Study on the Evolution of SCO Agricultural Trade Network Pattern and Its Influencing Mechanism

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The paper studies agricultural trade between China and SCO countries using the network approach (TERGM). While this statistical approach has been implemented in the analysis of international trade before, this is the first attempt to study agricultural trade with China. In short, I am positive about the publication of this article, but I see several clarifications and further discussions that are necessary before publication.

 

Main points

1. There are several concerns about the export decision function, i.e., Equation (6). 

1) Please justify the $1 million threshold value used to create the dependent variable. Given that trade volume differs among countries, I fail to understand the appropriateness of this common value as a threshold. 

2) This equation should be explicit about the timing of explanatory variables recorded (e.g., t+1, t, or t-1). It is not clear about information of which years are contributing to the current export decisions. 

3) Equation (6) should be expressed in the standard manner, using subscripts i, j, and t. For example, home GDP should be shown as beta_{1}lngdp_{i, t?} rather than beta_{i1}lngdp. This equation currently contains the same variables (e.g., lngdp) with different parameters. 

 

2. Given that the Chinese population and food demands are expected to decline soon, the paper will be more interesting if the authors discuss the future importance of food trade for China when discussing policy implications in conclusion.

 

3. Line 637: The paper focuses on Chinese trade and hardly discusses the benefits and drawbacks of trade with China. For win-win outcomes, both countries must benefit from agricultural trade. In that sense, why not provide statistical evidence showing that SCO countries benefit from agricultural trade with China if any? SCO countries may have merely altered their export destination and may not have benefited economically overall. This point arises because research is limited to SCO countries and, like Comment 2,  is important for publication in this journal since it is directly related to the sustainability of such trade policy.

 

Editorial points

4. The authors should explain the statistical method (TERGM) and its estimation method in detail. The paper refers to the R components but does not provide specific model settings.

 

5. The abbreviations (e.g., QAP and ERGM) that appear first in the paper should be spelled out.

 

6. What does the number in parentheses refer to in Tables 4 & 5?

 

7. Table 5 lacks the number of observations for Model 10 and the AIC, BIS, and Log Likelihood information for Model 9.

 

8. Equation (6) requires a closing parenthesis.

 

9. To my knowledge, the “time-indexed random graph model” (TERGM) is not a common expression in the literature or is inconsistent with the abbreviations. I wonder if you mean the Temporal Exponential-Family Random Graph Models.

 

10. What are the theoretical relationships between export probability and its determinants in Equation (6)? I failed to understand their expected sign and judge if the estimators were appropriate.

End

Author Response

Comments 1: There are several concerns about the export decision function, i.e., Equation (6). 

1) Please justify the $1 million threshold value used to create the dependent variable. Given that trade volume differs among countries, I fail to understand the appropriateness of this common value as a threshold. 

2) This equation should be explicit about the timing of explanatory variables recorded (e.g., t+1, t, or t-1). It is not clear about information of which years are contributing to the current export decisions. 

3) Equation (6) should be expressed in the standard manner, using subscripts i, j, and t. For example, home GDP should be shown as beta_{1}lngdp_{i, t?} rather than beta_{i1}lngdp. This equation currently contains the same variables (e.g., lngdp) with different parameters. 

Response 1）: Thank you for your valuable suggestion. For creating the critical value of US$1 million for the dependent variable, we mainly refer to Li Jing et al. (2017), Journal of Management World, Compound Impact Factor: 31.204, and Luo Chao-liang et al. (2022), Journal of International Trade Issues, Compound Impact Factor: 5.335; and also formulate it on the basis of the actual amount of agricultural trade of the SCO countries with a high degree of reasonableness.

Response 2）: Thank you for your valuable suggestion. The formation of the networks is interdependent in time, which means that the formation of the second network is dependent on the first network, the formation of the third network may be dependent on the first two networks, and so on In this paper, the agricultural trade networks from 2003 to 2022 are interdependent.

 

Response 3）: Thank you for your valuable suggestion. When constructing the TERGM model, all authoritative journals we checked used the format in beta_{i1}lngdp.

 

Comments 2: Given that the Chinese population and food demands are expected to decline soon, the paper will be more interesting if the authors discuss the future importance of food trade for China when discussing policy implications in conclusion.

 

Response 2: Thank you for your valuable suggestion.

 

Comments 3: Line 637: The paper focuses on Chinese trade and hardly discusses the benefits and drawbacks of trade with China. For win-win outcomes, both countries must benefit from agricultural trade. In that sense, why not provide statistical evidence showing that SCO countries benefit from agricultural trade with China if any? SCO countries may have merely altered their export destination and may not have benefited economically overall. This point arises because research is limited to SCO countries and, like Comment 2,  is important for publication in this journal since it is directly related to the sustainability of such trade policy.

 

Editorial points

 

Response 3: Thank you for your valuable suggestion. The increase in the volume of trade shows that both sides have benefited, and it is common sense to avoid harm. Furthermore, geographic proximity and proximity in itself saves a lot of costs and thus increases the benefits.

 

Comments 4: The authors should explain the statistical method (TERGM) and its estimation method in detail. The paper refers to the R components but does not provide specific model settings.

Response 4: Thank you for your valuable suggestion. We have revised the suggestion part in the Point 4, and labelled yellow in the article. We used the commands inside R components to complete the empirical analyses, not any model, so we didn't write about the process of model construction. The revisions are as follows:

TERGM (Temporal Exponential Random Graph Model) is an extension of ERGM (Exponential Random Graph Model) specifically designed for processing and analysing time-dependent network data. It combines theories from network science and social science and aims to explain relationships in networks and their evolution over time through a generative mechanism. The model is able to capture dynamic changes in the structure of a network, including the formation, disappearance and alteration of relationships between nodes, and how these changes are affected by other factors in the network (e.g., node attributes, network structural features, etc.)

 

Comments 5: The abbreviations (e.g., QAP and ERGM) that appear first in the paper should be spelled out.

Response 5: Thank you for your valuable suggestion. We have revised the suggestion part in the Point 5, and labelled yellow in the article. The revisions are as follows:

QAP（Quadratic Assignment Procedure）；ERGM (Exponential Random Graph Model)

 

Comments 6: What does the number in parentheses refer to in Tables 4 & 5?

Response 6: Thank you for your valuable suggestion. We have revised the suggestion part in the Point 6,.and labelled yellow in the article, Standard errors are in parentheses.

Comments 7: Table 5 lacks the number of observations for Model 10 and the AIC, BIS, and Log Likelihood information for Model 9.

 

Response 7: Thank you for your valuable suggestion. The number of observations for model 10 has been written as 2754, model 10 is based on a robustness test under the MPLE method, the software does not give the AIC, BIS, and Log Likelihood information, so it is not listed.

 

Comments 8: Equation (6) requires a closing parenthesis.

Response 8: Thank you for your valuable suggestion. We have revised the suggestion part in the Point 8. The revisions are as follows:

Comments 9: To my knowledge, the “time-indexed random graph model” (TERGM) is not a common expression in the literature or is inconsistent with the abbreviations. I wonder if you mean the Temporal Exponential-Family Random Graph Models.

 

Response 9: Thank you for your valuable suggestion. We use TERGM (Temporal Exponential Random Graph Model)

 

Comments 10: What are the theoretical relationships between export probability and its determinants in Equation (6)? I failed to understand their expected sign and judge if the estimators were appropriate.

 

Response 10: Thank you for your valuable suggestion.

The TERGM model is based on a generative mechanism to explain network relationships. This generative mechanism assumes that each edge (or relationship) in the network is independently generated according to a certain probability distribution, which is influenced by other factors in the network (e.g., attributes of nodes, structural features of the network, etc.).

Other questions have also been revised accordingly.  Please refer to the yellow section of the article for details.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The paper entitled “Study on the evolution of SCO agricultural trade network pat- 2 tern and its influencing mechanism” is an interesting topic but some problems can be addressed before publication.

1.     Multiple ideas are discussed into one paragraph which making it difficult to understand. Thus, my suggestion is to break down the  arguments/discussion into easier parts.  Further lack of fluency, some paragraphs jump from one into another without a clear connection, which can create confusion for readers.

2.     The authors adopt the networking Methodology, which is good, but it is suggested to explain more about the network methodology. What is “the time index random graph model (TERGM)” explain it.  Further, analysis, could be expanded. Some terms are not fully explained.

3.     Add flowcharts that summarize the methodology to make the approach more understandable.

4.     Avoid the inconsistency in the use of terminology, particularly in the discussion of network theory and trade relations. Define the key terms early in the paper, especially for complex concepts like network density and reciprocity coefficient.

5.     The literature review is comprehensive but could benefit from a deeper critique of existing studies. Further highlight the study gaps in current research and how this study addresses these gaps.

6.     Add recent literature , it would  strengthen the literature review.

7.     Analysis section, could be enhanced by providing more detail of the findings.

8.     Though the authors summarize the findings well but still it could be more significant by discussing the potential impact of the research on policy or future research, it will provide a stronger ending of the paper.

Author Response

Comments 1:     Multiple ideas are discussed into one paragraph which making it difficult to understand. Thus, my suggestion is to break down the arguments/discussion into easier parts.  Further lack of fluency, some paragraphs jump from one into another without a clear connection, which can create confusion for readers.

Response 1: Thank you for your valuable suggestion.

Comments 2:     The authors adopt the networking Methodology, which is good, but it is suggested to explain more about the network methodology. What is “the time index random graph model (TERGM)” explain it.  Further, analysis, could be expanded. Some terms are not fully explained.

Response 2: Thank you for your valuable suggestion. We have revised the suggestion part in the Point 2, and labelled yellow in the article. The revisions are as follows:

TERGM (Temporal Exponential Random Graph Model) is an extension of ERGM (Exponential Random Graph Model) specifically designed for processing and analysing time-dependent network data. It combines theories from network science and social science and aims to explain relationships in networks and their evolution over time through a generative mechanism. The model is able to capture dynamic changes in the structure of a network, including the formation, disappearance and alteration of relationships between nodes, and how these changes are affected by other factors in the network (e.g., node attributes, network structural features, etc.)

Comments 3:    Add flowcharts that summarize the methodology to make the approach more understandable.

Response 3: Thank you for your valuable suggestion. The authors mainly used the TERGM model, which was finally tested for robustness, so we would like to keep the original write-up.

Comments 4:     Avoid the inconsistency in the use of terminology, particularly in the discussion of network theory and trade relations. Define the key terms early in the paper, especially for complex concepts like network density and reciprocity coefficient.

Response 4: Thank you for your valuable suggestion. We have revised the suggestion part in the Point 4.

Comments 5:     The literature review is comprehensive but could benefit from a deeper critique of existing studies. Further highlight the study gaps in current research and how this study addresses these gaps.

Response 5: Thank you for your valuable suggestion.

Comments 6:     Add recent literature, it would strengthen the literature review.

Response 6: Thank you for your valuable suggestion. We have referred to the main up-to-date literature relevant to this paper.

Comments 7:     Analysis section, could be enhanced by providing more detail of the findings.

Response 7: Thank you for your valuable suggestion.

Comments 8:     Though the authors summarize the findings well but still it could be more significant by discussing the potential impact of the research on policy or future research, it will provide a stronger ending of the paper.

Response 8: Thank you for your valuable suggestion.

Other questions have also been revised accordingly.  Please refer to the yellow section of the article for details.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

First of all, I must say that I find this study highly interesting, not only because of the geographical environment in which the data is analyzed, but also because of the primary productive environment and the conclusions relevant to any area in which agriculture is an intensive sector.

 

I like that the abstract not only talks about the objective and scope, but also includes a brief reference to the main output of the study, which makes the reader foresee what the development of the study will be without taking away interest from the details of the study.

 

The Introduction is quite well written, however I think it would gain power if it included some more references. I also think that reference should be made to the socio-political environment of the region, which studies such as DOI: 10.1016/j.techfore.2023.122742 (which I recommend citing) directly link to the digital development of countries and productive structures, and which can therefore have a broad impact on this area of ​​study. I also recommend including at the end of the section some brief commentary on the structure of the article in the following sections.

 

In the Literature Review, I see quite current and valuable references, and I also like the segmentation of the content into subsections. Perhaps it could be expanded a bit, but what the authors have described in this section is enough.

 

Section 3 of analysis is very interesting for the reader. Here the authors structure the object/scope of the study, the research method and the outputs of the study very efficiently. The Figures and Tables are of great value, and the graphic content also helps greatly to assess and understand the results of the study. Well done!

 

I recommend that section 4 of theoretical analysis be included (as a subsection) within section 3, since it leads to the mistake of thinking that it is a discussion of the results. Otherwise, the content is valuable and well described, nothing to object to.

 

The approach of the section "model construction and empirical analysis" corresponding to section 5, where the authors create the model and the hypotheses that they are going to develop, is very good. The selection of variables is correct and well justified, as are the mechanism variables. The construction of the model is explained very briefly, but it is enough to move on to the section on Empirical analysis using the MCMC MLE method. Here a brief explanation and reference to this method could be included. Section 5.3 could also be segmented since it is somewhat tedious to read due to having a lot of condensed text without reading pauses. The images and figures are valuable.

 

The Conclusions are well founded on the results obtained, and well segmented into sections. Well done! I also like the proposal for future studies, however I miss a brief reference to the limitations of the study.

 

The bibliographical references are current and sufficient, nothing to add except as I said in a previous comment, the lack of any reference to the degree of digital development based on the sociological components of the country under study.

 

In general, I liked the flow of the article!

Author Response

Comments 1: I like that the abstract not only talks about the objective and scope, but also includes a brief reference to the main output of the study, which makes the reader foresee what the development of the study will be without taking away interest from the details of the study.

Response 1: Thank you for your valuable suggestion.

Comments 2: The Introduction is quite well written, however I think it would gain power if it included some more references. I also think that reference should be made to the socio-political environment of the region, which studies such as DOI: 10.1016/j.techfore.2023.122742 (which I recommend citing) directly link to the digital development of countries and productive structures, and which can therefore have a broad impact on this area of ​​study. I also recommend including at the end of the section some brief commentary on the structure of the article in the following sections.

Response 2: Thank you for your valuable suggestion.

Comments 3: In the Literature Review, I see quite current and valuable references, and I also like the segmentation of the content into subsections. Perhaps it could be expanded a bit, but what the authors have described in this section is enough.

Response 3: Thank you for your valuable suggestion.

Comments 4: Section 3 of analysis is very interesting for the reader. Here the authors structure the object/scope of the study, the research method and the outputs of the study very efficiently. The Figures and Tables are of great value, and the graphic content also helps greatly to assess and understand the results of the study. Well done!

Response 4: Thank you for your valuable suggestion.

Comments 5: I recommend that section 4 of theoretical analysis be included (as a subsection) within section 3, since it leads to the mistake of thinking that it is a discussion of the results. Otherwise, the content is valuable and well described, nothing to object to.

Response 5: Thank you for your valuable suggestion.

Comments 6: The approach of the section "model construction and empirical analysis" corresponding to section 5, where the authors create the model and the hypotheses that they are going to develop, is very good. The selection of variables is correct and well justified, as are the mechanism variables. The construction of the model is explained very briefly, but it is enough to move on to the section on Empirical analysis using the MCMC MLE method. Here a brief explanation and reference to this method could be included. Section 5.3 could also be segmented since it is somewhat tedious to read due to having a lot of condensed text without reading pauses. The images and figures are valuable.

Response 6: Thank you for your valuable suggestion.

Comments 7: The Conclusions are well founded on the results obtained, and well segmented into sections. Well done! I also like the proposal for future studies, however I miss a brief reference to the limitations of the study.

Response 7: Thank you for your valuable suggestion.

Comments 8: The bibliographical references are current and sufficient, nothing to add except as I said in a previous comment, the lack of any reference to the degree of digital development based on the sociological components of the country under study.

Response 8: Thank you for your valuable suggestion.

Comments 9: In general, I liked the flow of the article!

Response 9: Thank you for your valuable suggestion.

Other questions have also been revised accordingly.  Please refer to the yellow section of the article for details.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Accept