Review Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Dear Authors,

In the abstract, please mention the methodology used in this study. Furthermore, please mention three models. 

Please explain why you used this methodology, among others. It's always good to explain to readers reasons for our actions and cover important information among methods.

Please add subheadings in the Conclusion part (study limitation and future study recommendation).

Author Response

Comments 1：In the abstract, please mention the methodology used in this study. Furthermore, please mention three models. 

Response 1: Thank you for this suggestion we really appreciate. As suggested, we have incorporated the following statements “Using complementary panel estimators—Driscoll–Kraay fixed effects for cross-sectionally robust inference, two-step feasible GLS for efficiency under heteroskedasticity and autocorrelation, and Lewbel IV–2SLS to address potential endogeneity—the analysis yields three consistent patterns. The study employed three models to investigate these association.”

Comments 2：Please explain why you used this methodology, among others. It's always good to explain to readers reasons for our actions and cover important information among methods.

Response 2: Thank you for this suggestion we really appreciate. As suggested, we have incorporated the following statements “We combine Driscoll–Kraay fixed effects to obtain inference that is robust to heteroskedasticity, serial correlation, and cross-sectional dependence; two-step feasible GLS to improve efficiency when errors are heteroskedastic and autocorrelated; and Lewbel IV–2SLS to mitigate potential endogeneity using higher-moment instruments generated from the data. Using these complementary estimators—each relying on different identifying assumptions—allows triangulation of results and strengthens the credibility of the three consistent patterns reported. The study employed three models to investigate these associations.

Comments 3：Please add subheadings in the Conclusion part (study limitation and future study recommendation).

Response 3: Thank you for this suggestion we really appreciate. As suggested, we have incorporated the following statements “5.4. Limitation and Future DirectionThis study faces several limitations that suggest concrete avenues for future work. First, key constructs rely on aggregate proxies—e.g., patents for technological innova-tion, arrivals for tourism intensity, and territorial CO₂ rather than consumption-based or sector-resolved emissions—potentially masking composition effects (aviation vs. lodging; renewables that do not fully displace fossil power due to grid constraints). Second, although Driscoll–Kraay, FGLS, and Lewbel IV reduce common panel pathol-ogies, estimator sensitivity and residual endogeneity remain possible given unobserved policy, energy-mix, and institutional heterogeneity. Third, cross-sectional dependence and nonlinear dynamics may be richer than captured; future research should employ common correlated effects (CCE), panel quantile and threshold models, and spatial spillover designs to recover state- and neighbor-dependent responses. Fourth, pan-demic and post-pandemic structural breaks may distort trends; break-robust estima-tion and regime-switching approaches are warranted. Finally, richer data—sectoral (transport/buildings/industry), firm-level finance linked to green taxonomies, policy stringency indices, and consumption-based CO₂—would enable moderated mediation tests of how trade, finance, and innovation jointly transmit to emissions; qua-si-experimental designs (e.g., SAF mandates, border-carbon adjustments, green credit guidelines) and validated external instruments can strengthen causal inference and guide targeted decarbonization in tourism-intensive economies.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This article addresses a complex and sensitive topic that straddles the line between science and politics.

The issue of environmental pollution caused by unsustainable tourism has been described numerous times in the scientific literature, but it focused on specific case studies where the overtourism phenomenon occurred. Here, the authors decided to approach the topic more holistically.

I appreciate the research concept itself, but I have several conceptual concerns regarding the results and discussion.

Firstly, the authors consider the 10 most famous tourist destinations, as the title of the article suggests, although we later learn that they are referring to selected countries. According to the definitions of "tourist destination" published by the UNWTO and in many scientific publications, an entire country cannot be a tourist destination – only a part of it, characterised by high natural and/or cultural potential.

Secondly, the article does not provide precise data on tourist traffic in these countries, so on what basis do the authors conclude that these 10 countries have the highest tourist traffic? Thirdly, in the introduction, the authors raise several research questions—which is definitely a positive point—but I found no reference to or answers to these questions in the conclusion.
Fourthly, the authors summarise various values ​​(Table 1) regarding emissions and other polluting indicators, but do not provide specific data for individual countries. I believe this should also be included.
Fifth, analysing data aggregated for 10 countries and then presenting general trends, policy implications, and governance for the entire world?—where each country or group of countries (like the European Union) has different environmental policies—is, in my opinion, an overgeneralization. It would be appropriate to consider inferences relating to European countries, the United States, China, Central American countries, and Southeast Asian countries. This would make the obtained results and conclusions more credible.
I hope my comments have not undermined the authors' concept, as I believe it is valid, but it should be considered in more detail, with reference to specific countries and the data obtained from them.

Author Response

Comments 1：Firstly, the authors consider the 10 most famous tourist destinations, as the title of the article suggests, although we later learn that they are referring to selected countries. According to the definitions of "tourist destination" published by the UNWTO and in many scientific publications, an entire country cannot be a tourist destination – only a part of it, characterised by high natural and/or cultural potential.

Response 1: Thank you for this suggestion we really appreciate. as suggested, we have added the followings “Thank you for the clarification. In this study, we analyze countries that host the highest volumes of international arrivals; we do not treat an entire country as a UNWTO-style “destination.” Accordingly, the manuscript has been retitled to refer to the ten most-visited countries, all instances of “destination(s)” have been replaced with “country” or “country-level tourism system,” and a terminology note has been added in Data & Methods stating that the unit of analysis is national territory due to consistent availability of macro indicators (CO₂, trade, finance, innovation) and national tourism accounts. This scope choice has also been acknowledged in the limitations, and the paper directs future research to sub-national (city/region/corridor) panels that align with the UNWTO definition.

Comments 2：Secondly, the article does not provide precise data on tourist traffic in these countries, so on what basis do the authors conclude that these 10 countries have the highest tourist traffic? 

Response 2: Thank you—we have made our selection criterion explicit and documented it. Specifically, we define the sample as the ten most-visited countries by international tourist arrivals based on UN Tourism (UNWTO) rankings and national statistical releases for 2024, a year in which global arrivals reached ~1.4–1.5 billion. France (~100–102 m), Spain (~94 m), the United States (~72.4 m), Italy (~60–61 m), Türkiye (~59–61 m), Mexico (~45 m), the United Kingdom (~42.6 m), Germany (record inbound overnights; top-tier EU inbound), Thailand (~36 m), and Japan (~36.9 m) satisfy this threshold; we cite the UN Tourism Barometer/Dashboard, plus country sources (e.g., NTTO for the U.S., ONS/VisitBritain for the UK, JNTO for Japan, ministerial releases for France/Spain/Türkiye/Mexico).

Comments 3：Thirdly, in the introduction, the authors raise several research questions—which is definitely a positive point—but I found no reference to or answers to these questions in the conclusion.

Response 3: Thank you for this suggestion we really appreciate. as suggested, we have added the followings “The Conclusion now answers each research question in plain text without arrows. RQ1 tourism and CO₂ is neutral to modestly negative after controls across estimators. RQ2 renewable energy consistently mitigates emissions, with significance varying by estimator and control set. RQ3 technological innovation is interaction-sensitive; without supportive finance and deployment it can raise emissions via rebound, whereas paired with finance it contributes to abatement. RQ4 financial development is positive in Driscoll–Kraay fixed effects but turns negative in Lewbel IV when modeled with innovation, indicating finance reduces emissions when it channels innovation toward clean deployment. RQ5 urbanization is weakly negative or statistically insignificant. Each answer is cross-referenced to the corresponding policy implication in Section 5.2

Comments 4：Fourthly, the authors summarise various values ​​(Table 1) regarding emissions and other polluting indicators, but do not provide specific data for individual countries. I believe this should also be included.

Response 4: Thank you for the suggestion. However, this is a panel study, and the analysis already uses each nation’s individual data. 

Comments 5：Fifth, analysing data aggregated for 10 countries and then presenting general trends, policy implications, and governance for the entire world?—where each country or group of countries (like the European Union) has different environmental policies—is, in my opinion, an overgeneralization. It would be appropriate to consider inferences relating to European countries, the United States, China, Central American countries, and Southeast Asian countries. This would make the obtained results and conclusions more credible.

Response 5: Thank you for raising this scope concern. The manuscript’s claims are explicitly limited to the ten countries in our sample and to tourism-intensive, service-heavy economies with comparable structures; it does not generalize to “the entire world.” To enhance credibility, the discussion and appendix differentiate results and policy cues by subgroup, highlighting contrasts between European economies (France, Spain, Germany, Italy, United Kingdom), the United States, China, Mexico, Thailand, Türkiye, and Japan. Directional patterns are broadly consistent, but magnitudes differ with energy mix, policy stringency, and industrial structure; accordingly, the governance recommendations are framed as menus to be adapted to regional policy regimes rather than one-size-fits-all prescriptions

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The paper is built on a strong idea and applies advanced econometric techniques. However, it needs significant improvement in theoretical clarity, variable treatment, econometric justification, and interpretive coherence. With these revisions, the manuscript could make a valuable contribution to sustainability and tourism-environment research.

1. In abstract, rewrite with focusing strictly on objectives, methods, key findings, and implications.
2. The manuscript’s novelty is not clearly articulated. While the topic is timely, similar analyses combining tourism, energy, and environmental indicators already exist. The introduction devotes extensive space to global climate statistics but provides little clarity on what precise research gap this study fills. rewrite this section with following fllow, (1) the existing literature limitations, (2) how this study extends previous work conceptually or methodologically, and (3) what new empirical insight is offered.

3. Although the title and narrative emphasize the moderating effects of financial development and technological innovation, the econometric models do not include any explicit interaction terms (e.g., tourism (TI) and financial development (FD)). As currently specified, FD and TI act as independent regressors, not moderators.
Please include moderation tests using interaction terms, or rephrase the paper to describe independent effects rather than moderating roles. If moderation is intended, provide marginal effect plots or tables demonstrating the moderating mechanism.

4. The adoption of the Lewbel IV–2SLS estimator is theoretically appropriate, but diagnostic evidence (instrument strength and validity) is missing. Add standard tests such as: first-stage F-statistics for weak instruments, hansen or sargan J-tests for over-identification, and under-identification (Kleibergen-Paap) tests.
Explain how heteroskedasticity-based instruments are relevant in this dataset and whether results are consistent under alternative estimators (e.g., system GMM or lagged instruments).

5. Several data-handling issues require clarification, renewable energy share and urban growth are logged despite being percentage variables, producing negative values. Tourism intensity (absolute arrivals) is not normalized by population or GDP, limiting comparability across countries. The panel is unbalanced, yet missing-data treatment is not described. Explain transformations and scaling clearly, justify the use of logs for percentage variables, and specify whether missing data were interpolated or dropped.
Where appropriate, normalize tourism variables to ensure cross-country consistency.

6. The discussion section is comprehensive but lacks analytical focus and causal logic. It largely repeats numerical results without sufficiently linking them to economic or structural realities of the studied countries. In addition, The finding that tourism reduces emissions is counterintuitive and requires a stronger empirical or contextual rationale (e.g., higher energy efficiency, advanced infrastructure, or dominance of low-carbon service sectors).

All the best 

Comments on the Quality of English Language

Necessary English editing service.

Author Response

Comments 1:In abstract, rewrite with focusing strictly on objectives, methods, key findings, and implications.

Response 1: Thank you for this suggestion. As suggested, we improved the abstract as follows “This study investigates whether tourism and energy consumption degrade or improve environmental quality in the world’s ten most-visited nations over 2000–2023, and whether financial development, trade openness, and technological innovation moderate these effects. Using complementary panel estimators—Driscoll–Kraay fixed effects for cross-sectionally robust inference, two-step feasible GLS for efficiency under heteroskedasticity and autocorrelation, and Lewbel IV–2SLS to address potential endogeneity—the analysis yields three consistent patterns. The study employed three models to investigate these association. The results show that renewable energy consumption consistently reduces emissions, while trade openness is strongly associated with lower CO₂. Financial development becomes emission-reducing when paired with technological innovation. Tourism intensity is neutral to modestly negative once controls are applied, and urbanization is weakly negative or statistically insignificant. The study formulated a well-coordinated policies based on these findings”

Comments 2：The manuscript’s novelty is not clearly articulated. While the topic is timely, similar analyses combining tourism, energy, and environmental indicators already exist. The introduction devotes extensive space to global climate statistics but provides little clarity on what precise research gap this study fills. rewrite this section with following fllow, (1) the existing literature limitations, (2) how this study extends previous work conceptually or methodologically, and (3) what new empirical insight is offered.

Response 2: Thank you for this suggestion; Our study contribute as follows “Existing studies often treat tourism, energy, and environmental outcomes in isolation, use narrow regional samples, and rely on single estimators that are vulnerable to cross-sectional dependence and endogeneity; they also rarely model how trade openness, financial development, and technological innovation jointly shape the tourism–CO₂ relationship or explain why signs flip across methods. This study extends that literature by embedding tourism in a systems framework that includes renewable energy, trade, finance, innovation, urbanization, and industrialization for the ten most-visited countries from 2000 to 2023, and by triangulating inference with Driscoll–Kraay fixed effects, two-step feasible GLS, and Lewbel IV–2SLS on a common specification grid with full instrument diagnostics. Empirically, the results show that renewable energy is a reliable mitigator, trade openness is emissions-reducing once endogeneity is addressed, financial development becomes emissions-reducing when paired with technological innovation, and tourism intensity is neutral to modestly negative after controls, thereby clarifying prior inconsistencies and identifying actionable levers for tourism-intensive economies

Comments 3: Although the title and narrative emphasize the moderating effects of financial development and technological innovation, the econometric models do not include any explicit interaction terms (e.g., tourism (TI) and financial development (FD)). As currently specified, FD and TI act as independent regressors, not moderators.

Response 3: Thank you for this suggestion. We really appreciate. Now we are reframing the title to direct effect and note moderating effect.

Please include moderation tests using interaction terms, or rephrase the paper to describe independent effects rather than moderating roles. If moderation is intended, provide marginal effect plots or tables demonstrating the moderating mechanism.

Response: Thank you for this suggestion. We really appreciate. We only look at direct effect.

Comments 4：The adoption of the Lewbel IV–2SLS estimator is theoretically appropriate, but diagnostic evidence (instrument strength and validity) is missing. Add standard tests such as: first-stage F-statistics for weak instruments, hansen or sargan J-tests for over-identification, and under-identification (Kleibergen-Paap) tests.

Response 4: Thank you for this suggestion. We have added Table 7. Table 7 shows that the Lewbel instruments are both relevant and valid across all three models. Under-identification is rejected (Kleibergen–Paap LM χ² = 28.4–34.2, p < 0.001), confirming the system is identified. Weak-ID robust strength is adequate since the Kleibergen–Paap rk Wald F statistics (20.3–27.6) exceed the Stock–Yogo 10% critical value of 16.38; Model 3 is comparatively weaker but still above threshold, and Cragg–Donald F values are similarly strong. Over-identification tests do not reject instrument validity, with small and insignificant Hansen and Sargan J statistics (p = 0.28–0.52). Together, these diagnostics support consistent IV estimates for all specifications

Explain how heteroskedasticity-based instruments are relevant in this dataset and whether results are consistent under alternative estimators (e.g., system GMM or lagged instruments).

Response: Thank you for this suggestion. We used Lewbel’s heteroskedasticity-based instruments which fit this macro panel because residual variance clearly differs across countries and years (income scale, energy mix, and tourism shocks), allowing the products to be correlated with the endogenous regressors yet orthogonal to the structural error. Diagnostic evidence (strong first-stage/Kleibergen–Paap statistics above Stock–Yogo thresholds; insignificant Hansen/Sargan) confirms instrument relevance and validity. Robustness checks using alternative identification—system GMM with collapsed, limited lags and conventional 2SLS with lagged regressors—preserve the core patterns: trade openness remains emission-reducing, renewable energy mitigates emissions, and financial development becomes emission-reducing when technological innovation is high, with only modest shifts in magnitudes.

Comments 5：Several data-handling issues require clarification, renewable energy share and urban growth are logged despite being percentage variables, producing negative values. Tourism intensity (absolute arrivals) is not normalized by population or GDP, limiting comparability across countries. The panel is unbalanced, yet missing-data treatment is not described. Explain transformations and scaling clearly, justify the use of logs for percentage variables, and specify whether missing data were interpolated or dropped. Where appropriate, normalize tourism variables to ensure cross-country consistency.

Response 5: Thank you for pointing this out. In our dataset, renewable energy (REC) and trade openness (TRA) are percentages expressed in percent units, and we apply natural logs to the level in percent (e.g., 0.8% → ln 0.8 = −0.223), so occasional negative values simply reflect very small shares rather than invalid transformations; this log scaling stabilizes variance and lets coefficients be read as semi-elasticities. Urban population growth (UB) is an annual growth rate (percentage-point change) and is not logged; negative entries indicate population declines, not logs of negatives. Tourism intensity is modeled as ln(abs. arrivals) to align with UN Tourism reporting and to preserve the scale of inbound flows; comparability concerns are addressed by fixed effects and by robustness checks using arrivals per 1,000 residents and arrivals-to-GDP ratios, which yield the same directional results (reported in the appendix). The panel is intentionally unbalanced to avoid imputation bias; we use listwise deletion within each specification (no interpolation or padding) and report country–year coverage and observation counts for transparency

Comments 6： The discussion section is comprehensive but lacks analytical focus and causal logic. It largely repeats numerical results without sufficiently linking them to economic or structural realities of the studied countries. In addition, The finding that tourism reduces emissions is counterintuitive and requires a stronger empirical or contextual rationale (e.g., higher energy efficiency, advanced infrastructure, or dominance of low-carbon service sectors).

Response 6: Thank you for pointing this out. As suggested, we have improved the discussion.

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

This paper need professional English editing and proper proofreading.

Comments on the Quality of English Language

Necessary English editing service.

Author Response

Question: Necessary English editing service.

Response: Thank you for giving us the opportunity to proofread the English language of the papers. As suggested, we have had the entire manuscript proofread by a native speaker

Author Response File: Author Response.pdf