Valuation of Urban Parks Under the Three-Level Park System in Shenzhen: A Hedonic Analysis

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

In scientific literature, it has been scientifically proven that the presence of urban parks significantly contributes to the improvement of urban areas and the quality of life of residents. By using the hedonic price method, specifically the 3-D spatial multilevel autoregressive model, the authors studied the influence of three types of urban parks in Shenzhen, China, on property values.

While the article is well-structured and tackles an interesting topic, it is worth noting the complete absence of a paragraph regarding the literature review, where the results obtained by other scholars along with the methodologies previously used are reported. The method employed effectively explains the phenomenon, and the conclusions clearly elucidate the research findings.

It is recommended that the authors specify the contribution of their study to the literature, highlighting how it fits into existing literature and what new insights it brings.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

LAND -3380712 Valuation of urban parks under the three- level park system in Shenzhen: a hedonic analysis

On the surface this is a well written research article, but more work is needed before it is publishable. Specific recommendations are made to the authors.

Abstract: is OK

Introduction 

The literature review needs to be more extensive and nuanced. There are more international studies of hedonic valuation of views of parks or greenspaces. This literature needs to be separated from more general amenity values related to urban park proximity versus recreational value access to park proximity. Then the authors need to posit their research questions.

Methodology

Lines 137-148 As this reviewer understands it – there are two sets of research questions. One addresses the hedonic price relationship to views of urban parks and the other addresses the hedonic relationship of distance access to the three types of urban parks. What about the differences between rental apartments and owner-occupied residences?

Also, how to we know that those residences or apartments actually have views to the urban parks?  Will this make a difference in the results?

Line156 – should be “employs”?

Line 187- can be neglected?

Line 211- please define “burn in period”

Results and Discussion

Lines 265-272- how to you know that these are the reasons for the value of views of the country park? Could it be that the residences near the country park have more visual access because there is a large viewshed from residences to the park? 

Line 281- how do you know that community parks have limited viewsheds? Did anyone do a viewshed analysis? It seems the whole issue of amenity viewshed valuation is skewed by the location of the larger country park plus its size. This may be similar in some ways to Central Park in NY City or other larger urban parks which have high visual access and prized views for those surrounding the urban park.

Accessibility results and discussion is more straight forward but should be discussed in relationship with other hedonic valuation studies.

Line 349- should be “zone”?

Conclusion

The authors should have a more nuanced conclusion for country parks regarding both view and physical access with mediating factors. Then the same for both city parks and community parks with mediating factors.

It seems that a future research study could be actual viewshed analysis for residential units facing different urban park types tied to a hedonic value analysis would be merited.

Author Response

1. Comment for Introduction: The literature review needs to be more extensive and nuanced. There are more international studies of hedonic valuation of views of parks or greenspaces. This literature needs to be separated from more general amenity values related to urban park proximity versus recreational value access to park proximity. Then the authors need to posit their research questions.

Response: The literature review has been expanded to incorporate additional international studies on the hedonic valuation of views of parks or greenspaces (Lines 76-83). Furthermore, the literature on amenity values associated with urban park proximity has been distinctly separated from discussions on recreational value access to parks, offering a clearer differentiation between these concepts (Lines 84-88). This refinement enhances the clarity and depth of the review, ensuring a more comprehensive understanding of the topic.

 

2. Comment for Methodology:

1) Lines 137-148 As this reviewer understands it – there are two sets of research questions. One addresses the hedonic price relationship to views of urban parks and the other addresses the hedonic relationship of distance access to the three types of urban parks. What about the differences between rental apartments and owner-occupied residences?

2) Also, how to we know that those residences or apartments actually have views to the urban parks?  Will this make a difference in the results?

3) Line156 – should be “employs”?

4) Line 187- can be neglected?

5) Line 211- please define “burn in period”.

Response:

1) The study acknowledges the potential differences between rental apartments and owner-occupied residences, particularly in how they value urban park views. To address this, the methodology focuses on commercial housing transaction data, ensuring consistency in the analysis (Lines 117-130). Future research could further explore how housing tenure influences the perceived value of park views and accessibility.

2) The study utilizes detailed geographic information systems (GIS) data and aerial imagery from Baidu Maps to accurately determine the line of sight and proximity of each property to the nearest park, confirming the presence of park views (Lines 119-123). This approach ensures the reliability of the results regarding the impact of park views on housing prices. The distinction between properties with and without park views is crucial for understanding the hedonic value of urban parks, and future research could further explore this dynamic.

3) It has been revised in the corresponding text (Line 226).

4) Corrections have been made in the corresponding text (Line 259).

5) Burn in period refers to the initial phase of the sampling process where the chain is allowed to reach a state of convergence and stability before collecting samples for inference. During the burn-in period, the chain explores the parameter space and adjusts its trajectory to reach the target distribution. The purpose of the burn-in period is to mitigate the influence of the chain's starting point or initial conditions on the final results. By discarding the samples generated during the burn-in phase, the MCMC algorithm allows the chain to settle into the target distribution and ensures that the subsequent samples are representative of the posterior distribution of interest. We have add the description in Lines 282-283.

 

3. Comment for Results and Discussion:

1) Lines 265-272- how to you know that these are the reasons for the value of views of the country park? Could it be that the residences near the country park have more visual access because there is a large viewshed from residences to the park? 

2) Line 281- how do you know that community parks have limited viewsheds? Did anyone do a viewshed analysis? It seems the whole issue of amenity viewshed valuation is skewed by the location of the larger country park plus its size. This may be similar in some ways to Central Park in NY City or other larger urban parks which have high visual access and prized views for those surrounding the urban park.

3) Accessibility results and discussion is more straight forward but should be discussed in relationship with other hedonic valuation studies.

4) Line 349- should be “zone”?

Response:

1) The limitation in our methodology regarding the lack of a viewshed analysis has been acknowledged, and we have added a discussion on this point (Lines 436-440), noting the potential influence of viewshed on the value of park views and the need for future research to address this aspect. And the discussion has been revised to incorporate the possibility that visual access to country parks might be affected by a broader viewshed (Lines 337-345). Additionally, the importance of considering mediating factors, including transportation and recreational opportunities, when assessing the influence of park views on housing prices has been highlighted (Lines 488-495).

2) Acknowledgment is given that a formal viewshed analysis was not conducted in the study. The conclusions were instead based on the general characteristics of community parks in Shenzhen, which are often smaller and situated in "fragmented land" within urban corners, as reported by the Shenzhen Municipal Bureau of Planning and Natural Resources (Lines 348-351). These parks are primarily designed for specific functional use, such as recreation and exercise for nearby residents (Lines 351-352). However, the importance of a more detailed viewshed analysis to accurately assess the visibility of parks from residences is recognized, and a discussion on the need for such analysis in future studies has been added (Lines 441-448).

3) The discussion on accessibility has been expanded to include a comparison with other hedonic valuation studies, particularly those focusing on urban parks and greenspaces (Lines 495-505). This revision highlights how the findings align with or diverge from existing literature, providing a more comprehensive context for the results. Specifically, the study now references key works that explore the relationship between park accessibility and housing prices, such as studies by Wu et al. (2017) and Kong et al. (2007), to better situate the findings within the broader academic discourse.

4) It has been modified to the corresponding text (Line 419).

 

4. Comment for conclusion:

The authors should have a more nuanced conclusion for country parks regarding both view and physical access with mediating factors. Then the same for both city parks and community parks with mediating factors.

Response:

Appreciation is expressed for the suggestion to provide more nuanced conclusions regarding the view and physical access to different types of parks, considering mediating factors. In response, the discussion has been expanded to include the following (Lines 488-505):

1. Country Park: The conclusions have been refined to consider the mediating factors that may influence the value of views and physical access to country parks. These factors include the size and location of the park, which can affect the viewshed from residences, as well as the recreational opportunities and conservation initiatives offered by these parks.
2. City Park: The revised manuscript now discusses the importance of city parks in providing recreational opportunities and their central location, which can influence their visibility and accessibility.
3. Community Park: The conclusions have been updated to reflect the potential mediating factors that may affect the appeal of community parks' views, such as noise pollution from community events and the limited viewshed due to their smaller size and location.

Reviewer 3 Report

Comments and Suggestions for Authors

*The author should add some parallel research and emphasize the basis of hedonic price method in the introduction to highlight the uniqueness and innovation of the research.

 

*The manuscript does not adequately explain why a 3-D spatial multilevel autoregressive model was selected. Furthermore, the criteria for choosing a specific spatial weight matrix were insufficiently discussed, which affects the transparency and replicability of the results. Additionally, the use of a semi-logarithmic function may not fully capture nonlinear relationships between housing prices and independent variables, particularly at extreme values. A more thorough justification of these methodological choices would strengthen the study's credibility.

 

*Treating parks of the same type as homogeneous oversimplifies the analysis. Differences in characteristic indicators within park types should be accounted for in calculations and analyses to control for spatial features that might otherwise skew the results.

 

*The absence of survey or qualitative data to verify whether proximity and visibility of parks align with residents' preferences and behaviors limits the explanatory power of the hedonic pricing model. Including such data would provide a more robust interpretation of the findings.

 

*Line113-118. The rationale for using data from January 2016 to December 2017 is unclear. Since then, significant events, including real estate turbulence, the COVID-19 pandemic, and shifts in national urban construction police, have profoundly impacted the urban construction in China. Data from 7-8 years ago may no longer be representative for evaluating the value of urban parks in Shenzhen today.

 

*Line 287. I don't think a simple walking distance can reflect on accessibility results very well.

Comments for author File: Comments.pdf

Author Response

Comment 1: The author should add some parallel research and emphasize the basis of the hedonic price method in the introduction to highlight the uniqueness and innovation of the research.

Response 1: A more detailed comparison with parallel research in the field has been incorporated (Lines 48-53), emphasizing the distinctive aspects of the study. Additionally, the theoretical underpinnings of the hedonic price method have been expanded upon (Lines 56-62), drawing attention to how the approach advances the current understanding of this methodology and its application to the research.

 

Comment 2: The manuscript does not adequately explain why a 3-D spatial multilevel autoregressive model was selected. Furthermore, the criteria for choosing a specific spatial weight matrix were insufficiently discussed, which affects the transparency and replicability of the results. Additionally, the use of a semi-logarithmic function may not fully capture nonlinear relationships between housing prices and independent variables, particularly at extreme values. A more thorough justification of these methodological choices would strengthen the study's credibility.

Response 2: The reason of using 3-D spatial multilevel autoregressive model is given in Lines 226-230. This model has been examined to be able to adequately account for all potential spatial effects (including spatial heterogeneity and spatial dependence) in the 3-D spatial context of a populous modern city, so as to enhance the model robustness and improve the estimate accuracy. The use of semi-logarithmic function is a common practice in hedonic price modelling, which is used to mitigate heteroscedasticity issue. The justification of the methodological choices of spatial weights has been given clearly in Lines 363-375.

 

Comment 3: Treating parks of the same type as homogeneous oversimplifies the analysis. Differences in characteristic indicators within park types should be accounted for in calculations and analyses to control for spatial features that might otherwise skew the results.

Response 3: Appreciation is expressed for the reviewer's comment on the oversimplification of treating parks of the same type as homogeneous. Future research will address this by examining characteristic indicators within park types to account for spatial disparities that could skew results (Lines 131-135). The study will provide a more accurate assessment of the impact of urban parks on housing prices in Shenzhen by incorporating a detailed analysis of park features such as size, amenities, and maintenance, which influence visitors' positive emotions and park use, as well as their potential effect on property values (Lines 135-140).

 

Comment 4: The absence of survey or qualitative data to verify whether proximity and visibility of parks align with residents' preferences and behaviors limits the explanatory power of the hedonic pricing model. Including such data would provide a more robust interpretation of the findings.

Response 4: The acknowledgment is made that the absence of survey or qualitative data limits the explanatory power of the hedonic pricing model in this context. Future research will address this limitation by incorporating such data to verify the alignment between park proximity, visibility, and residents' preferences and behaviors (Lines 141-144). This approach will provide a more robust interpretation of the findings by integrating residents' perspectives on the value of urban parks, which is essential for understanding the true impact on housing prices (Lines 144-150).

 

Comment 5: Line 113-118. The rationale for using data from January 2016 to December 2017 is unclear. Since then, significant events, including real estate turbulence, the COVID-19 pandemic, and shifts in national urban construction policy, have profoundly impacted the urban construction in China. Data from 7-8 years ago may no longer be representative for evaluating the value of urban parks in Shenzhen today.

Response 5: The dataset spanning from January 2016 to December 2017 offers a historical perspective on the relationship between urban parks and housing prices in Shenzhen. However, it is recognized that the period since has been marked by significant events that have reshaped the urban landscape and housing market dynamics, including real estate turbulence, the COVID-19 pandemic, and shifts in national urban construction policy (Lines 178-184). These factors have undoubtedly influenced the value of urban parks in the context of current housing prices. To ensure that the assessment of urban parks' impact on housing values remains current and reflective of the evolving urban environment in Shenzhen, it is essential to incorporate more recent data (Lines 184-188).

 

Comment 6: Line 287. I don't think a simple walking distance can reflect on accessibility results very well.

Response 6: Accessibility in urban parks is indeed a multifaceted construct that extends beyond mere proximity, as indicated by the study's initial focus on walking distance as an initial measure of accessibility. Recognizing this, a more holistic approach is necessary to fully capture the accessibility premium and its impact on housing prices in Shenzhen (Lines 449-453). Future research will encompass a broader range of factors that contribute to accessibility, such as traffic intensity, road width, and walkability, to provide a more comprehensive assessment of how these elements influence the value of housing in relation to urban green spaces (Lines 453-458).

In addition to these revisions, we have made several other minor changes to improve the clarity and coherence of the manuscript. We believe that these revisions have strengthened our study and made a significant contribution to the field. We appreciate the opportunity to revise and resubmit our work to Land and hope that it is now acceptable for publication. Please do not hesitate to contact us if you require any further information or clarification.

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Authors have adequately responded to all reviewer comments.

Reviewer 3 Report

Comments and Suggestions for Authors

I still have doubts about using data from January 2016 to December 2017, as outdated data may not accurately reflect the current situation, leading to a significant decline in the research's relevance and value.