The Population Flow under Regional Cooperation of “City-Helps-City”: The Case of Mountain-Sea Project in Zhejiang

Round 1

Reviewer 1 Report

The article ,,The Population Flow Network Under Regional Cooperation of "City-Helps-City": The Case of Mountain-Sea Project in Zhejiang" presented by the authors does not correspond to the research goal, which ambitiously states to contribute to a theoretical, empirical understanding of spatial political economy in transition China and to suggest implications for combating regional inequality through inter-city cooperation. By the way, the goal is indeed ambitious and interesting, and it is a pity that the authors did not realize this goal in each section. On the merits of the study a number of points should be clarified.

First, there is no unified research methodology. It is simply not stated, and we ended up with a mixture of different approaches, models, and methods. Second, the authors rigidly state the problem, making an affirmative conclusion without any evidence. They simply refer to documents and well-known methodologies and techniques. For example, to study the spatial structure of population flows, the materials state that an analysis of social networks (SNA) has been carried out in order to build a network structure of inter-jurisdictional. While social network analysis (SNA) has long been known in science, the authors do not explain the weighted network of inter-jurisdictional population flows based on the use of cell phone signals. It is also not clear how the semantic analysis of texts extracted from cell phones was conducted.

Further the authors used the Jieba package to segment the news into words and thus created a corpus of texts and gives the methodology including Shannon entropy and inter-jurisdictional population flows at different levels. That is, there is an affirmative text without evidence, in other words, the authors offer a completely separate study, which is detached from its stated purpose. Moreover, in section 4.3 a new purpose of the study is given, in which the authors try to find out whether poor counties can improve their position in the regional network by cooperating with other municipalities on the basis of the Mountain-Sea project.

Population flows left out, section and categorically stated, "digital collaboration was to connect poor counties with other places through digital platforms for retail and investment, technical assistance for manufacturing," and the conclusion: the results show that this variable? is not significant in terms of influencing population flows. It turns out that crucial digital cooperation has no effect on population flows. This may be true in China, but such important statements need proof. By the way, there are many such inaccuracies in the article. For example, the authors call on the central or provincial government to strengthen the institutional legitimacy of cooperation by expanding various regional development programs and policies. The call is good, but before that the authors conclude that, according to their models, institutional cooperation does not have a significant impact on the flow of population between counties.

In Section 6, the authors also categorically call for regional cooperation, especially City-Helps-City projects, to consider both the scope and depth of cooperation and to emphasize the role of institutional action for policy effectiveness. It would be good if this call was explored throughout the material and was, at the expense of appropriate methodologies and tools, evidence-based.

Overall, the article gives the impression of fragmented research and requires substantial revision. The title should be fixed in the purpose of the research, to show what methodology and methodology will be used and on the basis of what tools. The material should be systematized and the authors should explain the concepts and definitions used in the study. For example, what is meant by inter-jurisdictional ties, which flows are investigated with appropriate characterization and ranking.

Author Response

Responses to Reviewer 1

 

1. Overall

• The article presented by the authors does not correspond to the research goal, which ambitiously states to contribute to a theoretical, empirical understanding of spatial political economy in transition China and to suggest implications for combating regional inequality through inter-city cooperation.

 

Re: We appreciate the reviewer’s insightful comments that help us to improve the paper. The goal of this paper is to examine whether and how inter-city cooperation can increase population flows, enhance linkages and improve the situation of the backward areas. Our main contribution is to provide the perspective of the “city-helps-city” cooperation and understand its potentials to reduce regional inequalities. The current version corresponds better to the research goal. We have modified the paper following your suggestions below.

 

2. Methodology

• There is no unified research methodology. It is simply not stated, and we ended up with a mixture of different approaches, models, and methods.

 

Re: The methodology of previous version includes a mixture of different approaches of social network analysis, text semantic analysis, Shannon entropy, regression analysis and study on some cases for elaborations. This may confuse the readers regarding the research methodology. We have clarified that our primary methodology (unified approach) is the regression models since our research question is to examine whether inter-jurisdictional cooperation in “city-helps-city” types (Mountain-Sea cooperation projects) affects population flows between poor places and other developed localities in Zhejiang. The cases from the cooperation news are concrete examples to better interpret model results and understand details of cooperation modes in different fields.

 

We reorganized section 3 on methodology to highlight the central role of regression models. Other approaches are just employed to generate the dependent variables of population flows (social network analysis) and the main explanatory variables of Mountain-Sea cooperation (text semantic analysis and Shannon entropy) for the regression models.

 

Specifically, the dependent variables of population flows (inflow and outflow) are generated through social network analysis, using the mobile phone signal data that tracks the locations of users. To measure the independent variables of Mountain-Sea cooperation, we rely on the text semantic analysis to extract the key information (cooperation fields, county names, etc.) of cooperation from the news data on the official website of Zhejiang Provincial Development and Reform Commission. Then, text semantic analysis calculates word counts of each identified field to measure the intensity of cooperation (CopFoc); the Shannon entropy is used to calculate the diversity (scope) of cooperation fields (CopDiv); we further identify whether a county pair is mandated officially by the central and provincial governments or not to determine the legitimacy of cooperation (CopMand). These approaches are essential to generating our dependent variables of population flows and the independent variables of the intensity of cooperation on different fields, the diversity of cooperation, and the legitimacy of cooperation.

 

• Moreover, in section 4.3 a new purpose of the study is given, in which the authors try to find out whether poor counties can improve their position in the regional network by cooperating with other municipalities on the basis of the Mountain-Sea project.

 

In other words, the authors offer a completely separate study, which is detached from its stated purpose.

 

Re: Yes, “try to find out whether poor counties can improve their position in the regional network by cooperating with other municipalities on the basis of the Mountain-Sea project” is a part of our research question, which has been emphasized in the new version. We agree with your comments that section 4 might contain diverse materials that may distract readers. In fact, 4.3 is our primary research question, corresponding to our regression models. Therefore, we modified the section 4 to ensure our analysis to be more focused on our research question. We shortened and changed the texts of 4.1 and 4.2 into the descriptive analysis of our dependent and independent variables. The descriptive statistical analysis and possible visualization for model variables allow us to better understand our Y (population flows between the poverty counties and the developed places) and X (their cooperation in different measures that are intensity, diversity and legitimacy). This step is necessarily conducted before regression models.

 

We also moved the part of degree centrality and the top 20 connected counties (population flows) to the appendix since it is not that relevant to our regression models and just provides information on how well the poverty counties link with the developed places in terms of population flows. This might be the main reason that you get distracted. In addition, we moved Table 1 into earlier texts to make readers aware that 4.2 is about the description of independent variables. Last but not least, we added the denotations in the regression models to remind readers how the texts in 4.1 and 4.2 correspond to our regression models.    

 

• The authors used the Jieba package to segment the news into words and thus created a corpus of texts and gives the methodology including Shannon entropy and inter-jurisdictional population flows at different levels.

 

Re: The Jieba package is used for the first step to decompose the cooperation news from government websites so that we can identify the key information of cooperation fields for each county pair under Mountain-Sea projects. This step is the foundation for our construction of independent variables. For example, for the intensity of cooperation field (e.g., industrial development, educational services, tourism etc.), we can calculate the word counts for the identified 9 major categories of cooperation. The Shannon entropy can be further applied to measure the diversity (scope) of cooperation fields by considering both the number and the intensity of collaboration.

 

We only focus on the county-level in this study for both our Y (population flows) and X (cooperation) variables. The county-level unity in our study refers to 3 types of county-level municipalities in China, including counties, districts and county city. This information can be found in section 3.1 when we introduced our study region and unit. In the second paragraph of section 3.2, we further clarified that our models only covered county pairs between poverty places and developed places since we were interested more on the “city-helps-city” cooperation under the state-driven agenda of Mountain-Sea Projects in Zhejiang (pilot). 

 

• While social network analysis (SNA) has long been known in science, the authors do not explain the weighted network of inter-jurisdictional population flows based on the use of cell phone signals. It is also not clear how the semantic analysis of texts extracted from cell phones was conducted.

 

Re: We appreciate that the reviewer has pointed out these two issues in the methodology part. First, for the SNA, we moved the explanation of network pattern and degree centrality to the appendix. As our dependent variables are the number of population flows between any two county-level municipalities based on our mobile phone signaling data, we want to make sure readers will not be distracted by other contents that are not directly related to our main methodology (regression models). Second, the semantic analysis extracted information from the Mountain-Sea cooperation news, which were retrieved from the official website of Zhejiang Provincial Development and Reform Commission. In the 3.2 of data sources, we have the materials about these two data sources. The process to extract information from the news data using Jieba is described in the third paragraph of section 3.3.3. Based on the calculation of the weight of TF-IDF (Term Frequency - Inverse Document Frequency) through Scikit-learn (Sklearn), we can extract the keywords of text corpus, and finally identify the key cooperation fields. In other words, we have a collected 203 pieces of cooperation news (texts) and applied text semantic analysis to extract information from them. However, again, this is not the separate analysis but for the generation of independent variables (X). 

 

To sum up, the dependent variables of population inflow and outflow (Y) were constructed based on cell phone data since it is location-based spatial big data that tracks people’s movement. Our partner, the China Mobile, provides this database of Zhejiang province for our study. The independent variables (X) of cooperation intensity for each field, diversity of cooperation and legitimacy (whether it is mandated) are constructed through text semantic analysis on the Mountain-Sea cooperation news data instead of cell phone data.

 

The relevant texts have been modified to clarify the above details of methodology and resolve your concerns. We appreciate your comments to help enhance our description of methodology.

 

• The title should be fixed in the purpose of the research, to show what methodology and methodology will be used and on the basis of what tools. The material should be systematized…

 

The article gives the impression of fragmented research.

 

Re: The fragmented impression aroused from the diverse analysis, which distracted readers. After we reorganized our materials, removed less relevant analysis, shortened the texts, changed languages (description for dependent and independent variables 4.1 and 4.2) and kept focused on our question (whether and how Mountain-Sea cooperation increase inter-jurisdictional linkages on population flows of poverty places with other developed places), we believe that current version is more systematized (by regression models) and can address the fragmented problems.

 

For the title, we get rid of the word “network” since our dependent variable is the population flows between poverty counties and their paired municipalities in economic advantages. The network pattern is just the outcome of descriptive analysis on our dependent variables (Y). We feel the rest of the title fits the current version of paper when we clearly stated our research question (goal) is to examine whether and how inter-jurisdictional cooperation in the type of “ city-helps-city” under Mountain-Sea agenda impacts the population flow between the poor and developed places.

 

 

3. Evidence

 

• The authors rigidly state the problem, making an affirmative conclusion without any evidence.

 

That is, there is an affirmative text without evidence,

 

Re: This point is very valuable to help us improve the paper. Our conclusions are drawn from three sources, including the analytical results (descriptive and regression results), the news data (for examples of cooperation), and the literature. The news data not only contributes to the generation of independent variables on cooperation, but also provides concrete examples of “city-helps-city” cooperation to better understand how Mountain-Sea projects impacts population flows. In this sense, the methodology of this paper comprises of both quantitative models and qualitative interpretations and discussions. 

 

In the revision, we followed your suggestions to highlight the source of our points/conclusions through the paper. For example, in 4.3, we clarified what points are directly from the model results and what are interpretations based on the news data examples. Moreover, in the section 5 of discussion, we added the citations to support our argument. Again, the examples were retrieved from the official web news on the Mountain-Sea Cooperation Projects. We also gave the suffix of the county pair names. In addition, when we discussed the results in section 5, we also referred to the evidence in descriptive results in 4.2 that is about all the measures of cooperation. For example, cooperation on digital service, educational and health services are all found popular in Mountain-Sea agenda in 4.2 (Figure 4). The last section of conclusion has been modified in the same way. For example, when we tried to highlight the connections (population flow) between poverty counties and developed places, we referred to the descriptive results of our dependent variables in 4.1 to give evidence for this argument. We also added citations in the last section.

 

 

• Population flows left out, section and categorically stated, "digital collaboration was to connect poor counties with other places through digital platforms for retail and investment, technical assistance for manufacturing," and the conclusion: the results show that this variable? is not significant in terms of influencing population flows. It turns out that crucial digital cooperation has no effect on population flows. This may be true in China, but such important statements need proof.

 

Re: We modified the arguments on digital cooperation.

 

In the section 4.2, we changed the texts to directly say what we have found from Figure 4 that cooperation on tourism and digital technology account for relatively higher share comparing to other categories.

 

In 4.3 ("digital collaboration was to connect poor counties with other places through digital platforms for retail and investment, technical assistance for manufacturing,"), we clarified this as the expected role of digital cooperation based on the news data examples. In the Mountain-Sea cooperation news data, we see that county pairs are actively pursuing this type of cooperation on digital services for retail (e-commerce), investment and manufacturing. However, we also pointed out digital cooperation is not significant to yield effects according to our regression results. We also gave possible explanation on this difference between “hypothesis” and model results. The insignificant impacts of digital cooperation on increasing population flows and enhancing linkages of poverty counties with others might be attributed to the lack of digital capacity in the backward areas.

 

In 5.2, we listed the example of Taishun County-Lucheng District to illustrate how digital cooperation is pursued to develop e-commerce in the poverty county of Taishun under Mountain-Sea projects. We added citations, explained why we failed to observe the significant effects of digital cooperation in our models and gave recommendations for future efforts.

 

• For example, the authors call on the central or provincial government to strengthen the institutional legitimacy of cooperation by expanding various regional development programs and policies. The call is good, but before that the authors conclude that, according to their models, institutional cooperation does not have a significant impact on the flow of population between counties.

 

In Section 6, the authors also categorically call for regional cooperation, especially City-Helps-City projects, to consider both the scope and depth of cooperation and to emphasize the role of institutional action for policy effectiveness. It would be good if this call was explored throughout the material and was, at the expense of appropriate methodologies and tools, evidence-based.

 

Re: Our model results of how institutional legitimacy of cooperation affects population flows are mixed. In the paper, there are indeed two variables that can be regarded as the proxies for the institutional legitimacy building.

 

One is the CopPlan for the inter-city cooperation on the joint spatial plan or regional planning (urban construction, management etc.). The literature [25-27] argues that the institutional cooperation is a “higher stage” of inter-city cooperation to secure the implementation of programs. Even though we have found examples in the news data (e.g., section 5.3 Pingyang County-Yueqing City) on such type of cooperation, this variable is not significant in our models. This might be partly attributed to the fragmented institutional settings of China's regional government systems and the conflicts of local interests.

 

The call on the central or provincial government to strengthen the institutional legitimacy is from results on the other important variable (CopMand). According to our models, the mandated cooperation by Mountain-Sea agenda is significant to influence population flows. Therefore, compared with the voluntary county pairs in “city-helps-city” collaboration (CopMand=0), the top-down assignment (CopMand=1) imposes legitimacy to stimulate more effective cooperation and to improve the situation of poverty places.

 

Therefore, we call for the higher-level government to build institutional legitimacy for inter-jurisdictional cooperation, which might be through mandates or local empowerment (CopMand). For the joint spatial planning or regional agreement (CopPlan), we only argued for future improvement (in both section 5.3 and 6) to overcome the fragmentation of institutional settings and potential conflicts of local interests [25-27], which may undermine the effectiveness of institutional cooperation. We believe that this revision is able to address the inaccuracies with respect to the arguments on institutional cooperation to build legitimacy.

 

• The authors should explain the concepts and definitions used in the study. For example, what is meant by inter-jurisdictional ties, which flows are investigated with appropriate characterization and ranking.

 

Re: In the revised version, the important concepts and definitions were explained, especially for model variables. Our dependent variables are both population inflows and outflows between poor places and more advantaged municipalities. This kind of linkages may go beyond geographic distance, which is not like the flows within e.g., metropolitan areas (central cities and adjacent places). We want to examine if this type of linkages can be explained by “ city-helps-city” cooperation.

 

The ranking of population inflow and outflow (degree centrality) for each county have been moved to appendix.

 

The three dimensions of independent variables are defined in the paper. We would like to keep readers in mind that the intensity of cooperation focuses (for different fields), the diversity of cooperation fields, and whether the cooperation is legitimated by the central and provincial government are the variables of our main concern for the investigation on whether and how cooperation affects population flows. 

Author Response File: Author Response.pdf

Reviewer 2 Report

This paper is interesting for the journal as it contributes to theoretical and empirical understandings of the new spatial political economy in transitional China and provides implications for reducing regional inequalities through inter-city cooperation. Nevertheless, it still needs some changes before its publication. 

The case study is quite good, but the theoretical framework needs more international-based insights on the following issues:

- the distortions provoked by capitalist urbanization processes by taking into references some examples of global cities:

- 2020. Alpha City: How London Was Captured by the Super-Rich. London: Verso

- 2019. Capital City. Gentrification and the real estate state. London-New York: Verso

 

- 2017. The icon project: architecture, cities and capitalist globalization. New York: Oxford University Press

- the cooperation on the ways to foster urban planning towards healthier cities and their features, such as mobility, https://www.sciencedirect.com/science/article/pii/S2214109X22000699, https://link.springer.com/chapter/10.1007/978-3-031-10542-5_18

- the cooperation to organize large mega-events, https://www.tandfonline.com/doi/full/10.1080/09669582.2020.1870989?src=recsys

- the cooperation on how to redevelop brownfield infrastructure, https://www.elgaronline.com/view/book/9781800375611/book-part-9781800375611-18.xml

Moreover, in the conclusions try to better address the lessons learned from the spatial regime of “city-helps-city” cooperation for other countries.

 am expecting the authors will address these issues before stating whether the paper is ready for publication.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The article is recommended for publication

Reviewer 2 Report

Th paper is ready for publication