Research on Spatial Differentiation of Housing Prices Along the Rail Transit Lines in Qingdao City Based on Multi-Scale Geographically Weighted Regression (MGWR) Analysis

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This paper takes the spatial differentiation effect of Qingdao Metro on housing prices along its routes as the research object. By using the multi-scale geographically weighted regression model and integrating GIS and POI data, it analyzes the multi-dimensional influence of the metro on housing prices. The topic is of practical significance, but there are deficiencies in variable definition, theoretical depth and policy recommendations. It is necessary to focus on strengthening the rigor of analytical logic and the feasibility of policy recommendations.

1. In the literature review section, the critical analysis of existing studies is insufficient. It is suggested that during the review process, not only should the main findings of the existing studies be introduced, but also their shortcomings and the improvement points of this study should be pointed out to highlight the innovation and contribution of this study.
2. The literature coverage is uneven. The research mainly focuses on China and Europe and America, with insufficient citations of similar studies in developing countries such as Southeast Asia and South America, which weakens the universality of the conclusion. Although many relevant recent literatures are cited, the citations of classic literature are insufficient, and there is a lack of systematic review of early research.
3. Terminology and language expression issues: Some sentences are overly long, such as the first paragraph of the abstract, and should be simplified to enhance readability. The terms "bandwidth value" and "TOD model" should be clearly defined when they first appear.
4. Apart from the MGWR model, there are other spatial statistical models that can be used to analyze the spatial heterogeneity of house prices. It is suggested to compare these models to prove that the MGWR model is the optimal choice. The assumptions and application scope of the MGWR model need to be further discussed. For instance, the MGWR model assumes that the influence of variables is local and varies with space. However, in some cases, certain variables may have global effects. It is recommended to discuss whether these assumptions are fully applicable to the actual situation in Qingdao.
5. Data analysis issues:

1) The research is based solely on data from 2024 and lacks time series analysis. Housing prices are influenced by various factors such as economic cycles and policy changes. Can data at a single point in time fully reflect the long-term impact of rail transit on housing prices?

2) The data preprocessing is opaque. The criteria for excluding "villa and basement" are not clear, which may introduce sample bias. It is not explained how to handle missing values or outliers in POI data.

3) Categorical variables (such as "high-end decoration", "low/middle/high floors") lack operational definitions. It is suggested to clearly define the reference standard for "high-end decoration".

4) The residual spatial autocorrelation of the MGWR model has not been tested. It is recommended to supplement residual spatial autocorrelation analysis to ensure the validity of the model.

6. The analysis section elaborates on the research results in detail, but the discussion is relatively brief and lacks more in-depth exploration and interpretation of the research results. Regarding the spatial heterogeneity phenomenon of the impact of urban rail transit on housing prices, further analysis can be conducted on the underlying reasons and mechanisms in terms of social economy, history and culture, urban planning, etc., as well as the possible impacts and significance of this phenomenon on urban development and residents' quality of life.
7. In the discussion section, the results of this study can be further compared and analyzed with other similar studies to explore the innovation points, advantages and disadvantages of this study, as well as the similarities and differences with other research results and possible reasons, so as to better highlight the value and contribution of this study.
8. The research scope of this paper mainly focuses on Qingdao City, lacking an expansion analysis for other cities. It is suggested to discuss the extensibility of the research results at the end and analyze their implications for other cities.
9. The format of the reference list is not standardized enough, and there are errors in the citation formats of some references. For instance, the abbreviations of journal names are not standardized, and some references are missing page numbers.

10.Some additional comments on the tables and figures:

1) The color gradient of the house price interpolation map in Figure (1) is excessive, failing to clearly display the spatial correspondence between the subway lines and the distribution of house prices. It is suggested to simplify it to highlight the key areas.

2) The formulas are not standardized. For instance, such as Moran’s I and Kriging interpolation formula may have problems like the undefined definitions of variables when they first appear in the formulas, and the mixed use of capital and lowercase letters for weight symbols. It is recommended to reformat them using a professional formula editor.

3) The combination of charts and text content is not tight. The clarity of the charts, the accuracy of the annotations, and the correspondence with the text content all need to be carefully checked and adjusted to ensure that the charts can effectively assist in explaining the research questions and results.

Author Response

Comments 1: In the literature review section, the critical analysis of existing studies is insufficient. It is suggested that during the review process, not only should the main findings of the existing studies be introduced, but also their shortcomings and the improvement points of this study should be pointed out to highlight the innovation and contribution of this study.

Response 1: Based on the reviewers' suggestions, supplement and summarize the literature review section. On the basis of critical thinking, point out the shortcomings of previous studies and emphasize the innovation of this paper. The specific location of this modification can be found on page 2, lines 139-151.

Comments 2: The literature coverage is uneven. The research mainly focuses on China and Europe and America, with insufficient citations of similar studies in developing countries such as Southeast Asia and South America, which weakens the universality of the conclusion. Although many relevant recent literatures are cited, the citations of classic literature are insufficient, and there is a lack of systematic review of early research.

Response 2: Thank you very much for the reviewers' comments. This paper further supplements the literature on the developing countries in Southeast Asia and South America with relatively developed urban rail transit systems: Thailand, Malaysia, Brazil, and Singapore. The specific location of this modification can be found on page 2 and 3, lines 87-114 and 123-127. Furthermore, classic literature is incorporated to enhance the feasibility of the research topic. The specific location of this modification can be found on page 2, lines 59-63 and 76.

Comments 3: Terminology and language expression issues: Some sentences are overly long, such as the first paragraph of the abstract, and should be simplified to enhance readability. The terms "bandwidth value" and "TOD model" should be clearly defined when they first appear.

Response 3: Based on the suggestions from the reviewers, this article has been revised and polished by friends who are native English speakers to enhance the quality of the literature. Terms such as "bandwidth value" and "TOD model" that first appear have been clearly defined to facilitate readers' comprehension. The specific location of this modification can be found on page 3, lines 119-122 and 459.

Comments 4: Apart from the MGWR model, there are other spatial statistical models that can be used to analyze the spatial heterogeneity of house prices. It is suggested to compare these models to prove that the MGWR model is the optimal choice. The assumptions and application scope of the MGWR model need to be further discussed. For instance, the MGWR model assumes that the influence of variables is local and varies with space. However, in some cases, certain variables may have global effects. It is recommended to discuss whether these assumptions are fully applicable to the actual situation in Qingdao.

Response 4: We concur with the reviewers' opinions. Through the comparison of three models, namely OLS, GWR and MGWR, in this paper, it has been found that MGWR has a better fitting degree. Based on the statistical results of the MGWR model, the phenomena presented in Qingdao and the actual situation have been further analyzed. The specific location of this modification can be found on page 8 and 9, lines 314-351.

Comments 5: The research is based solely on data from 2024 and lacks time series analysis. Housing prices are influenced by various factors such as economic cycles and policy changes. Can data at a single point in time fully reflect the long-term impact of rail transit on housing prices?

Response 5: This study utilized the cross-sectional data from 2024. It is true that there are limitations in this study that prevent capturing the long-term dynamic impact of rail transit on housing prices. However, this year was a crucial juncture in the development of rail transit in Qingdao. The number of operating lines of Qingdao's urban rail transit increased to 8, connecting the main transportation hubs, business districts, hospitals, and scenic spots in seven districts and one municipality. This significantly enhanced the coverage and service capacity of rail transit. This paper reveals the spatial differentiation patterns of the impact of rail transit on housing prices. By comparing the changes in housing prices in different regions, it can reveal the influence of rail transit on the equilibrium of regional housing prices, assess the impact of rail transit on the supply and demand relationship in the real estate market, and provide a reference for future housing price prediction. The MGWR model can effectively separate the independent effects of influencing factors on cross-sectional data. We fully agree with the reviewer's viewpoint on the significance of time series analysis. Currently, Qingdao City has not made public the detailed and refined housing price and POI geocoding data covering multiple years. The 2024 dataset is the latest authoritative source integrating POI and housing prices of the third phase of rail transit stations. At present, our team is collaborating with the Qingdao Bureau of Natural Resources to obtain the panel data from 2013 to 2025, and intends to use the Difference-in-Differences (DID) model and event analysis method to separate the dynamic impacts of the planning, construction, and operation stages of rail transit. The related research results will serve as an important extension of this study, and future research achievements will further expand. In the paper, we have pointed out the shortcomings of the research and future prospects. The specific location of this modification can be found on page 17, lines 655-672.

Comments 6: The data preprocessing is opaque. The criteria for excluding "villa and basement" are not clear, which may introduce sample bias. It is not explained how to handle missing values or outliers in POI data.

Response 6: According to the reviewers' comments, explanations and supplements were provided for the processing of POI data, and explanations were also given for why villas and basements were excluded. The specific location of this modification can be found on page 5, lines 184-193.

Comments 7: Categorical variables (such as "high-end decoration", "low/middle/high floors") lack operational definitions. It is suggested to clearly define the reference standard for "high-end decoration".

Response 7: Thanks to the reviewers' comments, this paper has provided detailed explanations for the definitions of the categorical variable "refurbishment" and "low, medium and high floors". The specific location of this modification can be found on page 5, lines 210-220.

Comments 8: The residual spatial autocorrelation of the MGWR model has not been tested. It is recommended to supplement residual spatial autocorrelation analysis to ensure the validity of the model.

Response 8: According to the reviewers' suggestions, the residuals of the MGWR model results were analyzed for spatial autocorrelation using ArcGIS to verify the validity of the model. The specific location of this modification can be found on page 10, lines 371-380.

Comments 9: The analysis section elaborates on the research results in detail, but the discussion is relatively brief and lacks more in-depth exploration and interpretation of the research results. Regarding the spatial heterogeneity phenomenon of the impact of urban rail transit on housing prices, further analysis can be conducted on the underlying reasons and mechanisms in terms of social economy, history and culture, urban planning, etc., as well as the possible impacts and significance of this phenomenon on urban development and residents' quality of life.

Response 9: We are very grateful for the reviewers' suggestions. Based on these, we have conducted a more in-depth analysis of the underlying reasons and mechanisms of the impact of urban rail transit on housing prices from the perspectives of social economy, history and culture, and urban planning, taking into account the actual situation and background of Qingdao. This has enriched our research results. The specific location of this modification can be found on page 14, lines 511-513 and 525-558.

Comments 10: In the discussion section, the results of this study can be further compared and analyzed with other similar studies to explore the innovation points, advantages and disadvantages of this study, as well as the similarities and differences with other research results and possible reasons, so as to better highlight the value and contribution of this study.

Response 10: Thanks to the reviewers' suggestions, in the discussion section, the results of this study are compared and analyzed with those of the research on Bangkok urban rail transit. The innovation points and advantages and disadvantages of this paper are presented, and the persuasiveness of the research results is confirmed. The specific location of this modification can be found on page 14, lines 506-524.

Comments 11: The research scope of this paper mainly focuses on Qingdao City, lacking an expansion analysis for other cities. It is suggested to discuss the extensibility of the research results at the end and analyze their implications for other cities.

Response 11: Based on the reviewers' suggestions, in the discussion section, additional references to the inspirations from other cities similar to Qingdao domestically and internationally were provided, offering reference ideas for the sustainable development of other cities. The specific location of this modification can be found on page 15, lines 559-631.

Comments 12: The format of the reference list is not standardized enough, and there are errors in the citation formats of some references. For instance, the abbreviations of journal names are not standardized, and some references are missing page numbers.

Response 12: Thanks to the reviewers' suggestions, we have re-examined and revised the format of references, the names of journals, page numbers and punctuation one by one. The specific location of this modification can be found on page 17, lines 673-785.

Comments 13: The color gradient of the house price interpolation map in Figure (1) is excessive, failing to clearly display the spatial correspondence between the subway lines and the distribution of house prices. It is suggested to simplify it to highlight the key areas.

Response 13: Based on the suggestions made by the reviewers, we have modified Figure (1). We have simplified the number of breakpoints for Kriging interpolation and the decimal point annotations, deleted the complex gradient colors, and highlighted the key areas of residential prices along the urban rail transit lines. To comply with the concept of sustainable green development, we have accurately presented the price range from low to high by using the color gradient from light to dark in green. The specific location of this modification can be found on page 4, lines 168-170.

Comments 14: The formulas are not standardized. For instance, such as Moran’s I and Kriging interpolation formula may have problems like the undefined definitions of variables when they first appear in the formulas, and the mixed use of capital and lowercase letters for weight symbols. It is recommended to reformat them using a professional formula editor.

Response 14: Thank you for the reviewers' comments. We have re-formatted the formulas using the professional formula editor Mathtype. The specific location of this modification can be found on page 7, lines 256-281.

Comments 15: The combination of charts and text content is not tight. The clarity of the charts, the accuracy of the annotations, and the correspondence with the text content all need to be carefully checked and adjusted to ensure that the charts can effectively assist in explaining the research questions and results.

回应 15：根据审稿人的评论，我们重新核对了图表与正文中信息的对应关系，验证了表格中数据的完整性和正确性，并对正文中的部分信息和图表进行了修改。 此修改的具体位置可在第 6 页的第 243\291\305\333\350\354\451 行中找到

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

This manuscript addresses an increasingly relevant and timely issue: the spatial differentiation of housing prices in relation to urban rail infrastructure, specifically within the context of Qingdao, China. By employing the Multi-scale Geographically Weighted Regression (MGWR) model, the authors seek to uncover the non-stationary effects of various explanatory variables—distance to transit stations, floor area, green space coverage, and more—on housing values across a complex urban landscape. The study is carefully executed, methodologically sophisticated, and empirically rich. It offers significant technical contributions to spatial econometric modeling in urban studies, particularly by applying MGWR, a technique still relatively underused in this specific context.

The article is strongest in its empirical design and clarity of purpose. The authors succeed in demonstrating that the effect of rail transit on housing prices is neither uniform nor linear, and they convincingly show that MGWR provides superior explanatory power over traditional OLS or even GWR models. The use of a large and recent dataset (nearly 13,000 samples) enhances the reliability of the conclusions, while the maps and spatial visualizations are cleanly executed and informative. The findings—that housing prices show a peak responsiveness within approximately 373 meters of stations and that the effect varies by spatial and neighborhood characteristics—are both intuitive and substantively valuable for researchers and policymakers alike.

However, the manuscript would benefit greatly from more theoretical framing and reflexive interpretation. While the statistical results are presented with precision, they remain conceptually and institutionally under-contextualized. The analysis implicitly assumes a quasi-technocratic model of planning rationality, where housing prices respond to physical infrastructure in predictable ways. Yet urban development processes—especially in the Chinese context—are also shaped by fragmented institutional capacities, overlapping administrative jurisdictions, differential land-use governance, and uneven local policy execution. These elements are not fully accounted for in the current version of the paper.

he empirical sophistication of this paper, built on the application of MGWR to housing price differentiation near Qingdao’s rail infrastructure, constitutes a meaningful advancement in spatial modeling in the Chinese urban context. Yet its contribution could be markedly amplified through a more systematic engagement with the governance transformations, planning geographies, and epistemological debates that have characterized urban theory over the past two decades. To this end, four literatures—on soft spaces, risk-based regulation, social innovation in planning, and historical urban price dynamics—offer critical analytical leverage that would allow the authors to reposition their findings within broader scholarly conversations and policymaking dilemmas.

A first major axis of enrichment concerns the spatial governance of infrastructure development and its articulation across institutional boundaries. The notion of “soft spaces”, developed by Allmendinger, Haughton, Knieling, and Othengrafen (2015), is particularly relevant here. Soft spaces refer to flexible, often informal, and transcalar planning arenas—such as special economic zones, development corridors, or strategic spatial frameworks—that intersect, bypass, or overlap with formal administrative jurisdictions. These configurations, while not always legally binding, nonetheless exert real effects on spatial policy and development trajectories. In the case of Qingdao, the implementation of rail transit and associated urban renewal initiatives likely does not unfold exclusively through rigid administrative channels, but instead via evolving, hybrid governance spaces that escape full territorial fixity.

This framework would be especially well-suited to be introduced near the end of Section 2.2 (Research Framework), where the authors explain their choice of MGWR to capture multiscale effects. By recognizing that the observed spatial heterogeneity in housing prices may partly stem from planning through soft spaces, the authors could underscore that variability is not merely geographic—it is also institutionally produced through the ad hoc layering of planning instruments across nested governance levels. For instance, strategic rail investments or station-area regeneration initiatives might not map cleanly onto municipal boundaries, but instead reflect metropolitan-scale alliances or state-led urban entrepreneurialism. Incorporating a short paragraph in this section citing Allmendinger et al. (2015), and further substantiated by the genealogical insights of Purkarthofer and Granqvist (2021) on the global travel of soft space thinking, would provide a critical lens to interpret the spatial complexity their model reveals.

Secondly, to enrich the political and regulatory reading of the paper, the literature on risk-based regulation—particularly the work of Borraz et al. (2022)—offers a compelling analytical bridge. Borraz and colleagues explore how the implementation of supposedly standardized inspection and safety frameworks varies drastically depending on the national regulatory style, the organizational culture of the actors involved, and the institutional expectations of accountability. Although focused on food safety, their insights are readily applicable to urban planning regimes and the governance of infrastructure investment, especially in hybrid governance systems such as China’s. The differential valuation effects observed in Qingdao may not only be shaped by technical or spatial characteristics but may also be embedded within locally contingent logics of state intervention, risk tolerance, and administrative discretion.

This argument could be introduced in the Discussion section, as part of the reflection on why spatial effects differ so markedly across the territory. Citing Borraz et al. would allow the authors to frame their empirical findings within a broader concern about the heterogeneity of regulatory practice. In so doing, they would move from modeling what the price differentials are, to critically engaging with why the same infrastructure produces differentiated outcomes under varying configurations of public administration, investor confidence, and state-market relations. This line of reasoning would also provide a segue to bring in Le Galès and Vitale’s theory of fragmented governance, as discussed earlier, showing how the unevenness of governance practice is a constitutive feature of the urban political economy.

A third powerful vector of interpretive expansion lies in the literature on social innovation and transformative planning. The work of Nyseth and Hamdouch (2019) explicitly highlights the transformative potential of grassroots initiatives, civic networks, and alternative planning practices, particularly in cities facing complex urban challenges. While the current manuscript approaches housing price shifts from the perspective of rail proximity, it remains largely silent on the potential for civic agency or neighborhood-led adaptation in the face of rising values and speculative pressures. Integrating this perspective—perhaps in a brief forward-looking section in the Conclusion—would allow the authors to frame MGWR not just as a diagnostic tool, but also as a planning instrument that could support equitable governance responses. For example, MGWR could identify where housing market pressures are most acute, thereby enabling planners to prioritize these areas for inclusive zoning, rent stabilization, or participatory redevelopment schemes. Including a reference to Nyseth and Hamdouch would thus help link technical modeling with the broader urban question of “for whom” development occurs and how communities might contest, adapt to, or co-produce spatial change.

Relatedly, the transformative potential of the paper itself could be heightened by a reflection on how MGWR-based diagnostics might be more broadly embedded in socially responsive planning frameworks. This would help avoid the impression that the authors are simply identifying “winners” and “losers” of proximity-based value creation, and instead position them as contributing to an emerging literature on proactive, justice-oriented planning in the era of data-rich urbanism.

Also it is crucial not to diminish the ongoing relevance and analytical richness of hedonic price modeling, especially in the face of a sometimes uncritical valorization of newer, “smart” urban data tools. Despite the method’s long history, hedonic modeling remains an irreplaceable approach to capturing how context-specific characteristics influence urban value formation. Recent historical work—such as Barbot and Peroco’s (2019) study of 16th- and 17th-century Milan—demonstrates that even with sparse and incomplete data, hedonic principles can reveal deep insights into intra-urban social structure, residential choice, and price differentiation over the longue durée. The authors might integrate a paragraph—either in the Introduction or Conclusion—that not only defends but reclaims the value of hedonic approaches for understanding spatial dynamics. Such a paragraph could acknowledge that while machine learning, satellite data, and big data platforms have captured scholarly attention, the theoretical transparency, replicability, and interpretability of hedonic regression remain indispensable for longitudinal and comparative urban studies.

In the current paper, the authors could emphasize that the advances in geospatial econometrics, as exemplified by MGWR, represent not a break from the hedonic tradition, but rather an evolutionary leap that brings context-sensitivity and scale-awareness into a venerable methodological lineage. This would both honor the rigor of the current research design and position the paper as a bridge between urban data science and urban theory. The Milan example could be briefly cited to remind readers of the historical breadth and conceptual depth of hedonic reasoning as a tool for understanding urban transformation—not merely as a pricing model, but as a window into the relational production of space.

The authors could significantly strengthen their interpretation of spatial heterogeneity by incorporating insights from the literature on urban governance, particularly the work of Patrick Le Galès on the incompleteness and discontinuity of governance (DOI : 10.31235/osf.io/95zsc). As they have argued, governance in contemporary cities often unfolds through partial alignments, institutional bricolage, and multi-actor negotiation rather than through cohesive, linear implementation. In this light, the spatial variance identified through MGWR might be understood not only as a product of locational or structural factors but as a reflection of political fragmentation, planning ambiguity, and differentiated investment regimes.

This perspective could be easily integrated into Section 4.3, where the authors present the MGWR findings. One or two paragraphs—perhaps inserted at the end of that section—could introduce the idea that spatial modeling results should be situated within the broader institutional and governance ecologies in which they emerge. This would shift the analytical lens from a narrowly econometric view to one that is both spatial and political, enriching the interpretive horizon of the work.

Furthermore, the conclusion section should briefly revisit this theme, offering a more sociologically informed reflection on what the results imply for urban equity, planning responsiveness, and infrastructure-led development. Currently, the conclusion reiterates the technical superiority of MGWR and summarizes key spatial effects but does not extend into a discussion of how these patterns intersect with policy capacities or institutional limits. Doing so would offer stronger contributions to debates on transit-oriented development, spatial justice, and the politics of data-driven governance. In this sense 

Also, another area of improvement concerns the social implications of housing price increases near transit infrastructure. The authors commendably focus on quantifying spatial gradients of value, but they stop short of considering who benefits from these effects and who may be excluded. Given the growing literature on displacement, gentrification, and unequal accessibility in Chinese cities, it would be valuable to reflect—perhaps in a short discussion subsection—on how rising housing values might reinforce existing socio-spatial inequalities or generate new forms of exclusion. This would further align the study with the journal’s commitment to sustainable urbanism.

A final point pertains to the paper’s overall conceptual scaffolding. The authors rely on strong empirical methods, but they could frame their research questions and findings more explicitly in relation to urban policy debates. Posing the analysis within the language of urban transformation, infrastructural equity, and multiscalar governance would help readers understand not just what the results show, but why they matter. The paper makes a compelling case for using MGWR in urban economics, but it has the potential to say something broader—namely, that spatial modeling, when thoughtfully contextualized, can reveal the territorial traces of governance fragmentation and uneven development.

To summarize, this is a technically excellent, well-written article that makes a valuable contribution to the field of urban spatial analysis. With modest additions—particularly a sociological deepening of its interpretive claims, an engagement with governance theory, and a reflection on equity implications—it could become a reference point for both applied spatial analysts and critical urban scholars. I encourage the authors to take up these suggestions, which would not require major restructuring but rather thoughtful extensions in the discussion and conclusion.

 

Author Response

Comments 1: The notion of “soft spaces”, developed by Allmendinger, Haughton, Knieling, and Othengrafen (2015), is particularly relevant here. Soft spaces refer to flexible, often informal, and transcalar planning arenas—such as special economic zones, development corridors, or strategic spatial frameworks—that intersect, bypass, or overlap with formal administrative jurisdictions. This framework would be especially well-suited to be introduced near the end of Section 2.2 (Research Framework), where the authors explain their choice of MGWR to capture multiscale effects. By recognizing that the observed spatial heterogeneity in housing prices may partly stem from planning through soft spaces, the authors could underscore that variability is not merely geographic—it is also institutionally produced through the ad hoc layering of planning instruments across nested governance levels.

Response1：Thank you very much for your valuable comments. We have carefully read the "soft space" concept proposed by Allmendinger, Haughton, Knieling and Othengrafen (2015), combined with the literature of Purkarthofer and Granqvist (2021), and integrated it with the content of the MGWR model in this paper. The literature of Nyseth and Hamdouch (2019) helped the authors broaden their thinking in the discussion section, enhancing and enriching the practicality of the MGWR model. The specific location of this modification can be found on page 5, lines 221-242.

Comments 2: Secondly, to enrich the political and regulatory reading of the paper, the literature on risk-based regulation—particularly the work of Borraz et al. (2022)—offers a compelling analytical bridge. Borraz and colleagues explore how the implementation of supposedly standardized inspection and safety frameworks varies drastically depending on the national regulatory style, the organizational culture of the actors involved, and the institutional expectations of accountability. Although focused on food safety, their insights are readily applicable to urban planning regimes and the governance of infrastructure investment, especially in hybrid governance systems such as China’s. The differential valuation effects observed in Qingdao may not only be shaped by technical or spatial characteristics but may also be embedded within locally contingent logics of state intervention, risk tolerance, and administrative discretion. This argument could be introduced in the Discussion section, as part of the reflection on why spatial effects differ so markedly across the territory. Citing Borraz et al. would allow the authors to frame their empirical findings within a broader concern about the heterogeneity of regulatory practice.

Response 2：Thank you for the reviewers' comments. In the discussion section, we have incorporated the literature you recommended. The content of the literature has provided us with many inspirations and broadened our thinking. We have further supplemented the content of the discussion section of the article. The specific location of this modification can be found on page 15, lines 570-631.

Comments 3: Also it is crucial not to diminish the ongoing relevance and analytical richness of hedonic price modeling, especially in the face of a sometimes uncritical valorization of newer, “smart” urban data tools. Despite the method’s long history, hedonic modeling remains an irreplaceable approach to capturing how context-specific characteristics influence urban value formation. Recent historical work—such as Barbot and Peroco’s (2019) study of 16th- and 17th-century Milan—demonstrates that even with sparse and incomplete data, hedonic principles can reveal deep insights into intra-urban social structure, residential choice, and price differentiation over the longue durée. The authors might integrate a paragraph—either in the Introduction or Conclusion—that not only defends but reclaims the value of hedonic approaches for understanding spatial dynamics. Such a paragraph could acknowledge that while machine learning, satellite data, and big data platforms have captured scholarly attention, the theoretical transparency, replicability, and interpretability of hedonic regression remain indispensable for longitudinal and comparative urban studies.

Response 3：Based on your valuable suggestions, we have read and incorporated Barbot and Peroco (2019)'s research on Milan in the 16th and 17th centuries. The scholars therein reemphasize the significance of the characteristic price model. We also reaffirm the importance of the HPM in the concluding part of the article. Not only do we defend but also restate the value of the hedonic approach in understanding spatial dynamics, which remains the foundation of the new method in the future. The specific location of this modification can be found on page 17, lines 651-672.

Comments 4: Also, another area of improvement concerns the social implications of housing price increases near transit infrastructure. The authors commendably focus on quantifying spatial gradients of value, but they stop short of considering who benefits from these effects and who may be excluded. Given the growing literature on displacement, gentrification, and unequal accessibility in Chinese cities, it would be valuable to reflect—perhaps in a short discussion subsection—on how rising housing values might reinforce existing socio-spatial inequalities or generate new forms of exclusion. This would further align the study with the journal’s commitment to sustainable urbanism.

Response 4：We are grateful to the reviewers for their valuable suggestions, which have opened up new perspectives for the discussion part of this paper. We do not confine the statistical results to a single angle, but conduct a divergent and multi-angle analysis of the value of this research outcome, and further provide corresponding suggestions for the sustainable development of cities. The specific location of this modification can be found on page 16, lines 559-631. Through these modifications, the conclusion section has been made more specific and quantified, and it is capable of presenting the research's innovative points and its contributions to the scientific and practical fields clearly. At the same time, it ensures the consistency between the research content and the title, and emphasizes the sustainable value of this study.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Accept in present form.

Author Response

comments 1：Accept in present form.

Response 1：We sincerely thank the reviewers for their appreciation of our work and are honored by the decision to accept the manuscript in its current form.

Reviewer 2 Report

Comments and Suggestions for Authors

I really thank the authors for valuable improvements

I appreciate the great work they have done to take into account most of my suggestions. 

The paper it is almost ready for the publication, and it will have a lot of impact.

I just insist on one point that has not been considered, but it is of a great importance for Chinese cities, and for urban science in general. 

The authors could significantly strengthen their interpretation of spatial heterogeneity by incorporating insights from the literature on urban governance, particularly the work of Patrick Le Galès on the incompleteness and discontinuity of governance (DOI : 10.31235/osf.io/95zsc). As they have argued, governance in contemporary cities often unfolds through partial alignments, institutional bricolage, and multi-actor negotiation rather than through cohesive, linear implementation. In this light, the spatial variance identified through MGWR might be understood not only as a product of locational or structural factors but as a reflection of political fragmentation, planning ambiguity, and differentiated investment regimes.

This perspective could be easily integrated into Section 4.3, where the authors present the MGWR findings. One or two paragraphs—perhaps inserted at the end of that section—could introduce the idea that spatial modeling results should be situated within the broader institutional and governance ecologies in which they emerge. This would shift the analytical lens from a narrowly econometric view to one that is both spatial and political, enriching the interpretive horizon of the work.

Furthermore, the conclusion section should briefly revisit this theme, offering a more sociologically informed reflection on what the results imply for urban equity, planning responsiveness, and infrastructure-led development. Currently, the conclusion reiterates the technical superiority of MGWR and summarizes key spatial effects but does not extend into a discussion of how these patterns intersect with policy capacities or institutional limits. Doing so would offer stronger contributions to debates on transit-oriented development, spatial justice, and the politics of data-driven governance. 

Author Response

comments 1：The authors could significantly strengthen their interpretation of spatial heterogeneity by incorporating insights from the literature on urban governance, particularly the work of Patrick Le Galès on the incompleteness and discontinuity of governance (DOI : 10.31235/osf.io/95zsc). As they have argued, governance in contemporary cities often unfolds through partial alignments, institutional bricolage, and multi-actor negotiation rather than through cohesive, linear implementation. In this light, the spatial variance identified through MGWR might be understood not only as a product of locational or structural factors but as a reflection of political fragmentation, planning ambiguity, and differentiated investment regimes.

This perspective could be easily integrated into Section 4.3, where the authors present the MGWR findings. One or two paragraphs—perhaps inserted at the end of that section—could introduce the idea that spatial modeling results should be situated within the broader institutional and governance ecologies in which they emerge. This would shift the analytical lens from a narrowly econometric view to one that is both spatial and political, enriching the interpretive horizon of the work.

Furthermore, the conclusion section should briefly revisit this theme, offering a more sociologically informed reflection on what the results imply for urban equity, planning responsiveness, and infrastructure-led development. Currently, the conclusion reiterates the technical superiority of MGWR and summarizes key spatial effects but does not extend into a discussion of how these patterns intersect with policy capacities or institutional limits. Doing so would offer stronger contributions to debates on transit-oriented development, spatial justice, and the politics of data-driven governance. 

Response 2：

Thank you for the reviewer's suggestion. We have carefully read the literature you recommended. Patrick Le Gales pointed out in his work that urban governance is not a linear process, but a dynamic, incomplete and very diverse process. He emphasized the importance of comparing different urban governance practices. The paper criticizes the view that postmodern cities are seen as disorderly expansion. Governance structures still shape urban order through policy coordination, public services and infrastructure. Taking Cairo and Istanbul as examples, most of the existing literature focuses on social and cultural dynamics, while ignoring the analysis of power structures, policy alliances and multi-level governance mechanisms. The author emphasizes that governance is a process of coordination between multiple actors, and its success or failure directly affects inequality issues such as housing, education and security.

We introduced the views of scholar Patrick Le Gales at the end of Section 4.3. The specific location of this modification can be found in the Conclusion section on page 16, lines 559-572.In, we supplemented how the MGWR model intersects with policy capacity or institutional constraints and revised it according to your suggestions. The specific location of this modification can be found on page 18, lines 679-694.