute an extensive bedroom community with little self- of Korea, 2012). Located 25 km southeast of Seoul, sufficiency. They purchased semi-agricultural it was one of the first initiated projects among the lands that allows up-zoning to semi-urban use five first-generation new towns and is the largest which accommodates residential and commercial among all state-led new towns in the country. The uses. Rather than promoting large-scale residen- Korea Land Corporation executed Bundang’s tial projects, they strategically kept developments development swiftly, first announcing the project small but sizable enough to be profitable so as to in April 1989 and finalizing its master plan only 3 receive exemptions on infrastructure provision months later. Construction of sites and infrastruc- requirements such as wider entry roads, public ture began before the end of the same year (Korea transit, recreational and civic facilities, parks, and Land Corporation, 1997). The main goal of con- schools (Korea Land Corporation, 2000; Gyeonggi structing Bundang was to provide a livable sub- Research Institute, 2003).5 Some developers split urban settlement furnished with advanced transit parcels into multiple smaller ones in order to meet system, accessible recreational space and safe the requirements for exemption (Roh, 2000). dwelling environment for all income groups Undersupply of infrastructure and the extremely (Korea Land Corporation, 1997). The state paid rapid population growth in Suji burdened the specific attention to promoting self-sufficiency by local government with enormous cost on infra- supplying more than 1.6 million m2 of land for structure to support the new residents (Korea office and retail uses and another half a million for Land Corporation, 2000) and characterized the

Figure 3: Locations of Bundang and Suji, and the selected areas for comparative analysis (B1, B2, S1 and S2; and Ba, Bb, Bc, Sa, Sb and Sc). Note: S1 and Sa include a small part of Giheung District due to their direct proximity.

area as one of the most sprawling parts of the regard to employment. A total of 41.4 per cent of country (Chung and Kim, 2003).6 commuters in Bundang and 36.0 in Suji travel to As Harris (2010) and Forsyth (2012) pointed out, Seoul, outnumbering those who have jobs in their little has been discussed on the constituents of a cities (Table 1).7 Third, the two suburbs are located suburb, and the diversity in geographical contexts within 1-hour driving distance from Seoul. Lastly, and academic disciplines make it difﬁcult to deﬁne the vast majority of existing studies on the SMA, the term. However, a few factors characterize for example, Lee and Ahn (2005), Lee and Shin Bundang and Suji more suburban than urban. (2012) and Jun et al (2013), perceive them as typical First, the population sizes of Bundang and Suji are elements of the country’s suburban landscape. incomparably smaller than that of Seoul (Table 1). With regard to population density, which often is Second, both show heavy reliance on Seoul with accepted as a key characteristic of suburbs,

Table 1: Demographic and socio-economic characteristics and commuting patterns of Bundang and Suji in comparison with Seoul

Characteristic Original plan of bundang Bundang Suji Seoul

Demographic characteristicsa Population 390 320 405 814 340 420 10 103 233 Number of households 97 580 141 384 117 288 194 176 Average household size 4.0 2.9 2.9 2.4 Population density (persons/hectare) 198.7 207.0 80.9 166.9

Socio-economic characteristicsb Average annual household income (in US dollars) NA 54 540 50 760 45 434 Average number of cars per household NA 1.16 1.23 0.83

Commuting patternsc Proportion of workforce commuting to Seoul (%) NA 41.4 36.0 88.9 Proportion of workforce commuting within their city (%) NA 37.4 29.6 88.9 aAs of year 2014. bAs of year 2013. cAs of year 2010. Sources: Korea Land Corporation (1997); Korea Statistical Information Service (kosis.kr); 2010 National Household Travel Survey in South Korea (www.ktdb.go.kr); Gyeonggi Province Transportation Database Center (gtdb.gg.go.kr/GTDBWeb/); Seongnam City (2013); Yongin City (2013).

Bundang shows a higher value than Seoul Among the aforementioned studies, we employ (Table 1). This is not an unusual case since many the analytical frameworks used by Southworth suburbs around the world are denser than their (1997) and Lee and Ahn (2003).8 We also incorpo- central cities (Forsyth, 2012). rate the livability measures of urban form, which Lynch (1981), Jacobs and Appleyard (1987) and Southworth (2003) suggested.9 In this sense, we Comparative Analysis of Urban Form selected ﬁve urban form characteristics to compare Characteristics Bundang and Suji as listed below.

This analysis focuses on literature that studied ● Land use pattern: Land use types and mix, den- comparatively the urban form characteristics of sity, and their arrangements. various suburban developments. At the regional ● Street pattern and circulation: Street types and scale, Southworth and Owens (1993) examined their arrangements, vehicular and pedestrian patterns of eight suburban areas in the San Fran- networks, and street connectivity. cisco Bay area and witnessed the decline of the ● Public open space: Recreational spaces and public public realm over time. Wheeler (2008) observed squares. six metropolitan areas in the US and identified a ● Built form pattern: Building footprint and three- set of historic patterns and recent typologies of dimensional urban form. suburban development. With respect to new urba- ● Control: Degree to which the land is under public nist developments, Southworth (1997) compared ownership or control. Kentlands and Laguna West with a typical streetcar suburb and questioned the effectiveness of The five characteristics may not be the most com- making walkable suburbs without adopting a prehensive list of urban form attributes and may regional approach. Lee and Ahn (2003) analyzed fall short of fully addressing its complex nature, urban form of Kentlands with Radburn and com- and the findings of the aforementioned similar mended Radburn’s prescription for better walk- studies are coming from American suburbs. There- ability. In addition, Scheer and Petkov (1998) fore, the appropriateness of applying these five contrasted urban form of three edge cities with characteristics to the SMA may be questioned. that of a traditional urban core in the US and However, they are not dogmatic principles that suggested implications for future suburban devel- present customarily set guidelines or standards on opment. Zacharias (2005) examined two state-led various elements of urban form such as block size, new towns in Hong Kong and identified design street width, set-back and land use, but essential features that promote walkability and bikability. dimensions of urban form that provide a

6 © 2015 Macmillan Publishers Ltd. 1357-5317 URBAN DESIGN International 1–20 Shaping Suburbia conceptual framework for discussion and analysis overwhelming dependence on high density and on urban environment. The studies that compara- commonly assign a fair amount of land to com- tively analyzed various urban form patterns with mercial/office, civic/institutional and recreational differing development backgrounds ascertain the uses. While Bundang incorporates more mixed and effectiveness and robustness of the five character- recreational uses, Suji has more low-density resi- istics as an analytic framework. dential and agricultural uses. The land use mix in Since Bundang and Suji are too large to be the original plan of Bundang does not significantly analyzed at a reasonable scale in their entirety, we differ in B1 and B2, while that in S1 and S2 in Suji selected parts of the two based on location and land shows substantial disparity. Within Suji, S2, com- use with some randomness introduced within each pared with S1, has more land assigned to residen- criterion. We placed two 1.6 km-by-1.6 km squares – tial use with heavier dependence on high-density, B1 and B2 in Bundang, and S1 and S2 in Suji – in roads and agricultural uses, while less commer- analyzing land use pattern, public open space, built cial/office, mixed and recreational uses. form pattern and control; and three 800 m-by-800 m In Bundang, a systematic arrangement of land squares – Ba, Bb and Bc in Bundang, and Sa, Sb and uses is distinct. In both B1 and B2, the centrality of Sc in Suji – in analyzing street pattern and circulation commercial areas in the form of transit-oriented (Figure 3). B1, B2, S1 and S2 correspond to whole developments is clearly visible (Figure 6). In B1, the neighborhoods. Ba and Sa relate to central commer- commercial core in the northwestern corner oper- cial areas, Bb and Sb to high-density residential ates as one of the two main centers of Bundang and areas, and Bc and Sc to residential areas at all extends to the southeast along a linear corridor that densities.10 Our selection process minimizes the accommodates mixed, commercial and recreational potential misrepresentation of the two areas since uses. Access streets and pedestrian walkways the design principles of Bundang are kept generally serve the corridor. In B2, residential blocks with constant throughout the area, and S1 and S2 in Suji civic/institutional and recreational uses that are coveralmosthalfofitstotaldevelopedarea.Inthe evenly dispersed surround a compact commercial next sections, we compare Bundang and Suji accord- core. High-density residential blocks are located ing to the five urban form characteristics we elabo- toward the center and north, whereas medium- and rated earlier. low-density to the periphery and south. A web of car-free walkways connects commercial and civic/ institutional uses with recreational blocks. Land use pattern In Suji, such an arrangement of land uses is less clear. Although the layout of residential and Bundang and Suji share many land use types in commercial lands and their parcelization are rela- common (Table 2; Figures 4 and 5). They show tively well organized in the central, eastern and a similar mix of residential densities with an northeastern parts of S1, the remaining parts of S1

Table 2: Percentage (%) of land use types in B1 and B2 in Bundang (in comparison with original plan), and S1 and S2 in Suji

Land use type Bundang Suji

Original plan B1 B2 S1 S2

Residential 32.4 36.6 28.2 28.4 40.7 Low-density residential 0.4 0.0 0.0 0.2 2.3 Low-density residential with mixed use 3.3 4.7 2.0 5.0 0.4 Medium-density residential 4.0 0.5 4.3 1.3 0.1 High-density residential 24.0 25.6 17.5 21.9 37.1 High-density residential with mixed use 0.7 5.8 4.4 0.0 0.8 Commercial/ofﬁce 8.3 5.9 5.1 12.5 0.6 Civic/institutional 11.7 2.3 3.8 4.5 3.7 Recreational/pedestrian and bicycle pathways 25.2 28.7 34.7 15.9 5.5 Forest/green buffer 2.8 2.7 2.0 11.6 8.1 Agricultural 0.0 0.0 0.0 2.6 5.3 Public road 19.6 23.8 26.2 24.5 36.1 Total 100.0 100.0 100.0 100.0 100.0

Source: Korea Land Corporation (1997).

Figure 4: Land use types in Bundang: (a) Low-density residential with mixed use; (b) medium-density residential; (c) high-density residential; (d) high-density residential with mixed use; (e) neighborhood commercial; (f) central commercial.

Figure 5: Land use types in Suji: (a) low-density residential with mixed use; (b) high-density residential; (c) neighborhood commercial; (d) central commercial; (e) big-box retail; (f) agricultural. and almost everywhere in S2 show little organiza- fragmented. Also in S2, a substantial amount of tion in land use. Especially in S2, residential and agricultural land remains directly adjacent to high- commercial lands are unevenly scattered through- density parcels, showing patches of leap-frog out the area. Unlike Bundang, it is difficult to find development patterns that contribute to the ineffi- any planning principles at work in these parts of cient use of land and insufficient supply of urban Suji. Their land use is not mixed use but highly infrastructure (Ewing, 1994).

Figure 6: Land use patterns in B1 and B2 in Bundang and S1 and S2 in Suji. Squares are 1.6 km on each side.

Street pattern and circulation from those observed in Radburn, where the roads are public. The third is the wider street The two areas show several common features in widths than in other South Korean cities which street pattern (Figure 6). The first is the dominance reflect the increasing automobile ownership in the of superblocks that reach up to 500 m on one side. 1990s. The superblocks usually contain large-scale There are four types of streets in Bundang: residential complexes composed of high-rise apart- arterials (28–40 m wide), sub-arterials (20–30), col- ment buildings, neighborhood commercial build- lector streets (12–20) and access streets (6–10). The ings, and spaces for parking and recreation. The arterials carry 6–10 car lanes, a central median and second is the absence of culs-de-sac and loops. sidewalks. Usually, green buffers supplement They exist in many high-density residential devel- them on each side so as to protect the inhabitants opments but are no more than private internal in adjacent parcels from traffic and noise. The access roads, making the superblocks different arterials and sub-arterials frequently configure the

Figure 7: Pedestrian walkways in Bundang: (a) Residential area; (b)commercialarea;(c) vertical separation of pedestrians and motorists. Source:(c) Aerial photograph from © Naver Maps (map.naver.com). boundaries of high-density residential blocks, pub- vehicular traffic is widening and straightening lic open spaces and commercial areas (Figure 6). historical trails. The collector and access streets are mostly found in To expand the discussion, we carried out a commercial or low-density residential areas and quantitative comparison of vehicle and pedestrian limitedly in high-density residential areas. A con- networks by calculating linear meters of streets tinuous network of pedestrian walkways, also and number of blocks, intersections (nodes) and used as bicycle paths, overlaps Bundang’s street links in Ba, Bc and Bc in Bundang, and Sa, Sb and grid (Figure 7). They penetrate through the midst Sc in Suji (Tables 3 and 4). Among the two, the of the superblocks, providing car-free access to pedestrian network includes both the vehicular open spaces, schools and commercial areas for and pedestrian rights-of-way, assuming that side- pedestrians. Vertical separation of pedestrians and walks are located on both sides of street. Overall, motorists, as in Radburn, occurs in several loca- the vehicular network in Suji generally outper- tions (Figure 7). forms that in Bundang in total street length and Street hierarchy found in S1 of Suji resembles number of blocks, intersections and links per unit that in Bundang (Figure 6). The arrangement and area, indicating that Suji has a more fine-grained design of streets and their relationship to land use street network than Bundang (Table 3). Similar are generally alike, except that the blocks are preponderance is observed in the comparison of generally smaller. However, street patterns in most the pedestrian network. However, it is difficult to parts of Suji are far from being persistent. The say that Suji is more walkable. Pedestrian walk- hierarchy loses consistency, as demonstrated in S2 ways hardly exist in most parts of Suji, while where the arterial transforms into a highway as it Bundang has a more accessible pedestrian network heads southwest (Figure 6). Often, arterials and throughout the area. Also, Bundang shows a dra- sub-arterials provide access to low-density resi- matic increase in the linear meters of streets and dential blocks, and access and collector streets number of blocks, intersections and links when we serve high-density residential. Pedestrian walk- incorporate the pedestrian network into calcula- ways exist, but are extremely limited, providing tion (Table 4). minimal options for non-motorized travel. We also examined street connectivity, which is Interestingly, an organic street pattern prevails determined by the quantity and quality of connec- in many parts of Suji. A large number of residential tions in a network (Tables 3 and 4). Literature and commercial parcels are not in the usual suggests that better connectivity provides more rectangular shape but in irregular forms enclosed route options, shortens trip distances (Frank and by winding streets and oblique intersections. Engelke, 2005), and increases the likelihood of However, Suji’s organic street pattern does walking and bicycling (Handy et al, 2003). Here, not relate to what Camillo Sitte advocated. we measured the connectivity index, the ratio of the It depicts not the incremental growth of Suji but number of links to the number of intersections the indiscreet conversion of agricultural fields to (American Planning Association, 2006). Our calcu- residential or commercial use. Unfortunately, lation shows that the street network of Bundang uneven development of the area is deforming this has a higher connectivity level for both vehicles organicity. Gridded pods are popping up here and and pedestrians, providing a more efficient overall there, and the need for accommodating more street network with better circulation.

Area Ba Bb Bc Sa Sb Sc

05McilnPbihr t.1357-5317 Ltd. Publishers Macmillan 2015 © Street patterns

Intersections RA EINInternational DESIGN URBAN

Linear meters of streets 5920 4780 5289 9839 5890 7083 Number of blocks 21 18 27 54 16 41 Number of intersections 22 14 26 66 18 42 Number of links 40 28 54 120 36 70 Connectivity index 1.8 2.0 2.1 1.8 2.0 1.7

Notes: Squares are 800 m on each side. 1 – 20 hpn Suburbia Shaping 11 2©21 amla ulsesLd 1357-5317 Ltd. Publishers Macmillan 2015 © 12 i n Kim and Kim

Table 4: Pedestrian network

Area Ba Bb Bc Sa Sb Sc

Street patterns

Intersections RA EINInternational DESIGN URBAN

Linear meters of streets 6882 (961) 5636 (856) 7364 (2075) 10 551 (713) 5890 (0) 8607 (13) Number of blocks 30 22 46 67 16 42 Number of 28 (6) 20 (6) 60 (34) 85 (19) 18 (0) 42 (0) intersections Number of links 54 (14) 42 (14) 118 (64) 149 (32) 36 (0) 70 (1) Connectivity index 1.9 (0.4) 2.1 (2.3) 2.0 (1.9) 1.8 (1.7) 2.0 (N/A) 1.7 (N/A) 1 – 20 Notes: Values in parentheses apply to pedestrian-only rights-of-way. Squares are 800 m on each side. Shaping Suburbia

Public open space Bundang, paved pedestrian walkways, bicycle trails and greenery provide space for recreational The share of public open space, which includes activities (Figure 8b). Suji is more abundant with recreational spaces and public squares, is higher in creeks. Some parts along Jeongpyeong and Seong- Bundang than in Suji (Table 2). While an organized bok creeks have been converted to recreational system of public open spaces at various sizes is spaces since the late 2000s, while many parts still found in Bundang, Suji is without one (Figure 6). remain as an edge where developments directly Substantial forest land exists in Suji but is predom- border on the creeks (Figure 9b). inantly privately owned. Another type of public open space is pedestrian Types of recreational spaces and their arrange- squares. Those located in the Seohyeon commercial ments also differ across Bundang and Suji. Exis- center of Bundang are 20–30 m wide and designed tence of the 42-hectare Joongang Park, which with substantial street furniture which provides literally means ‘central park’ in Korean, symbol- space for pedestrians to linger or rest (Figure 8c). izes Bundang’saffluence of public open space (see On the other hand, those in the Pungdeokcheon top-right quadrant of B1 in Figure 6; Figure 8a). commercial center of Suji are generally 10–12 m Smaller neighborhood recreational spaces accessi- wide and without specific facilities for pedestrians ble by pedestrian walkways can be easily found all (Figure 9c). over Bundang. Meanwhile in Suji, such large-scale recreational spaces do not exist. Rather, blocks with irregular shapes, on which profitable residen- Built form pattern tial development seemed to have been difficult, are developed as neighborhood parks (S1 and S2 in The overall built form patterns of Bundang and Figure 6; Figure 9a). Individual residential devel- Suji show a strong concordance with the under- opments usually accommodate recreational space lying land use. They also share a similar mix of within their premises but provide restricted access grain, implying that the dominant landscapes in to the public. both areas are relatively uniform and that building Both Bundang and Suji have a fair amount of types lack diversity. In Bundang and parts of Suji, creek-side public open space. Along Tan Creek in clustering of similar built form patterns is frequent

Figure 8: Public open spaces in Bundang: (a) Joongang Park; (b) Tan Creek; (c) Seohyeon commercial center.

Figure 9: Public open spaces in Suji: (a) Sanghyeon Park; (b) Jeongpyeong Creek; (c) Pungdeokcheon commercial center.

(Figure 10). In the remaining parts of Suji, the greenery (Figures 11a and 11b). However, in patterns are unevenly scattered. Footprints of parts, especially where the land use is designated farmhouses and cottages in very fine grain still as high-density residential with mixed use, rows exist, but those of high-rise apartment buildings of tall residential towers in uniform height often often sandwich them. block view corridors (Figure 11c). In Suji, without The most dominant built form pattern in both any similar approach, high-rise apartment build- areas is that by the high-rise apartment buildings ings with uniform heights dominate the skyline in coarse grain, reflecting the form of typical (Figure 12a). Openness to nature, secured view residential buildings built in the 1990s (Figure 10). corridors and diversity in building heights are They are usually 15 or more stories high and in rare. The indiscreet conversion of agricultural linear rectangular shapes that usually run 60–90 m lands to residential or commercial use dwarfs the with six to eight units on each floor. In Bundang, traditional landscape (Figures 12b and 12c). these buildings are usually placed in a block or parcel so as to constitute clusters of communities, creating a notion of enclosure and scale. This Control pattern repeats in parts of Suji, but predominantly the buildings line up in a row facing single direc- The concept of control, originally proposed by tion, creating endless arrays of walls rather than Lynch (1981) as a performance dimension for city forming neighborhood clusters. form, is defined as the extent to which users of a The high-rise apartment buildings also domi- place regulate it. Southworth (2003) narrowed the nate the three-dimensional built form pattern of meaning down as the degree to which the land is the two areas. Their bulk, height and arrangement under public ownership or public control. The determine major view corridors and skylines, difference between shares of publicly and privately affecting residents’ visual perception. In Bun- owned lands signifies the progress toward privati- dang, the overall urban landscape design was a zation of land. Although this dimension may seem key element in its original plan. Buildings vary to disadvantage Suji from the outset since the in their heights and secure view corridors to private sector played a bigger role than the public,

Figure 10: Built form patterns in B1 and B2 in Bundang and S1 and S2 in Suji. Squares are 1.6 km on each side.

Figure 11: Three-dimensional built form pattern in Bundang: (a) Variation in building heights; (b) secured view corridors to greenery; (c) blocked view corridors by monotonous skyline.

Figure 12: Three-dimensional built form pattern in Suji: (a) Endless arrays of high-rise apartment buildings: (b) uncomfortable cohabitation of agricultural and development landscapes; (c) dwarfed traditional landscape.

Figure 13: Comparison of control: Public use versus private land control in B1 and B2 in Bundang and S1 and S2 in Suji. Squares are 1.6 km on each side. it still gives relevance to the discussion on decline public transit service and use and residential satis- of the public realm due to privatization of cities. faction. By doing so, we are able to understand the The share of publicly owned land or those under performance in citizens’ daily life of the two areas public use in Bundang is higher than in Suji. 53.8 that may provide complementary perspectives to per cent of B1 and 58.9 per cent of B2 are under planners’ bird’s eye view and the regularity of public control, while only 38.2 per cent of S1 and physical form analysis. 28.2 per cent of S2 are public (Figure 13). In this calculation, we regarded all publicly owned lands as public, thus including public open spaces, public Public transit service and use rights-of-way, and parcels owned by public entities and governments. We also regarded schools, both Bundang has a better public transit service than public or private, as public. The larger private Suji. Its residents enjoy more bus routes that serve land ownership in Suji implies that the control of more frequently at nearer stops (Table 5). As of the public sector in its urban environment is 2014, Bundang provides almost twice the number relatively limited. This may lead to higher prob- of local and regional bus routes in service and four ability of individualism and marginalization times as many stops than Suji. Average bus service of public spaces in Suji, as literature suggests intervals during both peak and non-peak hours are (Ellin, 1996; Southworth and Parthasarathy, 1996; shorter in Bundang. Residents experience shorter Madanipour, 2004). travel times to reach the nearest bus stop. While most bus services in Bundang came together with the construction of its residential neighborhoods, Residents’ Livability and Satisfaction Suji was without proper bus service for years. Bus operators, mostly private, were reluctant to serve We investigated resident’s livability and satisfac- many neighborhoods of Suji due to insufficient tion in Bundang and Suji with a specific focus on profitability (Gyeonggi Research Institute, 2003).

Table 5: Public transit service and access and modal split in Bundang and Suji

Mode Service indicator Bundang Suji

Public transit service and access Bus Number of bus routesa 76 46 Number of bus stops per 100 Hab 21 5 Average interval during peak hours and 13 and 19 15 and 35 non-peak hours (minutes)a Perceived average time to reach the nearest bus 3.1 3.3 stop (minutes)c Subway Number of subway stationsd 71 Population per stationd 56 920 299 195 Annual rides per personc 200 36 Perceived average time to reach the nearest 10.6 26.8 subway station (minutes)c Modal Splitc Commute travel Private vehicle 52.7% 60.3% Bus 28.2% 27.5% Rail 10.6% 5.3% Walk or bike 7.7% 6.2% Other modes 0.8% 0.8% Leisure travel Private vehicle 41.7% 48.1% Bus 21.2% 20.7% Rail 5.9% 3.8% Walk or bike 30.9% 27.1% Other modes 0.3% 0.4% aAs of year 2014. bAs of year 2012. cAs of year 2010. dAs of year 2011. Sources: Gyeonggi-Do Bus Information System (www.gbis.go.kr/); Korea Railroad Corporation, Korea Airport Railroad and Korea Rail Network Authority (2012); Korea Statistical Information Service (kosis.kr); National Transport Survey 2010 (www.ktdb.go.kr); Seongnam City (2013); Yongin City (2013).

Bundang also outperforms Suji in its rail service. population using private vehicles for both com- Through state-level decision making, Bundang’s mute and leisure purposes in Suji is higher than master plan included establishment of the Bun- that in Bundang. In turn, residents in Suji take dang Line that connects Bundang directly to the public transit, walk, and bike less. SMA’s regional rail system since 1994. The New Bundang Line, also connecting Bundang to Seoul, began its service in 2011. As of 2015, seven rail Residents’ satisfaction stations along the two lines serve Bundang. Absence of requirements for the state, local The Social Survey of South Korea that is carried out government, or private developers to provide rail at the city level by the local governments every two service has long kept Suji’s residents from enjoying years presents information on the overall satisfac- the service. In 2004, one temporary station opened tion on the residential environment. The survey directly southeast of Suji, as a southern extension provides a summary of residents’ responses to of the Bundang Line. It was only in 2007, more various aspects of their social life which includes than 10 years after Suji’s population started to soar, income and consumption, education, public a proper rail station opened in the eastern part of health, residence and transportation, and safety. Suji. Although a further extension of the New Survey results in 2013 showed that 62.2 per cent of Bundang Line that stops at four stations in Suji residents in Bundang were satisfied with their will begin service in 2016, residents of Suji today residential environment while 56.1 per cent in Suji travel perceivably more than twice as much time as were satisfied (Figure 14). A total of 12.6 per cent in those of Bundang to reach the nearest rail station Bundang and 14.1 per cent in Suji were dissatisfied. (Table 5). This result is in line with findings from studies that Unsurprisingly, Suji shows a higher private identified that more affluent households tend vehicle use rate (Table 5). The proportion of the to be more satisfied with their neighborhoods

100% coordination and civic engagement. Nevertheless, 16.6% 21.7% Bundang’s superiority over Suji in various dimen- 80% sions of urban form and the resulting livability and Highly Satisfied 39.5% satisfaction successfully represent the benefits and 60% 40.5% Slightly Satisfied effectiveness of public planning in shaping subur- Neutral 40% bia. With the heavy involvement of the state in Slightly Dissatisfied 29.9% initiating the planning and development of Bun- 25.2% 20% Highly Dissatisfied dang, a set of principles of suburban community ’ 11.8% 11.9% planning were established. The public sector s 0% 0.8% 2.2% holistic approach allowed better strategies on land Bundang Suji use, circulation and urban design to be implemen- Figure 14: Residential satisfaction responses of Bundang and ted. The new town was able to supply efficient Suji residents in 2013.Sources: Seongnam City (2013); Yongin City (2013). transport infrastructure, accessible public spaces and amenities for the residents. In contrast, Suji shows many negative signs of (Hannscott, forthcoming; Lu, 1999; Mohan and sprawl which Bruegmann (2005) defined as ‘low- Twigg, 2007). However, as Parkes et al (2002) density, scattered, urban development without suggests, the quality of physical environment may systematic large-scale or regional public land use be an influential factor in determining residents’ planning’ (p. 18), and which Ewing (1994) satisfaction. described as being characterized by poor accessi- Another indicator that illustrates residents’ satis- bility and lack of functional open space. The faction is housing price. Major real estate database circumstances of Suji may differ from those of providers in South Korea, such as Naver Real sprawling American suburbs. However, Suji’s Estate and Real Estate 114, reported that the development approach, problems and concerns as average housing price per unit area as of 2013 in identified in this study make us concerned about Bundang was higher than in Suji by 46.1–61.5 per the higher probability that the ill effects of sprawl cent, whereas the annual household income of would iterate in places like Suji. These include Bundang’s residents was higher than those in Suji excessive automobile dependency (Newman and by 7.4 per cent (Table 1).11 From this, we can infer Kenworthy, 1999), fractured regional structure that home buyers are willing to pay more to live in (Barnett, 1995), deterioration of spatial quality a smaller unit in Bundang in order to enjoy the (Kunstler, 1993; Kelbaugh, 2002), segregated land better quality of its urban environment.12 use and increased privatization (Southworth and Owens, 1993; Gillham, 2002), monotonous every- day life and loosening of social ties (Baldassare, Concluding Reflections 1986; Putnam, 2000), and public health problems (Frumkin et al, 2004). Our comparative analysis demonstrates that Bun- This study reminds us of the long-lasting debate dang, the state-planned suburb in the SMA, in of market versus planning, which seems less active comparison with Suji, the market-led, performs these days but is still alive. Rooted in conservative better, with some variations, in providing systema- political traditions and neo-classical economics, tic arrangement of land uses, well-connected street advocates of market-determinism view planning patterns, diversity of quality public open spaces, as a less productive, authoritarian encumbrance clustered and controlled built form patterns, and (Klosterman, 1985; Banerjee, 1993a). They argued less privatization of land control. It also presents that ‘market approaches to planning have much enhanced livability supported by a larger share of better prospects than command-and-control and travel relying on public transit and non-motorized regulatory approaches’ (Richardson and Gordon, modes and a higher level of overall satisfaction by 1993, p. 348) and criticized that the benefits of anti- its residents. sprawl planning measures are often exaggerated However, Bundang is not without any faults. It (O’Toole, 2000; Gordon and Richardson, 2001, is essentially a large-scale greenfield development 1997, 1989; Cox, 2002). However, as counter-criti- that destroys farmlands and the natural environ- cized by Banerjee (1993b), who questioned the ment. Its physical landscape is dominated by effectiveness of market approaches on distribution arrays of cookie-cutter residential buildings. Its and securing rights and welfare, planning serves as planning and development process lacked regional a guide for the society, while the marketist view is

‘devoid of the values and ideals that go beyond subdivisions of a province. A city refers to an urban area efficiency’ (p. 356). This view was reinforced by that can be subdivided into districts if the population reaches more than 500 000. A county, more rural than Peiser (1984), Thompson (1993) and Ewing (1997) fi fi urban, is subdivided into either towns or townships, towns who empirically identi ed the bene ts of planning being small urbanized areas and townships rural. As of over market-driven or unplanned land policies. 2013, Bundang District is a subdivision of Seongnam City. Our comparative study of Bundang and Suji Suji District is a subdivision of Yongin City. supports the value of planning and provides 4 In the 1990s, the Korea Land Corporation executed three Housing Site Development Projects on 550 Ha of land in Suji, implications to especially those who face retro- fi which corresponds to 13 per cent of the total area. The tting of suburbs and establishing suburban cen- projects were an ex post facto measure to the uneven ters as key tasks at hand. In countries that suburban development led by private developers and institutionalize state-driven development policies, targeted on supplying residential lands in a more planned we suggest that the public sector should take more manner. Less attention was paid to providing commercial/ office and civic uses for self-containment in these projects, control in shaping suburbia to protect their resi- which differed from the state-led new towns (Hwang, 2000). dents from the ill effects of sprawl and seek an 5 For example, developing a residential community of 500 alternative future. In other parts of the world households required at least 12-m wide entry roads and was where the private sector plays a more proactive not obliged to provide childcare and recreational facilities. role, we propose that comprehensive master plan- On the other hand, a community of 1000 households mandated at least 15-m wide entry roads and supply of the ning and investment should take place. What fi two facilities (Korea Land Corporation, 2000). should not be forgotten is the signi cance of 6 According to Chung and Kim (2003), Yongin’s Sprawl Index, planning, whether by public or private sector, that defined as the rate of population growth divided by the rate cares about the public interest in making our cities of road area growth, between 1995 and 2000 was the second and regions more livable. highest among 206 jurisdictions of South Korea. 7 We also found that Bundang is a major employment center for Suji’s residents. A total of 12.9 per cent of Suji’s commuters work in Bundang. Acknowledgements 8 Southworth (1997) analyzed form and pattern, character, land use patterns, public open space, street design and This research was partially funded by the Pedes- circulation patterns, pedestrian access, transit, suitability for trian Environment Evaluation Project of the Archi- children, teens, and elderly, and market success. Lee and ’ tecture and Urban Research Institute, South Korea. Ahn (2003), elaborating Southworth s (1997) approach, fi examined form and pattern, land use pattern, public open We thank Suji District Of ce for sharing their space, housing, street pattern and circulation design, data with us and colleagues at the Lee Kuan Yew pedestrian and vehicle access, and unit and block plan. Centre for Innovative Cities, Singapore University 9 Lynch’s (1981) performance dimensions included vitality, of Technology and Design for their generosity. sense, fit, access, and control; and two meta-criteria which fi We are also grateful to Dominic Nishantha Cooray, are ef ciency and justice. Southworth (2003) suggested the following as measurable attributes of livable urban Susan Kinloch, Wee Kiat Lim, Michael Southworth environment: access, street and path systems, built form, and the anonymous reviewers for their excellent public spaces, activity, natural factors, views, control and suggestions for improving the manuscript. others (p. 346). 10 B1 corresponds to Sunae neighborhood, and B2 to Geumgok and Gumi neighborhoods in Bundang. S1 relate to Notes Poongdeokcheon neighborhood, and S2 to Sanghyeon neighborhood in Suji. 11 The South Korean Population Census or any other type of 1 The Korea Land Corporation was previously known as the social survey carried out by the public sector does not collect Korean Land Depository and Korea Land Development any information about housing price. Corporation. In 2009, the Korea Land Corporation merged 12 Another possible interpretation of the lower housing price in with the Korea National Housing Corporation and became Suji may be that the market-led suburban development can the Korea Land and Housing Corporation (LH Corporation). lead to more affordable housing options. However, in the 2 These population sizes may not have been the typical case long term, it imposes increased financial burden on the for suburbs in some parts of the world. However, they are public sector, as demonstrated in Suji, especially on quite frequently observed in many Asian countries such as infrastructure supply, as well as higher commuting costs Japan, Hong Kong, Singapore and China, those that have and congestion. gone or are currently going through rapid economic development and urbanization. 3 As a result of rapid population increase, Suji, originally a township of Yongin County, was promoted to a town while References Yongin County became Yongin City in 1996. In 2005, Suji was promoted to a district of Yongin City. Administratively Alterman, R. (eds.) (2001) National-Level Spatial Planning in in South Korea, city and county are independent Democratic Countries: An International Comparison of City and

Regional Policy-Making. Liverpool, UK: Liverpool University Gordon, P. and Richardson, H.W. (1989) Gasoline consumption Press. and cities: A reply. Journal of the American Planning Association American Planning Association. (2006) Planning and Urban 55(3): 342–346. Design Standards. Hoboken, NJ: John Wiley & Sons, Inc. Gordon, P. and Richardson, H.W. (1997) Are compact cities a Bae, Y. and Sellers, J.M. (2007) Globalization, the developmental desirable planning goal? Journal of the American Planning state and the politics of urban growth in Korea: A multilevel Association 63(1): 95–106. analysis. International Journal of Urban and Regional Research Gordon, P. and Richardson, H.W. (2001) The sprawl debate: Let 31(3): 543–560. markets plan. Publius 31(3): 131–149. Baldassare, M. (1986) Trouble in Paradise: The Suburban Transfor- Gyeonggi Research Institute. (2003) Study on Yongin Region. mation in America. New York: Columbia University Press. (No. 15) Suwon, South Korea: Gyeonggi Research Institute, Banerjee, T. (1993a) Antiplanning undercurrents in US planning [in Korean]. education: Antithesis or ideology? Environment and Planning Hall, P. (2002) Cities of Tomorrow: An Intellectual History of Urban B 20(5): 519–536. Planning and Design in the Twentieth Century, 3rd edn. Banerjee, T. (1993b) Market planning, market planners, and Malden, MA: Blackwell Publishing. planned markets. Journal of the American Planning Association Handy, S.L., Paterson, R.G. and Butler, K. (2003) Planning For 59(3): 353–360. Street Connectivity: Getting from Here to There. Chicago, IL: Barnett, J. (1995) The Fractured Metropolis: Improving the New American Planning Association. City, Restoring the Old City, Reshaping the Region. New York: Hannscott, L. (forthcoming) Individual and contextual socio- Icon Editions. economic status and community satisfaction. Urban Studies, Bruegmann, R. (2005) Sprawl: A Compact History. Chicago, IL: in press, doi: 10.1177/0042098015574811. University of Chicago Press. Harris, R. (2010) Meaningful types in a world of suburbs. In: Chapin, T.S. (2012) Introduction: From growth controls, to M. Clapson and R. Hutchinson (eds.) Suburbanization in comprehensive planning, to smart growth: Planning’s emer- Global Society. Bingley, UK: Emerald, pp. 15–50. ging fourth wave. Journal of the American Planning Association Hayden, D. (2004) Building Suburbia: Green Fields and Urban 78(1): 5–15. Growth, 1820–2000. New York: Pantheon Books. Chung, H. and Kim, J. (2003) The spatial distribution of urban Heid, J. (2004) Greenﬁeld Development without Sprawl: The Role of sprawl in metropolitan regions. Journal of Korea Planners Planned Communities. Washington DC: Urban Land Institute. Association 38(5): 7–20, [in Korean]. Hill, R.C. and Kim, J. (2000) Global cities and developmental Cox, W. (2002) Smart Growth and Housing Affordability. Washing- states: New York, Tokyo and Seoul. Urban Studies 37(12): ton DC: Millennial Housing Commission. 2167–2195. Demographia. (2015) Demographia world urban areas (Built- Hwang, K. (2000) Causes and solutions of sprawl in yongin Up urban areas or world agglomerations) 11th annual region. Land Research 11(3): 115–132, [in Korean]. edition, http://www.demographia.com/db-worldua.pdf, Jackson, K.T. (1985) Crabgrass Frontier: The Suburbanization of the accessed 13 July 2015. United States. New York: Oxford University Press. Ellin, N. (1996) Postmodern Urbanism. New York: Princeton Jacobs, A. and Appleyard, D. (1987) Toward an urban design Architectural Press. manifesto. Journal of the American Planning Association 53(1): Ewing, R. (1994) Characteristics, causes, and effects of sprawl: 112–120. A literature review. Environmental and Urban Studies Jun, M.J., Kim, J.I., Kwon, J.H. and Jeong, J.E. (2013) The effects 21(2): 1–15. of high-density suburban development on commuter mode Ewing, R. (1997) Is Los Angeles-style sprawl desirable? Journal choices in Seoul, Korea. Cities 31: 230–238. of the American Planning Association 63(1): 107–126. Kelbaugh, D. (2002) Repairing the American Metropolis: Common Fishman, R. (1987) Bourgeois Utopias: The Rise and Fall of Place Revisited. Seattle, WA: University of Washington Press. Suburbia. New York: Basic Books. Kim, H. and Han, S. (2012) Seoul. Cities 29(2): 142–154. Forsyth, A. (2005) Reforming Suburbia: The Planned Communities Klosterman, R.E. (1985) Arguments for and against planning. of Irvine, Columbia, and the Woodlands. Berkeley, CA: Univer- The Town Planning Review 56(1): 5–20. sity of California Press. Knox, P.L. (2008) Metroburbia, USA. New Brunswick, NJ: Forsyth, A. (2012) Deﬁning suburbs. Journal of Planning Litera- Rutgers University Press. ture 27(3): 270–281. Korea Land Corporation. (1997) Development History of Bundang Forsyth, A. (2014) Global suburbia and the transition century: New Town. Seoul, South Korea: Sung In, [in Korean]. Physical suburbs in the long term. Urban Design International Korea Land Corporation. (2000) Study on the Sprawling Condi- 19(4): 259–273. tions of Seoul Metropolitan Area and Schemes to Rationalize Land Frank, L. and Engelke, P. (2005) Multiple impacts of the built Policy. Seongnam, South Korea: Korea Land Corporation, [in environment on public health: Walkable places and the Korean]. exposure to air pollution. International Regional Science Review Korea Railroad Corporation, Korea Airport Railroad, Korea Rail 28(2): 193–216. Network Authority. (2012) Statistical Yearbook of Railroad. Frumkin, H., Frank, L. and Jackson, R. (2004) Urban Sprawl and Daejeon, South Korea: Korea Railroad Corporation, [in Public Health: Designing, Planning, and Building for Healthy Korean]. Communities. Washington DC: Island Press. Kunstler, J.H. (1993) The Geography of Nowhere: The Rise and Garreau, J. (1991) Edge City: Life on the New Frontier. New York: Decline of America’s Man-Made Landscape. New York: Anchor Books. Touchstone. Gibberd, F., Harvey, B.H. and White, L. (1980) Harlow: The Story Kwon, Y. (2003) Solving the problems of the Seoul metropolitan of a New Town. Stevenage, UK: Publications for Companies. region and the construction of the new administration capital Gillham, O. (2002) The Limitless City: A Primer on the Urban city. Journal of the Korean Geographical Society 38(2): 324–334, Sprawl Debate. Washington DC: Island Press. [in Korean].

Lang, R.E. and LeFurgy, J.B. (2007) Boomburbs: The Rise of America’s Phelps, N.A. and Wu, F. (eds.) (2011) International Perspectives on Accidental Cities. Washington DC: Brookings Institution Press. Suburbanization: A Post-Suburban World? New York: Palgrave Lee, C. and Ahn, K. (2003) Is Kentlands better than Radburn?: Macmillan. The American garden city and new urbanist paradigms. Putnam, R.D. (2000) Bowling Alone: The Collapse and Revival of Journal of the American Planning Association 69(1): 50–71. American Community. New York: Simon & Schuster. Lee, C. and Ahn, K. (2005) Five new towns in the Seoul Richardson, H.W. and Gordon, P. (1993) Market planning metropolitan area and their attractions in non-working trips: oxymoron or common sense? Journal of the American Planning Implications on self-containment of new towns. Habitat Association 59(3): 347–352. International 29(4): 647–666. Roh, T. (2000) Sprawl and Yongin. Yongin Region Research Lee, J. and Oh, D. (2012) Housing quality evaluation and 1: 45–65, [in Korean]. housing choice using PIF: A case of the Bundang new town Scheer, B.C. and Petkov, M. (1998) Edge city morphology: housing market in Korea. International Journal of Urban A comparison of commercial centers. Journal of the American Sciences 16(1): 63–83. Planning Association 64(3): 298–310. Lee, Y. and Shin, H. (2012) Negotiating the polycentric city-region: Seongnam City. (2013) The Fourth Social Survey in Developmental state politics of new town development in the Seongnam City. Seongnam, South Korea: Seongnam City, Seoul capital region. Urban Studies 49(6): 1333–1355. [in Korean]. Lu, M. (1999) Determinants of residential satisfaction: Ordered Southworth, M. (1997) Walkable suburbs?: An evaluation of logit vs. regression models. Growth and Change 30(2): 264–897. neotraditional communities at the urban edge. Journal of the Lynch, K. (1981) Good City Form. Cambridge, MA: The MIT American Planning Association 63(1): 28–44. Press. Southworth, M. (2003) Measuring the liveable city. Built Envir- Madanipour, A. (2004) Marginal public spaces in European onment 29(4): 343–354. cities. Journal of Urban Design 9(3): 267–286. Southworth, M. and Owens, P.M. (1993) The evolving metropo- Meier, R.L. (1970) Exploring development in great Asian lis: Studies of community, neighborhood, and street form at cities: Seoul. Journal of the American Institute of Planners 36(6): the urban edge. Journal of the American Planning Association 378–392. 59(3): 271–287. Mohan, J. and Twigg, L. (2007) Sense of place, quality of life and Southworth, M. and Parthasarathy, B. (1996) The suburban local socioeconomic context: Evidence from the survey of public realm I: Its emergence, growth and transformation in English housing, 2002/03. Urban Studies 44(10): 2029–2045. the American metropolis. Journal of Urban Design 1(3): Newman, P. and Kenworthy, J. (1999) Sustainability and Cities: 245–263. Overcoming Automobile Dependence. Washington DC: Island Thompson, G.L. (1993) Planning beats the market: The case of Press. paciﬁc greyhound lines in the 1930s. Journal of Planning O’Toole, R. (2000) The Vanishing Automobile and Other Urban Education and Research 13(1): 33–49. Myths: How Smart Growth Will Harm American Cities. Bandon Urban Design Institute of Korea. (2012) Urban Design in Korea. OR: Thoreau Institute. Seoul, South Korea: Bosunggak, [in Korean]. Oxford Brookes University. (2006) Transferable Lessons from the Wheeler, S.M. (2008) The evolution of built landscapes in New Towns. London: Department of Communities and Local metropolitan regions. Journal of Planning Education and Government. Research 27(4): 400–416. Parkes, A., Kearns, A. and Atkinson, R. (2002) What makes Yongin City. (2013) The Seventh Social Survey in Yongin City. people dissatisﬁed with their neighborhoods? Urban Studies Yongin, South Korea: Yongin City, [in Korean]. 39(13): 2413–2438. Zacharias, J. (2005) Generating urban lifestyle: The case of Hong Peiser, R. (1984) Does it pay to plan suburban growth? Journal of Kong new-town design and local travel behaviour. Journal of the American Planning Association 50(4): 419–433. Urban Design 10(3): 371–386.