Spatial Interaction Spillover Effect of Tourism Eco-Efficiency and Economic Development

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Thank you for giving me the possibility to review this paper. I have found the study presented in the paper interesting and undoubtedly relevant. This study uses the super SBM-DEA model and Malmquist model to measure the tourism eco-efficiency and proposes ways to enhance the tourism eco-efficiency of the Yangtze River Economic Belt. The paper will add to the literature on sustainable tourism, tourism planning, regional economic integration and tourism policy.

The abstract is informative. The keywords are in line with the terms used in the research. The Introduction section correctly puts the research topic in context and states the main research question.

The references are relevant and up to date. The literature review is appropriately structured, the theoretical framework and the hypothesis proposed based on the analysis conducted. The study area, data sources, and index system and variables are correctly detailed in the Methodology Section. A comprehensive analytical framework combines the 6 Super SBM-DEA model, the Malmquist index, and spatial econometric models, to analyze the spatial interplay between the tourism eco-efficiency and RGDP within the Yangtze River Economic Belt.

Based on the research results, the authors state the theoretical contribution of the study and provide strategic recommendations aimed at fostering regional collaborative advancement. May be it makes sense to add prospective avenue for future research.

Overall, I recommend this paper for publication.

Comments on the Quality of English Language

The English language is quite clear and understandable. Since I am not a native speaker, I have no additional recommendations.

Author Response

Thank you for giving me the possibility to review this paper. I have found the study presented in the paper interesting and undoubtedly relevant. This study uses the super SBM-DEA model and Malmquist model to measure the tourism eco-efficiency and proposes ways to enhance the tourism eco-efficiency of the Yangtze River Economic Belt. The paper will add to the literature on sustainable tourism, tourism planning, regional economic integration and tourism policy.

The abstract is informative. The keywords are in line with the terms used in the research. The Introduction section correctly puts the research topic in context and states the main research question.

The references are relevant and up to date. The literature review is appropriately structured, the theoretical framework and the hypothesis proposed based on the analysis conducted. The study area, data sources, and index system and variables are correctly detailed in the Methodology Section. A comprehensive analytical framework combines the 6 Super SBM-DEA model, the Malmquist index, and spatial econometric models, to analyze the spatial interplay between the tourism eco-efficiency and RGDP within the Yangtze River Economic Belt.

Based on the research results, the authors state the theoretical contribution of the study and provide strategic recommendations aimed at fostering regional collaborative advancement. May be it makes sense to add prospective avenue for future research.

Overall, I recommend this paper for publication.

Comments on the Quality of English Language

The English language is quite clear and understandable. Since I am not a native speaker, I have no additional recommendations.

 

Responses to reviewers

Reviewer #1：

Thank you very much for your time on our manuscript and the opportunity to revise the work.We are very grateful for the comments. We tried our best to improve the manuscript and made some changes to the manuscript.These changes will not influence the content and framework of the paper, and here we did not list the changes but marked in the revised paper.We appreciate for your warm work earnestly and hope that the correction will meet with approval.

In addition, all co-authors double revised the manuscript. It's our honor if our research can be published in Sustainability, thank you so much!

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

I suggest the following recommendation that are attached.

Comments for author File: Comments.pdf

Author Response

1.Comment:Figure 2 is not properly showing the data (as suggestion, use an image with better resolution).

1.Response:Thanks for your comment. We have increased the image resolution and annotated the source.

Specifically, revision as follows: Please see the revision MS (line225-229).

 

2.Comment:More references should be added, since a part of the research methodology (technique), is based on numerous research.

2.Response: Thanks for your comment.We have added the following references as well:

 

7.  Kong Y , Liu S .Spatio-temporal Evolution and Driving Factors of Carbon Dioxide Emissions from Energy Consumption in the Yellow River Basin，Journal of Physics: Conference Series, 2023, 2468(1).
8.  Sun, J.; Tang, D.; Kong, H.; Boamah, V. Impact of Industrial Structure Upgrading on Green Total Factor   Productivity in the Yangtze River Economic Belt. Int. J. Environ. Res. Public Health2022, 19, 3718.
9. Wu, C., Zhuo, L., Chen, Z., Tao, H., 2021. Spatial Spillover Effect and Influencing Factors of Information Flow in Urban Agglomerations—Case Study of China Based on Baidu Search Index.Sustainability,13(14), 8032.
10.  Tsaur S H , Lin Y C , Lin J H .Evaluating ecotourism sustainability from the integrated perspective of resource, community and tourism，tourism management, 2006, 27(4):640-653.
11. Nugroho P , Numata S .Resident support of community-based tourism development: Evidence from Gunung Ciremai National Park, Indonesia,Journal of Sustainable Tourism, 2020(5):1-16.
12. Bazargani R H Z , Kili H .Tourism competitiveness and tourism sector performance: Empirical insights from new data,Journal of Hospitality and Tourism Management, 2021, 46:73-82.
13. Cheng,L.,Zhou F Y.,Research on the Spatiotemporal Pattern and Growth Effect of Tourism Ecological Efficiency in the Yangtze River Economic Belt, Journal of Chongqing Technology and Business University (Social Sciences Edition), 1-16.(I(In Chinese)
14. Li Y F.,Li J.,Xu M., The interaction multiplier and spatial spillover effect between manufacturing employment and service employment, Finance and Trade Economy, 2017, 38 (04): 115-129.(In Chinese)

 

3.Comment:Please reconsider conclusions (as suggestion, be more consistency and indicate novelty, limits and future research).

3.Response: Thanks for your comment!We have added the innovation, limitations, and future research prospects of this article in the conclusion section.

Specifically, revision as follows: Please see the revision MS (line579-591).

 

4.Comment:conclusions are not supported by the results presented in the article or referenced in secondary literature (as suggestion, please refer to other tourism recently research).

4.Response: Thanks for your comment! We have added tourism related literature to support the conclusions and discussions of this manuscript.

For example, Cheng H et al. (2020) explored the spatial characteristics and influencing factors of China's tourism ecological efficiency, and believed that there is an interactive relationship and spatial spillover effect between tourism ecological efficiency and economic development, especially tourism economic development, and this spatial spillover effect is influenced by the differences in economic development levels among provinces. Regions with greater differences in economic development levels are more prone to spatial spillover. But it did not further explore the positive and negative effects of spatial overflow.

Guo et al. (2022) found that the main impact of economic development on tourism ecological efficiency lies in the allocation of resources such as human and capital. The research object of this article spans across the eastern, central, and western regions of China, with significant regional development disparities. Therefore, when neighboring regions experience rapid economic growth, it may attract the outflow of resources such as capital, technology, and talent. The loss of such resources may affect the sustainable development of the local tourism industry, thereby inhibiting the improvement of TEE.

Specifically, revision as follows: Please see the revision MS (lines 176-180、lines 574).

 

We hope the response to your questions will make you satisfied, we are very grateful for the valuable comments and suggestions. In addition, all co-authors double revised the manuscript. It's our honor if our research can be published in Sustainability, thank you so much!

 

Thanks and beat regards，

Qunli Tang and all authors

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Thank you for the opportunity to review the paper “Spatial Interaction Spillover Effect of Tourism Eco-Efficiency and Economic Development”. The main theme of the paper is in line with the aims and scope of the journal. Moreover, it is well-structured and meets the academic standards for a research paper. I believe it has the merits to be accepted, but I have some comments which I present below that the authors should take into account. My comments aim to improve the theoretical discussion of the paper and increase the validity and readability of the results section. I will provide my comments with reference to the text.

Section 2.2.1: I find it hard to follow the discussion in this section. First, how does the improved economic capacity of the local population enhance the TEE in the place they live? Is the local population considered tourists? Second, how does the development of clusters affect TEE? Some discussion with sufficient citations should be made here.

Line 131: I would avoid the reference to Say’s Law as its validity is questionable. Say posited that markets always eliminate supply surpluses, which is not always the case, especially for the tourism sector.

Section 2.2.3: I believe that not all possible effects of the neighboring areas on local tourism dynamics are discussed here. For instance, based on a gravity rationale, economic growth in neighboring regions may increase the flows to local destinations, thus increasing one part of TEE (desirable outputs). Second, increased TEE in a neighboring region may increase the TEE in the local area in at least two ways. First, by improving the fame and attractiveness of the wider area, especially for international tourism, and second, by positive spillovers in environmental conservation issues, especially in bordering destinations. Some additional discussion with appropriate references is also needed in this part.

Table 1: I have some comments on the selection and presentation of the variables. First, what does the 3A attraction level mean? Second, why did the authors select only 3-star accommodations? This would be fine if they only measured the revenue of these hotels as an output, but this is not the case here. The revenue used as an output indicator is produced by many other businesses not captured by the input indicator. Therefore, I suggest incorporating a measure of all hotels and establishments in their input. Third, travel agencies in the local areas mostly serve outbound tourism of locals. Moreover, I do not see how their operation affects the undesirable indicator. Maybe the authors should consider leaving this input out of their model. Finally, a table with the descriptive statistics of all variables should be included in the text to allow readers to understand their particular characteristics and the respective value intervals.

Line 286: The model should be numbered. Second, the notations are not fully explained in the paragraph that follows. A more detailed description of all notations should be given. For instance, what do the symbols d and u mean? The same is true for the Malmquist model. The notations should be fully explained. For instance, what is the notation for the desirable and undesirable outputs? In addition, the intervals of the scores and what they represent for the levels of efficiency should be given.

Line 360: How is the shortest economic distance measured? Why do the authors choose to weight distance with an economic measure which is also tested in the main model (RGDP with WRGDP)? I would recommend estimating the models again by using a distance measure (physical distance, borders, or something similar) that does not incorporate any economic variable and compare the results with the existing one.

Line 409: The term “total factor productivity (TFP)” was not discussed in the description of the model. How is this connected to TEE? Please keep in mind that not all readers of “Sustainability” are familiar with these terms. A brief explanation would be much appreciated.

Table 7: The authors run a panel 3SLS model with weights. There are some issues to consider here. First, did they check for time autocorrelation and heteroskedasticity? Second, regarding the spatial models, how did they test which model is more suitable for their study? Why did they omit the Durbin model? In ordinary regressions, the type of spatial autocorrelation is defined by maximum likelihood tests, and the weights matrix is incorporated into the model according to the results of the test. Therefore, some justification for their selection is essential. Third, the core values used for the TEE model could very easily be incorporated in the TEE DEA model. Visitor scale measured by the total number of tourists and especially scale of the tourism economy which is measured by the total tourism incomes (10,000 RMB) are heavily correlated with the output of the TEE model. I do not see how these two variables should be used as explanatory variables of eco-efficiency. The authors should consider the relevant literature to find some other variables to explain TEE. Fourth, the vast literature on DEA explanatory frameworks employs truncated regressions because of the nature and statistical properties of DEA scores (see Simar and Wilson, 2007). The model here is a linear one, therefore it is not suitable for this kind of data. Therefore, a truncated specification should be employed to substitute or supplement the existing one.

Minor comments

Lines 33-34: “By assessing TEE, it is possible to achieve economic growth while minimizing environmental damage.” I would replace the word “assessing” with something like “improving” or “increasing” because I do not see how the assessment of TEE leads to economic growth.

Line 73: Replace "various" with "Various."

Line 78: Replace "Efthymios" with "Tsionas."

Line 106: Replace “focus” with something like “attention is paid.”

Figure 2: A source is needed.

Line 292: “inputs, desired outputs, and desired outputs.” Please check this sentence for errors or duplications.

References

Simar, L., & Wilson, P. W. (2007). Estimation and inference in two-stage, semi-parametric models of production processes. Journal of econometrics, 136(1), 31-64.

Comments on the Quality of English Language

There are soem grammar issues. Some examples are highlighted in my comments.

Author Response

Responses to reviewers(original comments by reviewers are in blue color)

Reviewer #3：

1.Comment: Section 2.2.1: I find it hard to follow the discussion in this section. First, how does the improved economic capacity of the local population enhance the TEE in the place they live? Is the local population considered tourists? Second, how does the development of clusters affect TEE? Some discussion with sufficient citations should be made here.

1.Response: Thank you very much for your time on our manuscript and the opportunity to revise the work.The lack of sufficient clarification in Section 2.2.1 is indeed our oversight.

In response to the question “how does the improved economic capacity of the local population enhance the TEE in the place they live?”，it is important to note that TEE is a quantitative indicator used to measure the degree of sustainable development and harmony between humans and the environment in a region’s tourism industry. Higher TEE indicate better tourism development and a more harmonious relationship between humans and the environment in the region.The study area is located in an international golden tourism belt, where the tourism industry is thriving. Scholars have found that the development of tourism increases residents’ income and improves their quality of life. Based on economic benefit perception, community members believe that tourism can benefit the local economy, leading to increased participation in tourism planning, decision-making, and activities. Therefore, the enhancement of residents’ economic capabilities motivates greater support for the development of the local tourism industry, with community members participating as tourists or tourism professionals, contributing to the industry’s positive development and consequently enhancing TEE. To support this discussion, we have added the following reference:

[1] Nugroho P , Numata S .Resident support of community-based tourism development: Evidence from Gunung Ciremai National Park, Indonesia[J].Journal of Sustainable Tourism, 2020(5):1-16.DOI:10.1080/09669582.2020.1755675.

Regarding the question “how does the development of clusters affect TEE?”, the issue of sustainable development of the industry also pertains to how the industry maintains long-term competitiveness. The development of tourism industry clusters enhances the competitiveness of the local tourism industry, thus benefiting the sustainable development of the industry. In terms of indicator measurement, TEE is improved. There are studies that specifically explore the relationship between tourism industry clusters and their competitiveness, emphasizing the positive impact of tourism competitiveness on the development of the tourism industry, that is, regions with tourism industry clusters exhibit stronger tourism competitiveness and promote the good development of the industry. To support this discussion, we have added the following reference:

[2] Bazargani R H Z , Kili H .Tourism competitiveness and tourism sector performance: Empirical insights from new data[J].Journal of Hospitality and Tourism Management, 2021, 46:73-82.DOI:10.1016/j.jhtm.2020.11.011.

According to your valuable suggestions, we finally made an explanation, and hope to get your satisfaction!

Specifically, revision as follows: Please see the revision MS (lines 114-146).

 

2.Comment: Line 131: I would avoid the reference to Say’s Law as its validity is questionable. Say posited that markets always eliminate supply surpluses, which is not always the case, especially for the tourism sector.

2.Response: Thanks for your comment！This section primarily aims to discuss how the market development of the tourism industry will, to a certain extent, increase the number of tourism practitioners, thereby addressing some employment issues. However, upon careful consideration of your suggestion, we agree that the application of Say’s Law in this section is indeed insufficient. Consequently, we have removed the reference to Say’s Law from this part and instead relevant literature as evidence to support our argument.

Specifically, revision as follows: Please see the revision MS (lines 155-158).

 

3.Comment: Section 2.2.3: I believe that not all possible effects of the neighboring areas on local tourism dynamics are discussed here. For instance, based on a gravity rationale, economic growth in neighboring regions may increase the flows to local destinations, thus increasing one part of TEE (desirable outputs). Second, increased TEE in a neighboring region may increase the TEE in the local area in at least two ways. First, by improving the fame and attractiveness of the wider area, especially for international tourism, and second, by positive spillovers in environmental conservation issues, especially in bordering destinations. Some additional discussion with appropriate references is also needed in this part.

3.Response: Thanks for your comment.Your considerations and suggestions are reasonable! The spatial relationship between TEE and RGDP is quite complex, and may even become more intricate due to differences in regional economic development levels and resource endowments. However, after conducting in-depth research on the relationship between two, we found that most of the literature has sorted out the one-way relationship between the two. Even if some studies have found spatial spillover effects, further research is needed to investigate the positive and negative effects of spatial spillover effects.

For example, Cheng et al. (2020) explored the spatial characteristics and influencing factors of China's tourism ecological efficiency, and believed that there is an interactive relationship and spatial spillover effect between TEE and economic development, especially tourism economic development, and this spatial spillover effect is influenced by the differences in economic development levels among provinces. Regions with greater differences in economic development levels are more prone to spatial spillover. But it did not further explore the positive and negative effects of spatial overflow. Guo et al. (2022) found that the main impact of economic development on TEE lies in the allocation of resources such as human and capital. The research object of this paper spans across the eastern, central, and western regions of China, with significant regional development disparities.

Therefore, when neighboring regions experience rapid economic growth, it may attract the outflow of resources such as capital, technology, and talent. The loss of such resources may affect the sustainable development of the local tourism industry, thereby inhibiting the improvement of TEE. Although this article proposes the hypothesis of siphoning, the subsequent empirical test results significantly validate this hypothesis. However, the spatial relationship between TEE and RGDP is still complex and requires further in-depth research in the future.We have placed the exploration of this part in the future research outlook.

According to your valuable suggestions, we finally made an explanation, and hope to get your understanding.

Specifically, revision as follows: Please see the revision MS (lines 176-180,lines 579-591).

4.Comment: Table 1: I have some comments on the selection and presentation of the variables. First, what does the 3A attraction level mean? Second, why did the authors select only 3-star accommodations? This would be fine if they only measured the revenue of these hotels as an output, but this is not the case here. The revenue used as an output indicator is produced by many other businesses not captured by the input indicator. Therefore, I suggest incorporating a measure of all hotels and establishments in their input. Third, travel agencies in the local areas mostly serve outbound tourism of locals. Moreover, I do not see how their operation affects the undesirable indicator. Maybe the authors should consider leaving this input out of their model. Finally, a table with the descriptive statistics of all variables should be included in the text to allow readers to understand their particular characteristics and the respective value intervals.

4.Response: Thanks for your comment！First, regarding the question of "what does the 3A attraction level mean?", the index belongs to tourism resource investment and represents the number of 3A and above scenic spots. This is a statistical indicator published in the China Yearbook of Cultural Relics and Tourism Statistics .

Second, the number of three-star hotels to some extent reflects the tourist reception capacity of the region (Guo et al., 2022), which is a capital investment indicator for calculating tourism ecological efficiency. Cheng et al. (2020) also used the number of three-star hotels as an input in their study. Due to the availability of tourism hotel data, it is difficult to collect all the data of star rated hotels during the research period. We hope for your understanding.

Third, travel agencies are important investment resources for the development of regional tourism industry and will occupy local tourism capital investment. Therefore, they are included as indicators of tourism capital investment (Cheng et al.,2023）。

Finally, in order to facilitate readers to quickly find the corresponding variables in the corresponding sections, we have listed the measurement indicators of tourism ecological efficiency and the empirical indicators in the corresponding sections of the article. Thank you for your feedback. For the convenience of readers, we have included a summary table of descriptive statistics for all variables involved in this article in Appendix A.

Specifically, revision as follows: Please see the revision MS (lines 644-646).

5.Comment: Line 286: The model should be numbered. Second, the notations are not fully explained in the paragraph that follows. A more detailed description of all notations should be given. For instance, what do the symbols d and u mean? The same is true for the Malmquist model. The notations should be fully explained. For instance, what is the notation for the desirable and undesirable outputs? In addition, the intervals of the scores and what they represent for the levels of efficiency should be given.

5.Response: Thanks for your comment！At line 286, the main data source was not found, and no formula was found. But according to your opinion, all models in the manuscript have been numbered. We also provided additional explanations on the specific meanings of the symbols in the model. For example, in the SBM-DEA model, “d and u represent expected output and unexpected output, respectively, and y^d and y^u represent elements in the expected output matrix and unexpected output matrix, respectively...”

In the Malmquist model, due to the requirement that various outputs change in the same direction, it is not suitable to treat unexpected outputs as negative outputs. This article draws on the method of Cheng (2021) and treats unexpected outputs as input indicators to directly use the Malmquist model.

Since Malmquist reflects efficiency changes from a dynamic perspective, generally speaking, efficiency values are divided into changes in efficiency (rising, unchanged, falling) by comparing them with 1. When the efficiency value is greater than 1, it indicates that the efficiency change in the current period has increased compared to the previous period. If the efficiency value is less than 1, it indicates that the efficiency in the current period has decreased compared to the previous period. If the efficiency value is equal to 1, it indicates that the efficiency has not changed.

According to your valuable suggestions, we finally made an explanation, and hope to get your understanding.

Specifically, revision as follows: Please see the revision MS (lines321-322,lines331-340).

6.Comment: Line 360: How is the shortest economic distance measured? Why do the authors choose to weight distance with an economic measure which is also tested in the main model (RGDP with WRGDP)? I would recommend estimating the models again by using a distance measure (physical distance, borders, or something similar) that does not incorporate any economic variable and compare the results with the existing one.

6.Response: Thanks for your comment！Sorry for the misunderstanding caused by our incorrect expression of d_ij. d_ij represents the shortest distance between the economies of each province and city, and this article calculates it using the Euclidean distance of the provincial capital city. We have made modifications to the original text.

According to the theory of new economic geography, the interactive status between two regions is unequal at different stages and highly correlated with the level of economic development. For example, the spillover effect of Shanghai on Anhui is not exactly the same as the spillover effect of Anhui on Shanghai, and is related to the economic development level of the two places. But traditional geographic spatial matrices assume that the interaction between two places is the same, which obviously leads to errors. Therefore, in order to more accurately reflect the interrelationships between regions, we drew on the weighted spatial economic geographic distance weight matrix proposed by Li et al. (2017). We use per capita GDP to represent the economic distance indicator in the matrix.

Specifically, revision as follows: Please see the revision MS (line380-line384, line393-line395).

7.Comment: Line 409: The term “total factor productivity (TFP)” was not discussed in the description of the model. How is this connected to TEE? Please keep in mind that not all readers of “Sustainability” are familiar with these terms. A brief explanation would be much appreciated.

7.Response: Thanks for your comment！Total factor productivity(TFP) is an indicator that measures the production efficiency of a region in all input factors (such as labor, capital, technology, etc.). It reflects the comprehensive impact of factors such as technological progress, scale efficiency, and production organization on output. The growth of TFP is often seen as an important source of economic growth and technological progress, and it is a broader concept that encompasses all input factors that affect output. However, in this article, it can be considered that the TFP value calculated by the application model represents the tourism ecological efficiency value. However, since tourism ecological efficiency is a variable in this article, two methods were used to calculate TEE (SBM-DEA Model, Malmquist Model).Therefore, TEE was not used here. But for the convenience of readers' understanding, we have made modifications based on your suggestions.

Specifically, revision as follows: Please see the revision MS (line445,line540).

8.Comment: Table 7: The authors run a panel 3SLS model with weights. There are some issues to consider here. First, did they check for time auto-correlation and heteroskedasticity? Second, regarding the spatial models, how did they test which model is more suitable for their study? Why did they omit the Durbin model? In ordinary regressions, the type of spatial auto-correlation is defined by maximum likelihood tests, and the weights matrix is incorporated into the model according to the results of the test. Therefore, some justification for their selection is essential. Third, the core values used for the TEE model could very easily be incorporated in the TEE DEA model. Visitor scale measured by the total number of tourists and especially scale of the tourism economy which is measured by the total tourism incomes (10,000 RMB) are heavily correlated with the output of the TEE model. I do not see how these two variables should be used as explanatory variables of eco-efficiency. The authors should consider the relevant literature to find some other variables to explain TEE. Fourth, the vast literature on DEA explanatory frameworks employs truncated regressions because of the nature and statistical properties of DEA scores (see Simar and Wilson, 2007). The model here is a linear one, therefore it is not suitable for this kind of data. Therefore, a truncated specification should be employed to substitute or supplement the existing one.

8.Response: Thanks for your comment！First, the Generalized Spatial Three Stage Least Squares (GS3SLS) method mainly focuses on spatial autocorrelation problems. It combines the characteristics of classical Three Stage Least Squares (3SLS) and spatial econometrics to address spatial autocorrelation and heterogeneity issues. The core of GS3SLS does not directly include checks for time autocorrelation and heteroscedasticity.

Second, as the purpose of this paper is to detect the bidirectional spatial interaction spillover relationship between two variables. Traditional single equation models tend to overlook the bidirectional interaction between TEE and RGDP. For example, spatial Durbin models are one-way equations that cannot test the interaction between core variables and may overlook endogeneity issues between core variables. The ordinary simultaneous equation model does not consider the spatial spillover effect between TEE and RGDP. Therefore, in order to verify the bidirectional spatial spillover effect between the two, this paper constructs a spatial simultaneous equation system model of TEE and RGDP, which can test both the interaction relationship of variables and the spatial spillover effect.

Third, the main purpose of this study is to examine the spatial relationship between TEE and RGDP, but other important influencing factors need to be controlled for. Therefore, tourist scale and tourism revenue are included as control variables in the model, rather than explanatory variables. TEE and GRDP in this article are explanatory and dependent variables for each other.

Finally, this paper investigates the estimation and inference problems of two-stage semi parametric models in the production process, using DEA to measure technical efficiency, and using truncated regression in the second stage to estimate the impact of environmental variables on efficiency.

Data Envelopment Analysis (DEA) is a method based on relative efficiency evaluation, used to evaluate the relative efficiency of decision units with multiple inputs and outputs. In 2002, Japanese scholar Tone proposed a non radial, relaxation based super efficiency SBM-DEA model to address the issue of traditional DEA models being unable to handle unexpected outputs such as pollutant emissions. Truncated Regression and Tone's SBM-DEA model based on slack measures differ in methodology and application areas, but there are some conceptual connections between them. Truncated regression estimates model parameters by maximizing the likelihood function, which can be seen as an optimization process. The SBM-DEA model maximizes the efficiency score of decision units through linear programming, which is also an optimization process. The data type of this article is also suitable for applying the SBM-DEA model, and existing literature on tourism ecological efficiency has also adopted the SBM-DEA model method.

According to your valuable suggestions, we finally made an explanation, Section 3.4.3 of our manuscript provides a description of this, and hope to get your understanding and satisfaction.

Specifically, revision as follows: Please see the revision MS (line351,lines347-359).

 

Minor Comments

a)Comment: Lines 33-34: “By assessing TEE, it is possible to achieve economic growth while minimizing environmental damage.” I would replace the word “assessing” with something like “improving” or “increasing” because I do not see how the assessment of TEE leads to economic growth.

a)Response:Thanks for your comment.

Specifically, revision as follows: Please see the revision MS (line39).

 

b)Comment: Line 73: Replace "various" with "Various."

b)Response:Thanks for your comment.

Specifically, revision as follows: Please see the revision MS (line77).

 

c)Comment: Line 78: Replace "Efthymios" with "Tsionas."

c)Response:Thanks for your comment.

Specifically, revision as follows: Please see the revision MS (line82).

 

d)Comment: Line 106: Replace “focus” with something like “attention is paid.”

d)Response:Thanks for your comment.

Specifically, revision as follows: Please see the revision MS (line108).

 

e)Comment: Figure 2: A source is needed.

e)Response:Thanks for your comment. We have increased the image resolution and annotated the source.

Specifically, revision as follows: Please see the revision MS (line225-229).

 

f)Comment: Line 292: “inputs, desired outputs, and desired outputs.” Please check this sentence for errors or duplications.

f)Response:Thanks for your comment.

Specifically, revision as follows: Please see the revision MS (line319).

 

g)Comment: Comments on the Quality of English Language

There are soem grammar issues. Some examples are highlighted in my comments.

g)Response:Thanks for your suggestion!We have tried our best to polish the language in the revised manuscript.

 

We hope the response to your questions will make you satisfied, we are very grateful for the valuable comments and suggestions. In addition, all co-authors double revised the manuscript. It's our honor if our research can be published in Sustainability, thank you so much!

 

Thank you and beat regards.

Qunli Tang and all authors

Author Response File: Author Response.pdf