How to Measure the Impact of Walking Accessibility of Suburban Rail Station Catchment Areas on the Commercial Premium Benefits of Joint Development

Abstract

As the primary solution to the issue of high passenger traffic in urban areas, rail transit has a significant impact on the structural form of cities and regional development. Additionally, it has varying degrees of a premium effect on land value around stations. Current research on the factors influencing the premium effect of rail transit station areas mainly focuses on the macro level of the station area circle, with more attention given to the premium caused by distance and functional differences. Most research objects are typically urban center lines or stations. However, this study focuses on the core area of the station and concentrates on the impact of the construction of integrated station–city facilities on the choice of pedestrian routes and the enhancement of pedestrian accessibility. It also explores whether this enhancement is associated with the premium benefits of ancillary commercial development. To achieve this goal, this paper integrates models from several related studies to conduct a comprehensive assessment. Firstly, it uses a spatial panel econometric model to improve the classical characteristic price method model. It then combines the ideas and models of the cost–benefit analysis method, taking the Odakyu Odawara Line of the Japanese suburban railroad as an example. This analysis explores the mechanism and factors influencing the rent premium of commercial facilities in the suburban rail station area and systematically assesses the combined station–city facilities. The study evaluates the social benefits (enhanced walkability) and economic value (premium value added from commercial facilities) of the combined station–city facilities systematically. The results of the study show that (1) the premium benefits of suburban railroad station area commercial facilities are significantly related to the type of station–city combination facilities, combination mode, and walking time and weakly related to the location factor. Additionally, (2) the results of the cost–benefit valuation analysis based on the Ebina Station verify that a reasonable design of station–city combination facilities can effectively enhance the proximity of commercial facilities to the station and improve the walking accessibility, thus promoting the premium benefits. The study demonstrates that a reasonable design of the combined station and city facilities can effectively enhance the proximity of commercial facilities to the station and improve pedestrian accessibility, promoting premium benefits, which can quickly feed the construction cost of the station and achieve positive revenue in the short term. Therefore, the results of the study provide a quantitative reference for the planning and design of suburban stations.

1. Introduction

Transit-oriented development (TOD) is a type of urban development that enhances the location value of land along a route or around a station, improving accessibility and generating a premium effect on land prices. However, in practice, TOD often faces challenges such as limited development benefits and the recovery of construction costs. To ensure the sustainability of future rail transit construction, capital recovery and operation remain critical issues. TOD can bring additional value, primarily commercial functions, to the urban functions of the station catchment area by strengthening public transportation investment and enhancing the premium effect, which is referred to as “value capture” [1]. Urban design plays a crucial role in justifying the premium effect of integrated rail transit development projects. The Rail + Property Development model, an early integrated development of rail transit stations and cities in Hong Kong, has proven to be effective in addressing the challenges of a high-density built environment and promoting the development of rail transit and urban space [2]. To ensure a swift recovery of rail transit construction costs and avoid operational inefficiencies of ancillary properties, a three-dimensional urban design strategy was employed. This approach enhances the connectivity of station buildings with commercial facilities and residential communities through comprehensive ground-level streets, air corridors, and pedestrian bridges across neighborhoods, improving the proximity of ancillary properties and accelerating the efficiency of premiums to the public. Furthermore, it enables integrated access to various transportation modes [3].

Railway station buildings, whether in the form of underground or elevated stations, remain independent of the original urban space and inevitably cause fragmentation to the surrounding area. Due to challenges such as detours or obstruction of street crossings, there are often situations around the station where the straight-line distance is close, but the actual walking distance is far or difficult. Therefore, a reasonable combination design plan is necessary to optimize the combination path between the station and businesses to avoid the embarrassing situation of being “adjacent but not close to each other.” Additionally, the use of three-dimensional integration facilities can effectively reduce bypass coefficients between development sites and stations, enhance the proximity of affiliated businesses, increase commercial footfall, and expand the premium benefits of joint development businesses. In typical station–city integrated development cases, underground passages and pedestrian bridges directly connected to each station entrance and exit are categorized as combined facilities in the form of line articulation, which are relatively low in construction cost and can be flexibly arranged. They are also the primary form of combined facilities design in most public-transportation-oriented urban areas.

In most cases, suburban stations are less complex than urban centers in terms of construction and surrounding elements, resulting in a large gap between straight-line distance and actual walking accessibility, particularly when the spatial distribution density of stores and street network density are both low in the station catchment area. The “contradiction” between straight-line distance and real walking distance is essentially a measure of “proximity” from a different perspective, directly manifested when the straight-line distance is closer but the real walking distance is farther, with the real walking distance based on the road network determining the real accessibility (Figure 1). In suburban areas, railway stations typically feature a linear combination of facilities designed to reduce pedestrian route directness (PRD) and shorten the actual walking distance. This approach aims to optimize the layout of the station to minimize the inconvenience for passengers and encourage the use of public transportation [4]. Moreover, influenced by the first place, the suburban rail station catchment area usually follows a development pattern of “close to the station and beyond a certain range, the density plummets”. In particular, the commercial premium effect is significantly influenced by the walking distance, and the premium change law based on the linear distance measurement of traditional research ideas can hardly reflect the characteristics of suburban stations and ignores the specific improvement mechanism and effect of the actual walking distance factor on the premium effect. Therefore, it is necessary to analyze the value-added performance of the premium effect of neighboring commercial facilities from a microscopic perspective, highlighting the individual attributes of suburban station stores and using a reasonable evaluation method and calculation model by replacing the straight-line measurement results with the actual walkable distance. In actual projects, enhancing the “proximity” of rail transit stations can facilitate the conversion of traffic flow to commercial flow and magnify the premium benefits of neighboring stores.

Station–city integrated facilities refer to a range of transportation infrastructure that connects railway stations and cities, with the aim of promoting closer integration at the spatial level and improving connectivity and spatial accessibility within urban areas [5]. These facilities encompass a variety of transportation infrastructure, such as flyovers, escalators, and underground passages, designed to enhance the integration of railway stations and cities and improve pedestrian flow and access. Although the use of combined station–city facilities can effectively improve pedestrian accessibility, the scale and form of the combined facility are issues that are easily overlooked. Insufficient service levels provided by such pedestrian facilities can lead to overcrowding, reducing pedestrian flow and negatively impacting commercial efficiency. Conversely, if the facilities are too wide, they can lead to dispersed passenger flow and diluted commercial agglomeration, resulting in wasted resources and a non-negligible cost investment for suburban station operation and investment. Thus, it is necessary to systematically evaluate and analyze the impact and value of such station–city integration facilities in terms of suburban rail transit stations. Such an evaluation will promote the development of station–city integration and improve the theoretical study of pedestrian accessibility.

Based on previous research, this study focuses on the Tokyo suburban railway system and proposes research ideas from an urban design perspective. The contributions of this study are threefold. Firstly, by selecting suburban rail stations as the object of study, it provides insights into the development approach of using public transportation to expand new cities and can bring greater economic benefits to suburban rail stations. Secondly, it reinterprets the design of station–city integration facilities to improve pedestrian accessibility in the process of station–city integration. Lastly, the study quantifies the economic value of combined facilities by introducing the cost–benefit evaluation method, reflecting the economic value of pedestrian accessibility from a microscopic perspective. Although this study falls under the category of pedestrian accessibility research, it offers a new perspective and entry point, providing new ideas for existing theoretical research and references for related practical projects. The remainder of this study is structured as follows: Section 1 provides a literature review, Section 2 presents the research methodology, Section 3 describes the study variables, Section 4 analyzes the relationship between station and city facilities and the impact of business premiums, Section 5 discusses the results, and Section 6 summarizes the study. Figure 2 show the study’s workflow.

1.1. Premium Effect from Rail Transit

Over the past few decades, scholars from Europe and America have actively investigated the extent of environmental impacts and premium effects of rail transit in mega-cities, such as London, Paris, and New York—where rail transit was first introduced. Recently, due to rapid urbanization in developing countries, the majority of cities have adopted a TOD (transit-oriented development) model, with rail transit as the backbone of development. As a result, the number of related studies in China and Southeast Asia has increased. These studies focus primarily on two areas: (i) the impact of rail transit on the premiums of different types of functional land and (ii) the extent and degree of impact on premiums caused by rail transit.

Rail transit has diverse impacts on land values of various land types. Previous studies have clearly categorized urban functions into standard categories, including residential, industrial, commercial, retail, education, entertainment, recreation, health, transportation, government, community, parking, vacant, and hospitality [6,7]. While a clear division of land types enables more precise measurement of the effects of rail transit on each function, an overly detailed classification can make the analysis cumbersome and obscure the presentation of findings. Most existing studies on the premium effect of rail transit on different functions have classified them into five main categories: residential, commercial, industrial, health, and education. Earlier studies have shown that the impact of rail transit stations on commercial real estate is greater than on residential real estate within the proximity of the station [8,9]. Research on the premium effect of residential properties shows that the price of residential properties and the type of housing will result in different premium benefits [10,11]. The premium effect of rail transit on industrial land is usually linked to its location in the urban area. Industrial land located near the city outskirts is significantly more impacted by rail transit, leading to a premium effect compared to land located in the city center [12]. However, due to the unique characteristics of industrial land, such as its large area and low demand for land prices, the premium for industrial land is generally minimal. Rail transit has a positive effect on house prices near schools, while it has a negative effect on hospitals [13]. However, education and health land typically covers a large area and is often a government project, making it challenging to reflect the premium effect in real estate value. In contrast, the premium effect on commercial real estate is influenced by macro market regulations, urban development policies, economic conditions, as well as surrounding environmental factors, public facilities, and crowd activities. Thus, the premium effect of rail transit is directly reflected in the rent of stores and the price of goods in commercial facilities.

The majority of existing studies have focused on the “sphere of influence” of railway stations, which refers to the area of impact and is visualized through the radiative effects of the station on distance, such as its economic impact. Most of these studies have concluded that land value decreases as distance from the station increases [14,15]. While some studies have observed opposing or fluctuating growth in specific distance ranges, the overall trend remains negative [16]. Studies examining the degree of premium effect have shown clear differences between European and American countries and Asian countries [6,17,18,19]. In multi-center cities, the premium of land surrounding rail transit is higher compared to single-center cities, and the premium of stations passing through multiple lines is more apparent [20]. Furthermore, the degree of premium effect caused by rail transit in Asian countries with higher population density is generally higher than that in European and American countries with lower population density. Additionally, the degree of premium effect is significantly higher in developed countries than in developing countries.

As for the research method for analyzing the premium effect, ANSELIN et al. [21] previously used the spatial econometric regression method to address the issue of errors, which has since been widely adopted in research related to rail transit and land development. With the advancement of research, more scientific and comprehensive methods have emerged, among which the hedonic price method [22,23] is one of the most typical. This method explains the spatial dependence generated by real estate samples in geographic space and shows spatial heterogeneity, which is relatively more powerful than the traditional characteristic price model. The explanatory power of spatial data is stronger, and the analysis results are more accurate. The spatial Durbin model and the geographically weighted regression model take into account the differences in spatial attributes, and the introduction of the spatial econometric model also makes the analysis results more quantitative and objective [16,17,24].

As urban industrial development transforms and built environments become more dense, the TOD (transit-oriented development)-model-oriented construction approach accelerates the high-intensity development of the rail transit station catchment area, leading to high functional diversity and a complex system with enhanced production values. To match the value, a large number of commercial functions are usually clustered near the stations, while the noise shielding of commercial buildings from the station core can provide a good environment for residential property development further away from the stations. The premium effect of rail transit on residential property is limited by the class of the residential property, and the lower the value of the residential property, the greater the premium effect. Conversely, the value of industrial property changes the least due to rail transit. Research results show that the area influenced by rail transit stations has a premium effect within roughly a 2 km radius circle, and the degree of influence on the surrounding area generally tends to rise, fall, rise again, and finally disappear. Comparing developed countries such as Europe and the United States with developing countries such as China and Southeast Asia, the premium effect due to rail transit is weaker in developing countries, which can be attributed to the level of land use in cities.

1.2. Walking-Accessible Impact Effects and Integrated Assessment

Luca Bertolini argues that the complexity of urban areas where rail transit stations are located is increasing due to the development of cities, the increase in mobility, the accumulation of activities, and the accumulation of conflicts between the relevant participants [25]. To avoid interference between the high-density built environment and pedestrian movement after leaving the station, attention is gradually being paid to the three-dimensional articulation of stations and urban space in the development of the station–city integration model. Three-dimensional pedestrian networks are being introduced into urban space as a medium space to connect different functional buildings and independent neighborhoods or to guide the flow of people. Currently, the impact of pedestrian accessibility on the premium effect can be divided into the improvement of pedestrian accessibility by 3D pedestrian networks and the impact of pedestrian accessibility improvement on the premium effect. Research cases are mostly focused on the influence area of rail transit stations, with a focus on people as the main actor and their actual walking behavior and routes to various destinations. The study covers the 3D walking network, actual walking distance, and the impact of different road network patterns on the premium.

The three-dimensional walking network can improve the walking accessibility and increase pedestrians’ willingness to choose routes by reducing the OD distance [26]. Enhancing the integrity of the pedestrian network can counter the effects of traffic congestion, and the city’s surface streets should be integrated with underground spaces, above-ground flyovers, and public transportation infrastructure [27]. Due to Hong Kong’s high-density urban construction and well-developed pedestrian system in the city center, more studies have been conducted on actual case studies, using Hong Kong as an example [28,29,30]. A continuous three-dimensional network consisting of a pedestrian bridge system, underground passages connected to subway stations, and paths connected to shopping centers can significantly improve pedestrian accessibility when well-connected [31], and the results of the study provide information for inclusive planning and design. The results of this study provide information for inclusive planning and design.

The theoretical basis for the conclusion that improved walking accessibility can effectively contribute to increased land values is derived from the locational land rent theory and the relationship between distance and value. Walking accessibility is used as a measure of land accessibility, and the benefits of better walking accessibility can be capitalized into higher office, retail, and apartment values [32]. Previous studies have examined the impact of enhanced walking accessibility on real estate premium effects in terms of walking distance [33], street layout [34], commute time [35], and other factors. These studies have shown that proximity to rail stations can provide significant benefits to residents within walking distance, that street layout has a substantial impact on price, and that pedestrian-oriented streets have a greater impact than car-oriented streets, especially for newer housing. The premium effect of improved walking accessibility is better reflected in new housing. Street networks can influence pedestrian flows, and these flows are highly correlated with neighborhood walking accessibility [36]. In a study measuring the premium effect generated by walking accessibility, Miko Kitamoto et al. developed a quantitative method for mastering street networks by replacing the street network with links and nodes and providing topographic information on the links to find centrality between streets [37]. Residential rent levels are characterized by significant spatial dependence and differentiation, and bus stops with high pedestrian accessibility can significantly increase residential rents [38].

The increase in pedestrian accessibility around rail transit stations can lead to a corresponding increase in the value of the surrounding real estate, particularly in neighborhoods that are primarily walkable. The impact of pedestrian accessibility varies across different functional categories, with a significant premium effect on surrounding businesses, but a negative effect on functional categories that require quiet spaces, such as hospitals. The implementation of a three-dimensional pedestrian network system around rail transit stations can ease large passenger and improve pedestrian accessibility, further enhancing the construction value of the surrounding area. However, theoretical research has not yet explored the extent to which the combined station–city facility can improve walking accessibility and its economic impact.

In summary, the main contributions of this paper are presented as follows:

(1)

Firstly, from a research object perspective, there are limited studies on the economic law changes based on the characteristics of suburban stations. International experience has shown that in order to develop satellite or new cities, large cities need to rely on transportation corridors to connect, and suburban stations usually become important nodes of the regional transportation network and the starting point of new city development, facing significant tidal phenomena. Therefore, using suburban stations as the research object is highly significant.

(2)

Secondly, regarding the study’s content, most existing studies discuss the influencing mechanisms of the real estate premium effect, with conclusions generally summarizing macroscopic laws, defining distance thresholds, or analyzing influencing factors. However, these findings often fail to guide comprehensive development and planning design at the neighborhood level for the station catchment area. Therefore, this study focuses on the core idea of locational land rent theory, using walking accessibility as the central content and examining its influence on the premium effect of ancillary commercial facilities. This reflects the value law of functional facilities in the station core area and quantifies the economic value of walking accessibility at a microscopic scale, which is an innovative work in the field of environmental behavior research. Additionally, it is important to highlight the economic value of individual pedestrian movements exiting the station, which is essential for most developing countries in the Asian region committed to developing public transportation to address population expansion and high built environment density.

(3)

Finally, this study proposes an integrated multi-model approach to comprehensively quantify the economic value of walking accessibility. The traditional hedonic price model is first modified through spatial econometric regression, constructing a regression model with the rent of commercial facilities as the dependent variable and attributes of different dimensions of affiliated commercial facilities as independent variables. A key parameter is introduced to measure walking accessibility, with walking time as the core variable and model fitting coefficients extracted. Subsequently, the cost–benefit approach is applied for scheme evaluation, using the station–city combination facility as a medium element to compare the commercial premiums that could be brought by walking time before construction with the construction cost of the facility. The necessity of the construction of the combined station–city facility and the economic value of the walking accessibility brought by it are comprehensively analyzed through this method, representing an exploratory attempt in this study.

2. Models and Methodology

2.1. Spatial Econometric Regression Models

The traditional HPM (hedonic price method) model is a common tool for evaluating real estate prices which focuses on real estate as an idiosyncratic commodity, with sample differences coming from various characteristics that the real estate itself has that can satisfy the consumer demand to explain the heterogeneity among commodities. The basic form can be represented by a multivariable linear regression model; the coefficients of the characteristic variables can be fitted by OLS least squares regression. Conversely, the spatial econometric regression method is based on the traditional cross-sectional and panel data by introducing spatial parameters (global constant parameters) and performing spatial autocorrelation and structural analysis on them. It primarily includes the spatial autoregressive model (SAR) [39], spatial autoregressive in error term model (SEM) [40], and SDM. The first two models can be generated by degenerating the coefficients of the SDM. Since the target of this study is the sample point data of shops under the same analysis space, the interaction between the independent variables of samples from adjacent areas is not considered for the time being; therefore, the models are primarily chosen from the first two. The difference between the SAR and SEM models is primarily the difference in the coefficients of the introduced variables, which are spatial lag variables and error variables, respectively. The specific selection method is determined by the coefficient test. The following figure shows the general process of constructing two classical spatial econometric regression models. The first step is to perform OLS regression (ordinary least squares regression) and use the OLS residuals for the Lagrange multiplier test. The test results consist of two statistics: LM-lag (Lagrange multiplier lag) and LM-error (Lagrange multiplier error). If neither of them is significant, OLS is the best model to use, suggesting that the data sample does not fit well with spatial econometric regression, and a simple OLS regression model is sufficient to capture the relationship among the variables. The spatial lag model is preferred if LM-lag is larger and more significant than LM-error, while the spatial error model is preferred otherwise. If both statistics are significant, a robustness test is required. The spatial lag model is chosen if Robust LM-lag is significant, and the spatial error model is chosen otherwise. Therefore, the following modeling process will strictly follow this procedure to test the coefficients and choose the optimal model type (Figure 3). Spatial econometric regression models take into account the spatial distribution effects and the potential error influences of the geographic information of the samples. They can effectively capture the relationships among the explanatory variables of the samples after adjusting for this issue and are widely used to study issues such as regional economic growth, housing prices, environmental pollution, and traffic flow planning. Among them, the impact of housing prices is one of the most fundamental applications of these models. This paper adopts the spatial econometric regression method to improve the traditional hedonic price model, considering the strong spatial correlation between the housing samples and the rail lines and stations.

Combined with the previous studies on real estate prices, considering the model fit, normal distribution hypothesis testing, and heteroskedasticity problems, the rental prices are taken as logarithmic, and a semi-logarithmic form of the model is used. The basic form of the SAR and SEM models is as follows:

$$\ln B_i={\alpha }_0+\rho W{B}_i+{\displaystyle \sum _{k=1}^m{\alpha }_k}{X}_{ki}+\beta d_i+{\epsilon }_i$$

(1)

$$\ln B_i={\alpha }_0+{\displaystyle \sum _{k=1}^m{\alpha }_k}{X}_{ki}+\beta d_i+{\epsilon }_i$$

(2)

$${\epsilon }_i=\lambda W{\epsilon }_i+\mu$$

(3)

where ρ is the spatial autocorrelation coefficient; λ is the spatial error variable influence coefficient; W is the spatial weight matrix; B_i is the single monthly rental price of the ith sample of shops (JPY 10,000); m is the total number of shops in the sample (k takes the values 1, 2, 3 ……, m − 1, m), $X_{ki}$ is the value of the kth characteristic variable of the ith real estate sample; $d_i$ is the ith sample of the shops’ distance from the nearest bus stop (m); ${\epsilon }_i$ is the possible random error for the ith sample of shops; μ is the random error from the model fit (the variables fit a normal distribution); and ${\alpha }_0$, $\beta$, and ${\alpha }_k$ are the impact coefficients.

2.2. Cost–Benefit Evaluation Model

The cost–benefit analysis (CBA) method is commonly employed in the evaluation of engineering projects. Its primary function is to establish construction objectives, propose alternative solutions, and employ relevant technical methods to provide detailed assessments of the expected costs and benefits of each solution. The method then prioritizes the alternatives and uses certain principles to determine optimal decisions by comparing them and rejecting solutions where the marginal social costs exceed the marginal benefits. While the CBA approach may have limitations in socio-economic impact assessment due to its tendency to favor “monetary” evaluation, its greatest advantage is its ability to offer decision makers critical ideas and quantitative methods to assess the comprehensive value of utilities in a more scientific manner. This approach is of great importance to the economy of suburban station construction and the rationality of planning and design schemes. In different urban design schemes for joint station–city development, there may be significant variations in the layout and form of ancillary commercial facilities, which could affect the proximity of the combined commercial facilities. When a particular type of combined facility is adopted to connect commercial facilities with station buildings, it is necessary to evaluate whether the additional construction cost can be offset by the increase in value-added premium revenue through a cost–benefit model. Therefore, a cost–benefit model must be employed to ascertain the viability of the investment of public resources in the project.

This study proposes to evaluate the design of selected target station bonding facilities by extracting key variable coefficients through spatial econometric regression analysis. Under reasonable prerequisites, the scenario simulation method is used to compare different urban design solutions and draw up pedestrian walking routes and actual walking distances before and after the construction of the station–city combination facilities to improve the proximity of stores and enhance the premium effect. The cost indicators primarily include the combination cost, that is, the additional cost to achieve a better station–city combination under the premise of its own scale, land cost, and construction costs, such as the construction cost and operation and maintenance cost of combination facilities, such as flyovers and plazas. The model is

$$C_t=C_L+C_F+C_M$$

(4)

where $C_t$ is the present value of the increase in construction cost due to the addition of the facility (JPY); $C_L$ is the additional cost of civil works for the linkage facility (JPY); $C_F$ is the cost of equipment required for the linkage facility, including the purchase of lighting equipment and escalator motors (JPY); and $C_M$ is the operation and maintenance cost, which primarily includes the capital investment required for the normal operation of the facility and the maintenance cost, including the daily cleaning of the facility and equipment maintenance costs, including routine cleaning of facilities and equipment repair and maintenance costs (JPY).

The benefits mainly include the value-added benefits of the ancillary businesses directly connected to the combined facilities and the benefits of the crossing time between the station and the businesses after the construction of the combined facilities. Owing to the uncertainty of the changes in station passenger flow, accurately determining the conversion of production activities is challenging. To avoid the inconvenience of valuing the cost of station bonding facilities, the evaluation model is simplified by considering only the differences in commercial value added due to changes in pedestrian routes before and after the erection of bonding facilities in critical areas, and all benefits are discounted to the present value at the beginning of the year of operation for analysis and comparison.

2.3. Model Integration

The core objective of this study is to evaluate the effect of station–city integration facilities on pedestrian accessibility improvement and their impact on the premium effect of affiliated commercial facilities, but there is no unified calculation model to address this issue. The spatial econometric regression model plays a critical role in correcting geographic information errors in the sample, resulting in a more reasonable form and accurate coefficients for the generated spatial lag model. This model is typically employed to analyze data that exhibit spatial patterns and may be influenced by spatial factors. In contrast, cost–benefit analysis is a method that compares the expected benefits of a policy or program with its expected costs to determine its viability. When used in conjunction, spatial econometric models and cost–benefit analysis can provide valuable insights into the economic viability of policies or projects that exhibit spatial dependence or interdependence. A recent study employed a spatial econometric model to extract variable coefficients associated with walking paths and calculate the resulting premium benefits. Subsequently, a cost–benefit analysis was conducted to evaluate the overall costs and benefits of the project while accounting for economic factors within the spatial pattern [41,42,43]. Therefore, a comprehensive analytical framework is presented in this study by integrating two commonly used models in related literature. First, a traditional spatial econometric model is utilized to construct a model with store rents as the dependent variable and the coefficients of key variables (e.g., walking time) combined with the attribute variables of stations and commercial facilities selected from existing studies as the independent variables. Subsequently, the key coefficients are introduced into the cost–benefit model, and scenarios are used to highlight the effect of the combined station–city facilities on pedestrian accessibility. Based on this, the costs of the facilities and their commercial effects are calculated, and a comprehensive evaluation is conducted to demonstrate the necessity and magnitude of the impact of the combined facilities on pedestrian paths. It should be noted that the process of model integration does not violate the premise of applying a single model with changing the meaning of the coefficients, which is a prerequisite for model integration. The process of model integration is shown in the Figure 4:

3. Study Area and Variables

3.1. Study Area

The suburban railway lines in Tokyo, Japan, were chosen as the research objects primarily due to the following factors: (1) Japan is a very representative country in Asia, where the construction of rail transit drives urban development, and its lines have a strong coupling relationship with the high-density built environment; (2) Japan has established the construction mode of station–city integration earlier and built several world-renowned rail transit complexes, and the joint development mode of the station city is more mature, with sufficient case samples and high ability to research; (3) Japan’s suburban lines are different from the traditional suburban railways in the sense that the construction around each station in the suburbs has a clear clustering effect, and beyond a certain range, the layout has a low density. The different types of stations in suburban and urban areas of Japan are shown in Figure 5.

Meanwhile, unlike the model of fully integrated station–city complexes, such as Shibuya and Shinjuku Stations in the city center, suburban businesses are relatively independent, which can ensure the spatial independence of the sample and, at the same time, the interference of vertical walking distance and other factors can be avoided in subsequent studies [44]. There are 18 suburban rail lines in Tokyo, of which the most prominent suburban lines include: the Ikegami Line, Oimachi Line, Higashi-Yoko Line, Meguro Line, and Tanen Urban Line of Tokyu Corporation; Keisei Line, Shin-Keisei Line, Keio Line, Inokashira Line, and Sagamihara Line of Keisei Corporation; and Odakyu Line of Odawara Corporation. If all rail lines were included in the study sample, tens of thousands of commercial data would be involved, resulting in an overwhelming amount of spatial information, which would instead be detrimental to targeted research.

The Odawara Line, which runs from Shinjuku Station to Odawara Station in the western part of Kanagawa Prefecture, was first opened in April 1927, and new stations have been built on the line since then. The Odawara Line has direct access to the Hakone Mountain Line and is interconnected with the Tokyo Metro Chiyoda Line and the Tokiwa Line of the East Japan Railway (JR East); it is also a limited express train to the Gotemba Line of the Tokai Railway (JR East). The line has 12 above-ground stations (elevated or at-grade) for both commuting and sightseeing (

Table 1. Statistics of combined facilities at each station on the Odakyu Co. Odawara Line.

Combined Design
Name of Station	Station Plan	Type of Station	Combined Facility Type	Combination Mode	Combined Facilities
Xiangqiu Garden		Through station	Separated type	Line type	Square + Street
Crane sichuan		Through station	Separated type	Line type	Square + Street
sutra hall		Through station	Fit type	Line type	Synthesis
Zushi Valley Treasure		Through station	Fit type	Line type	Street
Yuki Oda, Ami Ohno		Through station	Fit type	Line type	Synthesis + Square
Ebina		Transfer station	Fit type	Line type	Square + Street + Pedestrian bridge
Local thick wood		Through station	Fit type	Line type	Synthesis + Square + Street
A small field		Transfer station	Fit type	Line type	Synthesis + Square + Pedestrian bridge
Chengcheng School front		Through station	Wrap type	Face type	Synthesis
The Hill of New lilies		Transfer station	Fit type	Branch type	Square + Street + Pedestrian bridge
Sagami Ohno		Transfer station	Fit type	Face type + Branch type	Synthesis + Square + Street + Pedestrian bridge
Machida		Transfer station	Fit type + Wrap type	Face type + Branch type	Synthesis + Square + Street + Pedestrian bridge

). Most of the stations on the line have integrated facilities such as pedestrian bridges and underground passages and are highly accessible within 5 to 10 min on foot. Hence, The Odakyu Odawara Line is the best study target.

The group classified the rail transit stations based on the station type, combination facility type, combination mode, combination facility, and function type in the early stage and conducted certain statistical research according to different types. According to the different classifications of combining facilities, considering the frequency of use and the difficulty of construction and other comprehensive factors, the pedestrian bridge has the most obvious advantages of increasing the flow of people with lower construction costs among all kinds of combining methods, so this study chooses the pedestrian bridge as the object for research.

3.2. Feature Variable Selection

When selecting characteristic variables, attention must be paid to the differences between the commercial and traditional residential real estate samples while incorporating relevant research results. Commercial and residential real estate is closely related to people’s daily work and represents more complex commodities. There are significant differences between similar real estate commodities due to the characteristics that constitute their use value or affect their value fluctuation, and the characteristic price method considers that heterogeneous products are influenced by their characteristics. The construction of a characteristic price model for commercial real estate requires the identification of various characteristic factors that affect price fluctuation. The characteristic elements affecting the price can be divided into zoning, lot size, centrality, type of house, type of building, number of rooms, accessibility, land-use planning, natural scenery, wind power generator, location, neighborhood [45,46,47,48]. Since this study focuses on the relationship between sites and shops, the neighborhood characteristics are transformed into site characteristics adjacent to the sample of shops, as shown in

Table 2. Feature variable selection.

Store Structure Features	Store Location Features	Adjacent Station Features
Monthly rent per m2	Distance to Shinjuku Station	Type of station
Construction time		Number of interchanges
Store size	Straight-line distance to the nearest station	Combination method
Number of floors
Number of the store floors	Distance to Machida Station	Combination of facility types
Walking time (to nearest station)

.

The first group of variables describes the structural characteristics of the shops themselves, including the time of construction, size, number of floors, height of available shop floors, and walking time to the nearest site. The proximity of the commercial facility was measured using these variables. The walking time measurement was obtained from the data collection website At Home Data Measurement Source, which uses the shortest walking distance measured in the field at a normal adult walking speed (approximately 1.25 ms⁻¹) based on the street network. This measurement has some explanatory power and scientific validity.

The second group of variables describes the locational characteristics and focuses on the relationship between shops, stations, and the city, mainly considering the distance factor. This group includes the straight-line distance from the station, the distance from Shinjuku station, and the distance from Machida station. The selected line is a suburban non-circular line with a terminal station in Shinjuku, close to the city center, which was chosen to replace the influence of the city center. The distance from Machida station was included in the range of variables considering locational characteristics, as it serves as a hub station for the Odawara Line interchange to the JR Yokohama Line and can be interchanged to Tama New Town, Yokohama, and other critical nodal areas, which is presumed to significantly affect the sample. The straight-line distances were all measured directly using the ArcGIS calculation tool, as shown in Figure 4.

The third group of variables describes the characteristics of the shop-adjacent stations, including the station type, station–town combination method, type of combination facility, and average number of interchanges. The study established binary variables (0, 1) to describe the station type as single-line intermediate or multi-line interchange stations. The combination mode reference [6] was divided into separated (no combination), joined, fused, and integrated (joined + fused), and a scoring system was used to indicate the degree of combination for subsequent linear regression analysis. Based on the average daily passenger flow of the station and the evacuation efficiency of the passenger flow, the analysis shows that the separated type was the worst with a score of zero, the integrated type was the best with a score of three, and the combination facility type was divided into the types of combined facilities classified as linear, branch, surface, and integrated. The same scoring system was used to grade the participants according to convenience.

3.3. Data Analysis

The experimental data were obtained from the at-home property rental information network. Initially, we obtained data from 1989 stores along the Odawara Line, as shown in Figure 6. The data cleaning work primarily includes cleaning invalid data and null values and screening out some data that do not meet the research objectives. As the subsequent cost–benefit model primarily evaluates the improvement effect of the combined station–city facilities, it is necessary to judge the design scale of the facilities reasonably and conduct a two-step screening process. Firstly, stations with underground stations and the remaining 12 above-ground stations (including at-grade stations and elevated stations) were eliminated; secondly, the impact range of rail transit stations on the surrounding land premium is generally between 2 km and 3 km, but the urban construction density around the suburban stations is relatively low [14]. If the scope is too large, the sample size will decrease sharply, which will affect the accuracy of the model calculation. Considering that the comfortable walking distance is generally within 10–15 min of walking range [33,49,50], a range with a radius of 2 km from the site as the center is selected as a sample. Finally, we obtained 809 valid data points that could be used in this study. Figure 7 shows the distribution of the sample points of the original data.

4. Results

4.1. Analysis of the Premium Effect of Shops Based on Spatial Econometric Regression Analysis

The listed sample data of the characteristic variables were all included in the OLS model. The stepwise regression method was used to sequentially exclude the characteristic variables with similar characteristics and a significance of less than 10% and variance inflation factor greater than 3 to correct the multiple covariances of the model. According to the stepwise method, the variables were screened out based on the fit of the generated model. The correlation and significance of the shop area and rent are low, so they are removed. Since the station type and average interchange traffic have similar characteristics (interchange stations generally have higher traffic than intermediate stations), the variables of the station type and number of floors of buildings with low significance are removed. After removing the variables, the model fit was relatively good, with an adjusted R2 of 0.514 and no significant variation reduction. It was used to construct a spatial panel econometric model.

The construction of the spatial weight matrix is a crucial step before conducting the spatial correlation test. Generally, two methods, namely the neighborhood method based on contiguity and the distance threshold method, are employed. However, the choice of the method depends on the type of sample, and the optimal spatial weight matrix needs to be selected by comparing the correlation test metrics, such as the model fit R2, Moran’s I index, and significance level. The neighborhood method based on contiguity is mainly used for geometric shape elements, presupposing the presence of common edges or points between them. On the other hand, the distance threshold method can be used for the spatial correlation analysis of geometric shape and point samples by taking the geometric center of gravity of the samples and searching for sample points around the target point by the distance between them. In this study, the distance threshold method is applied to construct the spatial weight matrix, and the related indicators are compared to select the most suitable distance value. Distance threshold values of 650, 800, 1000, 1200, 1500, and 2000 m are chosen, and the lowest value of 650 is used to calculate the spatial distance weight matrix generated by the calculation software based on the spatial distance statistics of the sample taken. This ensures that each sample is surrounded by at least one neighboring sample. The distance values were then taken at an interval of about 200 m, which is a combination of suggested values given in the literature and is not absolute. The reason for choosing this value is that the selected samples are located around the suburban routes, and a review of the local road network through Google Earth reveals that 200 m roughly matches its neighborhood design scale and ensures that at least one sample will be distributed at every 200 m interval.

The

Table 3. Sample spatial OLS regression correlation test based on different spatial weight matrices (source: self-drawn).

	Distance Metric
	650 m	800 m	1000 m	1200 m	1500 m	2000 m
R2	0.524	0.530	0.545	0.529	0.442	0.410
Moran’s I	0.337 **	0.386 ***	0.501 ***	0.442 ***	0.470 ***	0.305 ***
Lagrange Multiplier (lag)	0.340 *	2.373 **	21.754 ***	15.906 **	12.098 **	6.731 ***
Robust LM (lag)	4.604 **	8.736 **	9.672 ***	15.676 **	0.017	7.449 *
Lagrange Multiplier (error)	0.459 *	1.459	4.042 *	2.928	2.839 *	0.671 ***
Robust LM (error)	5.851 *	7.821 ***	3.811 *	7.193 *	1.133 *	1.390 ***
Lagrange Multiplier (SARMA)	7.533 **	10.195 ***	19.718 ***	2.041 **	2.857 **	8.121 ***

Note: *, **, and *** denote p ≤ 0.1, p ≤ 0.05, and p ≤ 0.01, respectively, and represent significance at the 10%, 5%, and 1% levels, respectively.

displays the basic values of the spatial weight matrix generated using the aforementioned distance threshold values. Based on the Moran’s I index, it can be observed that the distance weight matrix with a threshold of 1000 m exhibits the highest spatial correlation performance. Therefore, this distance value is selected to generate the spatial weight matrix applicable to the construction model of this study. Additionally, Lagrange multipliers (LMs) and robustness tests are compared to determine the appropriate econometric model. For the spatial weight matrix generated by 1000 m, both LM-lag and LM-error are found to be significant, with the LM-lag value being greater than the LM-error value. As a result, further robustness checks are conducted, and the Robust LM-lag value is significantly higher than the Robust LM-error value. Based on the significance and model confidence, the SAR model is determined to be more suitable.

Figure 8 shows the univariate global spatial autocorrelation analysis that was conducted after taking the logarithm of the rent variable of the sample of shops. The value of Moran’s I (spatial autocorrelation coefficient) was 0.510, showing a moderate degree of spatial autocorrelation, and the significance was tested at 0.001. The lagged ln monthly rent is the spatially lagged (i.e., the weighted average of an observation and its surrounding “neighbors”) vector of the variable ln monthly rent.

Based on the SAR model, using a semi-logarithmic form of transformation (the reasons for adopting a semi-logarithmic model are explained in the previous sections), the impact of characteristic elements on premium benefits is converted into a percentage form to quantify the meaning of the coefficients as shown in

Table 4. SAR model analysis result.

Variables		Coefficient	Std. Error	Probability
W-ln (Monthly Rent)		0.4608940	0.0472351	0.00000
Constant		1.4955800	0.1900120	0.00000
Store features	Construction time	−0.0068320 ***	0.0007980	0.00501
	Number of floors where the store is located	−0.0667130 ***	0.0062810	0.00000
	Walking time	−0.0301790 ***	0.0061170	0.00000
Store location features	Distance to Shinjuku Station	−5.13814000 × 10−6	1.5238000 × 10−6	0.13363
	Straight-line distance to the nearest station	−0.0001070 *	0.0001580	0.08957
	Distance to Machida Station	2.9867000 × 10−6 **	1.550620 × 10−6	0.03514
Adjacent station features	Number of interchanges	1.1376100 × 10−7 ***	3.2749400 × 10−7	0.00045
	Combination method	0.0345834 ***	0.0347450	0.00000
	Combination of facility types	0.0486020 **	0.2421390	0.04063
Circle to which the store belongs (measured in straight-line distance)/m	(0–100]	0.0707650 ***	0.0962690	0.00154
	(100–200]	0.1339430 ***	0.0716340	0.00061
	(200–300]	0.0672810 ***	0.0716340	0.00422
	(30–400]	0.1274370 ***	0.0600130	0.00354
	(400–500]	0.0638960 ***	0.0541620	0.00434

Note: *, **, and *** denote p ≤ 0.1, p ≤ 0.05, and p ≤ 0.01, respectively, and represent significance at the 10%, 5%, and 1% levels, respectively.

: (1) First, consistent with traditional studies, location factors have a certain degree of influence on the prices of shops along the Odawara line (which has long been determined by the laws of urban economic geography), but the coefficient is low in this study. Among them, the coefficient of linear distance from the nearest station is negatively significant, indicating that the closer the station, the higher the premium index, with prices increasing by approximately 0.01% for every 100 m closer. (2) Second, the physical characteristics of shops have an impact on rents, and the regression coefficients are high. For example, the number of store floors and the construction time are negatively correlated, which can be explained by the fact that the older the store and the higher the number of store floors, the lower the rent of the store (here, we need to exclude the few cases where the rent of the store with certain historical district characteristics may have a positive correlation with the construction age). (3) Finally, the regression coefficients of the station–city combination mode and combined facility type are positive, indicating that they have a positive premium effect on shop rents, which confirms the theme and core content of this study. The design of the station–city combination mode and the combined facility significantly affect the premium effect of shops in joint development; the regression coefficient of the station interchange number is small, indicating that its influence on the premium effect of the neighboring commercial is low. Combined with the market law of shop operation, the rent of shops will follow the guidance of market policies and the fluctuation of pedestrian flow for dynamic adjustment. This study did not continuously track the dynamic changes in shop rent in the calculation process and analyze its relationship with pedestrian flow in depth. However, this will be addressed in future work. In this study, average daily interchange data were chosen without considering the instantaneous effect of important holidays on passenger flow. Consequently, the value of the regression coefficient is not high; nonetheless, the significance still passes the test, indicating that the coefficient is statistically significant. The study observed that when comparing the coefficient fits of the two characteristic variables of walking arrival time and straight-line distance simultaneously, the regression coefficient of the premium for walking arrival time is approximately −0.03. This is a high value based on the meaning of the variable and cross-sectional comparisons, indicating that the shorter the actual walking arrival time at the station, the more significant the increase in the premium benefit of shop rent. Meanwhile, the magnitude and significance of the former value are better than the latter, which reflects a common problem in related research. That is, the Euclidean distance is much less effective in small-scale studies than in large-scale studies. In neighborhood-scale walking situation studies, the true distance is the actual walking distance based on the road network, which needs to consider the possible existence of detours that lead to straight-line distance. The problem of not being allowed is the reason for the subsequent choice of this study to estimate by fitting values to the coefficients of the walking time variable, as shown in Figure 9.

The study narrowed its scope by analyzing the overall premium effect and combined it with existing studies to investigate the fluctuation of the circle to which the sample belonged. Binary variables (zero and one) were used to indicate whether the sample of shops was within a circle of a certain distance from the site. The sample was subdivided by a straight-line distance of 100 m from the site. The study found that the 0–100 m circle is not the most significant, and the conclusion that “the closer to the station, the higher the price” may not be applicable when the study scale is reduced. The regression coefficient for the 0–100 m circle is 0.0707, indicating a 7.07% increase in shop rent due to the proximity of the rail station. For the 100–200 m circle, the regression coefficient is 0.1339, indicating a 13.39% increase in shop rent due to the proximity of the station. The regression coefficient for the 300–400 m circle is 0.1274, but excluding the single reason of distance, this result may be influenced by the unevenness of the sample distribution, inhomogeneity, differences in road network structure, neighboring facilities, etc. The regression coefficient decreased after 400 m. It can be roughly determined that the positive premium impact of each station on shops on the Odawara Line is approximately 200 m. Based on the results, it is further speculated that 200 m may also be the ideal control range to guide the design of a station–city combination with the station catchment area and the distance boundary that can maximize the station premium effect, as shown in Table 4.

4.2. Cost–Benefit Estimates for the Station–City Combination Facility

In the previous section, a spatial lag model was used to analyze the premium effect of shops located around each station on the Odawara Line using spatial regression. The study classified the sample into circles and determined the “ideal premium control construction area” to be approximately 200 m. Based on the model coefficients, various scenarios were proposed for the design of the combined station–city facilities, and simulations were conducted to analyze the effect of the combined station–city facilities on the premium by comparing before and after.

To simplify the problem analysis process and control the analysis variables, Hirao Station, which is also located on the Odawara Line, was selected as the object of the cost–benefit evaluation. Ebina Station was chosen as a benchmark because it is an above-ground station with a cross-line type station building, equipped with pedestrian bridges that connect it to surrounding buildings, commercial complexes, and office buildings around the station [51]. The southern footbridge is a general pedestrian function that connects to the station’s second-floor exit and passes through several office buildings, leading to the shopping center “SHOPPERS PLAZA Ebina,” and finally to the Ebina three-dimensional parking building, which is in line with the objective of the analysis regarding the relationship between the building and the station [52]. Therefore, the “SHOPPERS PLAZA Ebina” was selected as the object of evaluation due to its four floors and a construction area of 22,244 m², with the commercial area being 14,184 m², which belongs to the traditional commercial complex consisting of the main shops for rent. The study examined the improvement effect of the facility on path selection, as shown in Figure 10.

According to the data related to other similar projects made public by the Tokyo construction authority, the civil construction cost of the station combined with the flyover facility is approximately JPY 500 thousand per m², and the total length of the flyover is approximately 243 m, of which the flat section is approximately 230 m, the projected length of the outdoor staircase is approximately 12 m, the average section width is 4 m, and the total construction area is approximately 900 m². The civil construction cost C_L = JPY 450 million, and according to relevant data, the average annual operation and maintenance cost of a new flyover in Japan over the past 10 years has been about JPY 6 thousand per m²; that is, the average annual operation and maintenance cost is about C_M = JPY 5.4 million. Moreover, since the flyover does not involve the costs of outdoor elevators and lighting equipment, the total cost is dominated by civil construction and operation and maintenance costs. That is, the total cost of C = C_L + C_M is JPY 455.4 million.

Based on the results of the SAR model analysis, various factors were controlled, including the characteristics of the shop, location, and neighborhood sites, as well as the role of spatial autocorrelation effects. The main focus was on the correlation between the improvement of proximity and the added premium value resulting from the combination of facilities. The value of time theory was applied to transform the problem into a consideration of the reduction in walking time and changes in the walking path, which can enhance walking accessibility, expand the effect of “passenger flow” to “commercial consumption flow,” and influence the commercial value of shops. By extracting the regression coefficients of walking time variables, spatial autocorrelation coefficients, spatial weight matrix, and constant terms (error terms, etc.), based on the SAR model, the estimation model of the value addition of shop premiums based on walking time is constructed as:

$$\ln B_i={\alpha }_0+\rho W\lg B_i+{\displaystyle \sum _{k=1}^m{\alpha }_k}{X}_{ki}+\beta d_i+{\epsilon }_i$$

(5)

where X is based on the road network walking time to the target shop (min). The rest of the parameters are explained in Equation (1). The formula is as follows:

$$X=T_W=T_V+T_h$$

(6)

where $T_V$ is the time taken to go up and down the outdoor steps (escalators) from the exit to the destination sample (min), and $T_h$ is the walking time from the exit to the target shop sample via horizontal facilities (streets, pedestrian bridge, etc.) (min).

This study uses walking time as the primary measure of station–commercial business proximity and integrates the coefficients of the walking time variable into the SAR model to derive a model of the value-added benefits of the single-month premium for shops based on the effect of walking time as follows:

$$\ln B_i=0.472\times 0.510B_i-0.006T_w+0.190012$$

(7)

The benefits created by walking time savings before and after the flyover are erected owing to the change in route choice. The route choices are shown in Figure 7. If the flyover is built, pedestrians exit the station, go directly from the second floor to the flyover, and reach the second floor of the target shop; the walking time through the flyover at normal speed is approximately $T_W=T_h=$2.95 min. Substituting $T_W$ into Equation (7), we can see that the flyover brings about a single month of premium revenue for a single shop after the flyover is built. $B_1$ = JPY 11.901 thousand; if the flyover is not built, i.e., pedestrians need to go down from the second floor to the first floor to exit the station then walk to the intersection according to the nearest principle, the total walking time is approximately 7 s. The mean total walking time is $T_W=T_V+T_h+T_d=$ 7.26 min (where $T_V$ is the walking time from the station to the first floor, $T_h$ is the horizontal walking time, and $T_d$ is the traffic light delay time under ideal conditions), which results in a single monthly premium of approximately $B_2$ = JPY 11.594 thousand per shop. Secondly, in the study case, the shopping center “SHOPPERS PLAZA Ebina” covers an area of 14,184 square meters. If only the store area is considered, a rough calculation shows that about 200 stores of 70 square meters can be set up and rented out to the public, and after this, the total annual average premium increase will reach approximately JPY 28.5 million. The new walking trails resulting from the connection facility improvements could generate more additional revenue compared to at-grade trails and could equalize the construction costs in about two years. As a type of transportation infrastructure led and invested in by the government, the pedestrian bridges that are constructed with durable materials such as steel, concrete, or composite materials can have a service life of 50 years or more [53,54]. The main cost investment during this period is the operation and maintenance cost. For a limited period of use, if the pedestrian bridge is solely used to guide people across the street, it will only generate social value and needs to be evaluated further by calculating the gross social product to assess the economic benefits it generates. However, this approach is not intuitive, and it is generally difficult to recover construction costs in the short term, resulting in this type of project usually being in a loss-making situation. On the other hand, if the pedestrian bridge is linked to ancillary commercial development, it can generate direct economic benefits and quickly recoup the construction costs.

After completing the benefit prediction, assessment, and comparative analysis, the group utilized another case to further substantiate the rationality of their conclusion and the necessity of the station–city combination facility design. The BRT station at the Municipal Service Center in Huli District, Xiamen, is connected to the Tianhong Mall and Wuyuanwan Leduhui Shopping Center through a joint air bike path and pedestrian bridge. However, this connection was not initially planned, and the construction history of the station–city integrated pedestrian facilities connecting these three locations is shown in the following figure. The construction plan was initiated in early 2016, and the first phase of the plan only considered connecting the BRT station to the Le Du Hui shopping center. In September 2016, the full-scale construction of the Xiamen City Air Bikeway began, extending the outbound pedestrian flow from the BRT station further to the adjacent block of Tianhong Mall. However, it did not connect directly to the mall; instead, it was focused on connecting the Air Bikeway to the surface streets. During the third phase of construction, which began in 2020, the aerial bicycle path was connected to the third floor of Rainbow Mall, and pedestrian-friendly facilities such as shade canopies and escalators were added around the aerial bicycle path. After the completion of the entire facility, passengers exiting through the BRT station could be effectively converted into commercial consumption streams, contributing significantly to the commercial premium benefits of Rainbow Mall and Wuyuanwan Loduhui Shopping Center in one year (Figure 11).

5. Discussion

From the main work and core findings of this study, this paper analyzes the law of the premium effect of commercial facilities attached to the station area, taking suburban railroads as an example, and on this basis, it explores the role of the improvement of pedestrian accessibility that station–city combination facilities can provide and the economic value they can generate, further refining the research dimension of the impact of rail transit on the premium effect of surrounding real estate. First, based on the results of the analysis, it is initially determined that the impact of proximity to suburban railroad stations on the premium effect of stores within the station area does not increase linearly, but fluctuates. This is contrary to the traditional perception that “the more proximity to a station, the better the value-added effect (or the higher the store rent)”. The main reason for this is that, from the perspective of the “distance” factor, unlike stations in urban centers, suburban stations have a relatively low-density road network; the difference between straight-line distance and real walkable distance is obvious, so the former is less sensitive to the premium effect, and the influence of walkability on commercial facilities relying on pedestrian streets or combined facilities is magnified. At the same time, due to the different levels of construction of suburban stations, the rent premiums for stores within the station area are not stable from an overall perspective and may vary as shown in the analysis results—the closer the distance, the greater the variation—but will continue to grow and reach a peak within a certain distance and then stabilize. After reaching the peak, the influence of distance is weakened and other factors (e.g., the level of public services, etc.) start to show a downward trend; secondly, comparing the results of the cost–benefit analysis, it is found that only one pedestrian bridge can recover the construction cost about two years earlier, which further demonstrates the importance of pedestrian accessibility and the fact that the three-dimensional station–city combination facility not only has the role of dispersing pedestrian flow but also generates potential additional economic value.

From the main work and core findings of this study, this paper aims to analyze the premium effect of commercial facilities attached to the station area, taking suburban railroads as an example. Based on this analysis, the role of pedestrian accessibility in generating economic value through station–city combination facilities is explored, further refining the research dimension of the impact of rail transit on the premium effect of surrounding real estate. The results of the analysis suggest that the impact of proximity to suburban railroad stations on the premium effect of stores within the station area does not increase linearly, but fluctuates. This finding challenges the traditional perception that “the more proximity to a station, the better the value-added effect (or the higher the store rent)”. The lower-density road network of suburban stations results in a difference between straight-line distance and real walkable distance, making the latter more sensitive to the premium effect. The influence of walkability on commercial facilities relying on pedestrian streets or combined facilities is magnified. Additionally, due to the varying levels of construction of suburban stations, rent premiums for stores within the station area are not stable overall and may vary as shown in the analysis results. The rent premiums may continue to grow and reach a peak within a certain distance and then stabilize. After reaching the peak, the influence of distance is weakened, and other factors such as the level of public services may start to show a downward trend. Comparing the results of the cost–benefit analysis, it is found that only one pedestrian bridge can recover the construction cost approximately two years earlier, demonstrating the importance of pedestrian accessibility. The three-dimensional station–city combination facility not only disperses pedestrian flow but also generates potential additional economic value. This concept is expressed in this paper in terms of store rents of commercial facilities; however, the rents of commercial office buildings and other buildings connected to the pedestrian facilities will be affected by them, which will further increase and expand the premium benefits brought by the combined facilities and accelerate the cost recovery, as in the case of Ebina Station. Simultaneously, the pedestrian bridge not only significantly enhances the proximity of the station to the surrounding commercial facilities (especially cross-street commercial) but also allows the station to be closely integrated with the surrounding office buildings and stations, effectively releasing more effective construction space. Furthermore, the flyover is also an urban crossing facility, which helps to improve the three-dimensional pedestrian system and realize the separation of pedestrian and vehicle traffic. The erection of flyovers avoids other difficult-to-estimate benefits such as traffic congestion caused by surface roads. Therefore, the positive value generated by the design idea of station–city integration facilities far exceeds the construction cost. This analysis idea and the conclusions drawn will be useful to assist in the comprehensive benefit assessment of different spatial connection scheme designs during the design process of station–city integration of rail transit stations.

In terms of theoretical contribution, previous studies have primarily focused on the premium effect of rail transit stations on surrounding land, typically analyzing and comparing changes in housing prices around stations before and after the completion of rail transit from an economic perspective. The results of these studies have established the foundation for this paper by identifying the type of function, scope of influence domain, and degree of influence. However, it is important to note that previous research on the station domain premium effect has been at a static level. This paper aims to refine the research scale by narrowing the premium effect down to a 500 m core circle of the station, focusing on the non-linear effect of the station on commercial facilities attached to development. The optimal influence area is roughly determined to be within a straight-line distance of 100–200 m from the station. Additionally, this study shifts the research dimension by analyzing the economic value generated by walking accessibility and simulating the possibility of walking activities and path selection of discrete individuals after exiting the station. By translating the walking time saved by pedestrians into the creation of social gross product, this paper quantifies the cost of different route options, as in line with the environmental behavior theory that “pedestrians choose the best route by weighing the time or physical cost of different route options.” By combining these two aspects, this study goes beyond theoretical macro guidance to optimize the layout of spatially connected facilities, obtain higher economic benefits, and improve the convenience of walking activities with less investment.

Besides the contribution to the theoretical research, as discussed in the previous case study of the station–city integration design of the BRT and two commercial complexes in Huli District, Xiamen City, the method of measuring the economic impact of walkability on the joint property development of the rail transit station area explored in this paper has a high practical value. It can greatly optimize the station–city integration design scheme and generate greater comprehensive value. Specifically, it is mainly reflected in two aspects: (1) The cost–benefit model itself is a scientific method to evaluate the feasibility and return on investment of the project. This paper systematically evaluates the composite value of an important station–city integration facility—the pedestrian bridge—through it. For most cities in China, urban transportation infrastructure investment is a huge expenditure. How to ensure the maximization of the benefits of this investment is a concern for both the government and the planning and design personnel. The pedestrian bridge is a relatively common, low-cost, and flexible transportation infrastructure facility, which will not only be used in the connection design scheme between the elevated rail transit station and the surrounding buildings, but also be set up in large numbers on the streets to divert the ground flow and avoid congestion. Therefore, we can also use this evaluation idea to analyze the construction projects with similar nature, which will help to further control the investment cost and predict the returns. Generally speaking, this kind of public infrastructure is invested in and constructed by the government, so its cost and benefit are directly related to the government’s economic policy and urban operation, and for the planning and design personnel, how to better control the cost is a long-standing challenge. Therefore, the research method and the conclusion drawn in this paper have a high practical significance for multiple subjects, as it can effectively promote the development of station–city integration and avoid the waste of economic resources. (2) In the process of comparative analysis, this paper proposes two walking path schemes to reflect the difference of walkability and finally finds the impact of walkability improvement on the value of joint development property. What cannot be ignored is that walkability itself is also the most basic demand of pedestrians in the rail transit station area, and its satisfaction ensures the satisfaction of higher-level needs (such as safety, comfort, pleasure, etc.) and helps pedestrians to carry out urban activities more effectively. This is highly consistent with the people-oriented planning and design concept, which is another important social impact value of this paper.

However, it is important to note that this study has several limitations and requires further empirical evidence. Firstly, the study focuses on suburban stations and aims to control for factors other than pedestrian linkages, but more research is needed to determine if similar premium effect patterns and improvement effects of station–city linkages remain significant in high-density urban centers. Secondly, this study only considers air corridors as integrated station–city facilities, and it is worth investigating if other types of pedestrian facilities, such as underground passages and surface paths, have a similar level of improvement effect. It is also important to consider if the presence of three types of pedestrian connections near a station will result in a non-linear superimposed improvement effect or if they will compete with each other, leading to redundancy and underutilization of space. To address these limitations, the research team plans to conduct further field research and analyze data from stations located in different parts of the city using a morphological approach to classify the types of station–city integration facilities. The team will also evaluate each integration facility and design scheme using the same workflow as this study. Finally, the team will refine their conclusions using cross-sectional comparisons to improve their generalizability and scientific validity.

6. Conclusions

Station–city joint development is a core strategy to promote station–city integration in rail transit construction in major cities worldwide. Under the trend of diversification of station catchment area functions, a reasonable and effective design of station–city integration can help coordinate the pedestrian relationship between stations and cities and strengthen the role of the law of value and the positive spillover effect of stations. This study constructs a spatial panel econometric model based on the Odawara Line of the Tokyo suburban railway in Japan and evaluates the effect of improved proximity on the spillover effect through a cost–benefit model based on the changing patterns and influencing factors of commercial facilities in the station catchment area from an urban design perspective. The main findings include the following: (1) The premium effect of shops around the Odawara Line station of the Tokyo suburban railway shows a strong spatial correlation. The combination method and type of combined facilities, as well as the actual walking time of the stores from the station, have a significant impact on the premium effect of shops. Macro location factors and some of the attributes of the stores (construction time, number of floors where the store is located) have a weaker effect. (2) The change in the premium benefit of shops within 500 m of the site fluctuates non-linearly. Overall, the peak could be reached within approximately 200 m. According to the sample routes selected in this study, when the combination of facilities is designed to be around 200 m, it is the most cost-effective and can effectively eliminate the negative impact of distance on price and turn it into a positive utility. (3) A comprehensive cost–benefit assessment based on scenario-based methods demonstrated a significant difference in the increase in the premium benefits of pedestrian bridge combination facilities before and after construction. Under the “tidal phenomenon” of suburban passenger flow, effective design of the combined facilities can spread out the construction cost and accelerate the goal of “return of premium to the public.” This is of great importance to the economy of suburban station construction and the soundness of the planning and design scheme. (4) The construction of pedestrian bridges can impact the path selection process of pedestrians, helping them to choose their routes out of the station. The combination of profit-making facilities around the station can change the walking trajectory and attract traffic to expand the consumption area, further demonstrating that the improvement of pedestrian accessibility can have both social and economic benefits.

Several issues need to be clarified in this research. Firstly, the case study only includes suburban rail transit stations in Japan. One reason for this is that the population density in Europe and the United States is relatively low, making the construction of large-scale centralized transportation infrastructure incompatible with the development plan of most cities. Additionally, the efficient transportation of people and the commercial premium effect attached to the stations are not apparent in these regions. Secondly, the contribution of rail transit to the regional economy is significantly better than that of private cars [55]. Japan’s rail transit system has undergone extensive development and construction, accumulating valuable experience. This experience is reflected in macro aspects, such as road network planning and integrated station–city development, and in constructing a three-dimensional pedestrian network of station–city connection facilities, which improves the quality of pedestrian space and the activities in its region [56] and promotes station catchment area economic activities. For densely populated developing countries such as China, which will continue to face the challenge of increasing urban population in the coming decades, rail transit, as the “artery” of the city, will also be at risk of a surge in passenger traffic. The control of excessive pedestrian traffic is an essential requirement to ensure pedestrian safety, whereas the more efficient use of pedestrian traffic is the key to improving the economic efficiency of the station catchment area. Pedestrians should be prioritized in a high-density urban environment to promote a consumer-oriented economy. Planning and designing a three-dimensional pedestrian network can improve the quality of the station catchment area in terms of safety and economy. The introduction of pedestrian flow indirectly affects the rental changes in commercial facilities, primarily because the fluctuation of continuous interchange traffic determines the vitality of the station catchment area and indirectly affects the revenue of shops. Whether “passenger flow” can be converted into commercial “consumption flow” is closely related to the pedestrian combination at stations and commercial facilities. This study provides theoretical validation for the next possible station catchment area transformation.