1. Introduction
Walking within the city has been overlooked, undervalued, or not adequately recognized as an important mode of movement with the appearance of car and development of high-speed road transport since the 1920s [
1,
2,
3]. In fact, it was reported that car-oriented transportation systems and land use policies nowadays have made walking less convenient and more dangerous, and pedestrian accessibility has declined steadily in conjunction with urban development in a large number of cities worldwide [
4,
5,
6]. The high dependence on car as means of transportation is a widely recognized urban issue, as it is associated with increasing urban sprawl and low density of land use, high dependence on energy, increasing atmospheric and noise pollution, and long commuting time caused by traffic congestion. There has thus been high level of advocacy to reduce car dependence and instead plan and design cities to be more walkable. Promoting walkability has become the main focus of environmental and policy initiatives in a number of cities in recent decades [
5,
7,
8,
9], which is not surprising as walking plays an important and unique role in an efficient transport system [
1,
3]. It provides benefits in mobility, physical activity, psychological pleasure, health, social connection, recreation, environment protection, energy conservation, etc. [
6,
10,
11]. Therefore, walkable cities have been advocated, and walkability has been treated as the foundation of a sustainable and livable city [
6,
10].
The concept of walkability essentially is about how conductive the built environment is for walking to various destinations [
5,
6,
10,
12,
13]. Walkability captures the proximity between functionally complementary land uses and the directness of a route, or the connectivity between destinations from the perspective of accessibility [
12], which is often associated with factors like street width, street connectivity, pedestrian crossings, number of lanes, distribution of facilities, safe speed, etc. [
5,
7,
9,
10,
14]. Apart from the accessibility perspective, walkability also can be defined as a pleasant walking environment, which is influenced by a number of factors such as cleanness of streets, safe crossings, sense of security, appearance of street trees, lighting at night, etc. [
3,
6,
8,
15,
16].
Accordingly, within the spectrum of walkability assessment, previous studies have focused on assessing pedestrian environment, which was either assessed subjectively by respondents or objectively by researchers [
8,
15,
17,
18]. In these studies, checklists assessment, stated preference survey, on-street survey, and on the move survey are commonly used methods to obtain pedestrians’ perceptions on walking environment [
8,
18]. There are also studies using multi-layer geographic information data to objectively assess walkability level in GIS, including residential density, street connectivity, land-use mix, intersections, land use accessibility and diversity of uses, net retail area, etc. [
5,
7,
9,
14,
19,
20,
21].
Walkability also can be assessed from the perspective of accessibility to daily facilities [
13,
22]. The literature indicates that accessibility-based factors are more indicative of walkability level and have a stronger effect on walking behavior [
10,
23]. Specifically, if a certain area is more supportive of walking to reach various daily facilities, then the walkability level in this area would be higher [
22]. This perspective is in line with the essence of the concept of walkability, since a walkable area should provide opportunities to reaching various destinations [
13]. This approach is actually a proxy of land-use mix and street connectivity, which were used to represent the diversity of land uses and facilities in some previous studies [
7,
14,
18]. In other words, walkability assessment based on distribution of public facilities (i.e., facility-based walkability assessment) treats accessibility and diversity of services as the leading predictor of walking behavior. Facility-based walkability assessment is superior in assessing and benchmarking walkability, using geospatial data in a direct and replicable manner, compared with measures of surveys, self-reporting perceptions, audits, and observational data employed in previous studies [
10]. Yet, while the existing literatures provide information and methodology for walkability assessment, facility-based walkability measurement is not widely used and needs further exploration. Walk Score
TM is a typical walkability measurement based on the assessment of accessibility to daily facilities. Walk Score was developed by Frontlane in partnership with academics [
24], and has been validated as an appropriate measurement of walkability [
24,
25,
26,
27]. More specifically, Walk Score was found to be significantly positively correlated with commonly used objective measures of walkability (such as street connectivity, residential density, accessibility to public transits, etc.), as well as subjective evaluation of walkable environment attributes [
24]. Therefore, Walk Score has the potential to be used as a comprehensive measurement of walkability, which is also more operational than traditional measurements. However, the Walk Score database currently is only available mainly for U.S. cities and several Canadian cities. Very few studies have been conducted elsewhere in the world, including China. This study applied Walk Score to a rapidly urbanizing district in Shenzhen and explored the spatial and temporal variations in walkability over two time periods. In addition, as walkability is associated with a wide range of urban features, we also investigated the relationship between walkability and real estate value in the study area.
The measurement of walkability has applications in a variety of research fields, including geography, urban planning, architecture, psychology, public health, economic studies, etc. For instance, it has been adopted in many cities around the world to investigate its relationship with people’s walking behavior, physical activities, modes of transportation, life quality, community safety, etc. [
2,
9,
14,
20,
28,
29,
30,
31]. A growing body of studies showed that a community with high walkability tends to lead to high social [
32,
33], economic [
10,
11,
34,
35,
36], environmental [
6,
37], and public health values [
21,
38,
39,
40,
41]. One particular application of walkability assessment in economic domain is the investigation of the relationship between walkability and real estate value [
35]. The economic value of a walkable neighborhood was obtained by encouraging economic transactions and social exchanges [
1,
11,
42]. A growing number of studies have contributed to the debate on the relationship between walkability level and property value [
1,
10,
34,
35,
36,
42,
43,
44,
45,
46]. For instance, it was reported that walkability level is positively correlated with residential estate value in 13 out of 15 housing markets sampled in U.S. metropolitan areas [
34]. Similarly, Pivo and Fisher [
35] found that greater walkability was capitalized into higher apartment values consistently. Gilderbloom [
10]’s study found a positive impact of walkability on housing value. Rauterkus and Miller [
36]’s finding is also in agreement with this positive correlation, and they found that this relationship is stable over time. However, Rauterkus and Miller [
36] reported that in more car-dependent neighborhoods the impact of walkability on residential land values becomes weaker. Likewise, Li, Joh [
45] reported that improvement of walkability yielded a greater housing price increase in a walkable neighborhood than in a car-dependent neighborhood. Apart from positive correlations, insignificant or negative correlations were also observed. Boyle, Barrilleaux [
46] found that walkability’s impact on housing value is statistically insignificant in Miami, Florida. Similarly, in Cortright [
34], two out of the 15 studies housing markets did not show positive correlation between walkability and housing value, and one of them (Las Vegas) exhibited negative correlation, while another one (Bakersfield) showed insignificant correlation.
The studies above demonstrated that there are inconsistencies among different studies, and there are still unanswered questions in the relationship between walkability and real estate value. Several reasons can be postulated, such as socio-economic conditions of the study area, car-dependent level of study area, urban morphology, etc. It thus remains largely unknown that if the positive correlation between walkability and housing value also holds in rapidly urbanizing cities in China. Few studies have been conducted in China to examine the economic impact of walkable neighborhoods, especially in Shenzhen as a typical rapidly urbanizing city where land use, transportation network, and facilities experienced huge changes since 1980s. Unlike some cities in Europe and U.S. with fine-grained pedestrian network, Futian District as the central part of Shenzhen was urbanized rapidly from a small fishing village to a big city in 30 years, from the early 1980s to the late 2000s [
47,
48,
49]. During this period, a large number of main roads with large scales were constructed rapidly to meet the demand of economic increase, which constituted the major components of road system in Shenzhen, with little consideration given to the smaller scale pedestrian design or high-density urban design for creating walkable urban spaces [
48,
49]. As a result, the planning and urban form of Shenzhen is more car-oriented, rather than pedestrian-oriented. The scale of neighborhood and city is also large and not intimate enough for pedestrians. In this context, this study is essential and would shed light on whether Futian District as a rapidly urbanizing area is walkable for pedestrians, and whether there exists positive correlation between walkability level and residential estate value in this area.
Therefore, the key aim of this study is to use modified Walk Score methodology to assess walkability objectively in a rapidly urbanizing district in China and to explore the relationship between walkability and real estate value, based on the distribution of public facilities and real estate value data collected in 2013 and 2018. The specific objectives are: (1) to see how this methodology can be applied in a Chinese context and to explore spatial and temporal variation of walkability level in a rapidly urbanizing district; (2) to examine if the presumed positive correlation between walkability level and residential estate value holds in this area.
4. Conclusions
A walkable environment plays an important role in supporting vibrant and sustainable urban communities. In China, its importance has not been fully recognized in urban planning process and public policy commitments. This study examined the spatial and temporal variation of walkability in year 2013 and 2018 using a modified Walk Score methodology in a rapidly urbanizing district in China, which has seldom been operationalized in a context like this. In addition, the relationship between walkability level and residential estate value was explored. It was found that although on average walkability in Futian District is not high, it has increased from 19.09 in 2013 to 23.59 in 2018. The area of “car-dependent” region has decreased, while the proportion of areas with walkability value above 30 out of built environment has increased considerably from 16% to 45% from 2013 to 2018. In addition, high spatial variation was observed across the study area: one eastern high center and three sub-high centers were identified. As for the relationship between walkability and residential estate value, in contrast with the positive correlations observed in some U.S. cities, negative correlation was observed for data in 2013 and 2018, although the relationship is only statistically significant in 2018. This is mainly attributed to the rapid urbanization process and car-dependent urban form of Futian District, supported by several previous studies with similar findings. Therefore, housing price in Futian District is more influenced by other factors rather than walkability.
Results of this study are of considerable importance in terms of extending current literature on walkability measurement and exploration of the relationship between walkable environment and housing values in a rapid urbanizing Chinese city context. Another key contribution of this study is that discussions on the underlying reasons for the incompatibility of spatial pattern of walkability and housing price provide implications for future planning activities and policy-making. Policies should be made to create walkable environment, improve environmental quality, and guide the development of housing market at the same time. Our findings lend credence to policies encouraging projects that would improve pedestrian infrastructure and walkability level. Local government also could provide incentives for developers to constructing walkable neighborhood. Developers should be willing to develop more walkable residential estates as long as additional expenses for walkable locations and environment do not exceed the benefits brought about by walkability. In addition, the walkability model used in this study has the potential to be applied in other cities in China as a powerful tool for urban planners, policy makers, and the public. A platform can be built to incorporate information on walkability and prices of local residential estates, which is helpful for residents to make informed decision in selecting walkable neighborhoods with acceptable prices.
Although the walkability model used in this study provides an operational measurement of facility-based walkability, it has some limitations. Euclidean distance was adopted in computing the distance between pedestrian and facility, due to the ease of capturing facilities within multiple distance ranges. However, it would be more accurate in measuring walkability if network distance along actual road networks was used. In addition, although the distribution of facilities is the focus of measurement of walkability in this index, it would be more comprehensive to incorporate the assessment of quality of walking environment in future studies, such as aesthetic attributes, cleanliness, safety, presence of roadside trees, pleasant urban design, etc. In addition, we suggest future studies to be conducted in China and other Asian countries, to provide more empirical examples, so as to contribute to the development and application of Walk Score in the Asian contexts.
In conclusion, Shenzhen as a typical rapidly urbanizing city in China requires further research on developing localized walkability measurement, exploring local influential factors on walkability level, investigating specific ways in which walkability has an effect on housing values, as well as approaches to creating a more walkable and sustainable environment.