Analyzing the Impact of Carbon Mitigation on the Eurozone’s Trade Dynamics with the US and China

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

I am pleased to review the article titled “Analyzing the Impact of Carbon Mitigation on the Eurozone's Trade Dynamics with the US and China”. This study addresses the core question of how carbon pricing mechanisms, particularly the EU's Carbon Border Adjustment Mechanism (CBAM), influence trade flows with the US and China and affect carbon emissions in the Eurozone. The research is both original and relevant, particularly given the growing global emphasis on climate policy integration with trade frameworks. It effectively addresses a pressing gap in the empirical literature by employing a novel econometric approach to disentangle the dynamic impacts of imported goods on carbon emissions, a relationship previously discussed mostly at an aggregated level or using static models.

The originality of the paper lies in its dual application of Bayesian variable selection techniques (BART and BASAD) and Time-Varying Structural Vector Autoregressive (TV-SVAR) models. This integrated methodology is particularly suitable for high-dimensional data, where relationships are dynamic and nonlinear. By narrowing down from 53 to three key trade variables and examining their evolving influence on EU carbon emissions, the paper adds depth and clarity to the discourse on environmental trade-offs in international commerce. Moreover, the paper's ability to forecast future impacts of trade on emissions elevates its strategic value for policymakers.

In terms of methodology, the choice of Bayesian frameworks is well justified, and the explanation of the variable selection processes is thorough. However, while the econometric tools are competently applied, the description of the state-space modelling and Bayesian estimation procedures could benefit from greater clarity, especially for replication purposes. For instance, more detail is needed on prior specifications, convergence diagnostics, and software implementation. Additionally, the paper would benefit from a more explicit discussion of ethical considerations and model limitations, particularly with regard to potential endogeneity or omitted variable bias, which could affect the interpretation of causal impacts.

The conclusions are largely consistent with the evidence presented. The TV-SVAR results illustrate how the impact of US and Chinese imports on EU carbon emissions has changed over time, offering nuanced insights. The claim that certain goods (like manufactured products from the US) may eventually be exempt from stricter regulation is supported by the declining marginal effect observed. Likewise, the forecasted increase in emissions from other categories (e.g., petroleum products from China) is well justified. These insights directly address the research question and support informed policy recommendations.

References are comprehensive, relevant, and current, citing key works in environmental economics and Bayesian modelling. However, more literature could be included on global trade law implications of CBAM and its reception in international fora (e.g., WTO challenges), which would strengthen the policy discussion.

The tables and figures are well-organised and add clarity to the statistical results. The statistics showing time-varying coefficients are especially effective in illustrating the paper’s central claims. That said, the captions could be expanded slightly to guide interpretation without requiring the reader to cross-reference the main text.

The paper aligns with the interests of an international audience, particularly those working at the intersection of econometrics, environmental policy, and trade. Its relevance to policymakers, economists, and sustainability professionals makes it a substantial contribution. The social and economic implications are well-articulated, especially in the discussion and conclusion sections, where the authors effectively bridge academic findings with real-world policy responses. This includes impact on regulatory design, trade partner adjustments, and global carbon governance.

The introduction clearly outlines the study’s relevance and places it within the broader context of EU trade and climate policy. The literature review is structured and comprehensive, drawing from a range of primary and secondary sources. It could be strengthened by explicitly stating search strategies and keywords used. The methodology section is strong in its technical design but should be made slightly more transparent for general readers. Results are presented with clarity, and implications are discussed rigorously, including a useful reflection on how the study contributes to future forecasting.

In conclusion, the study demonstrates scientific soundness, originality, and practical significance. It successfully communicates complex econometric content in an accessible way, though improvements in clarity, particularly in the methodology and policy context, would enhance its replicability and impact.

Recommendation:

Accept after minor revisions, particularly to enhance methodological clarity, expand literature on legal-policy implications, and provide more transparency on estimation techniques.

Suggestions

•  The Bayesian methods (BART and BASAD, TV-SVAR) are well chosen but would benefit from:
• Clarification of prior assumptions.
• Details on convergence diagnostics (e.g., trace plots, R-hat values).
• Explanation of how hyperparameters were tuned or chosen.
• Mention the software/packages used (e.g., R, Python) and the availability of code or data for reproducibility.

• Explain how contemporaneous relationships were identified (e.g., Cholesky decomposition), and justify the ordering of variables.
• Add 1–2 references on the legal implications of CBAM, such as compatibility with WTO rules or likely trade partner retaliation (e.g., China, US concerns).
• Briefly explain how literature was selected. For example, “A targeted search of Web of Science and Scopus databases using keywords like ‘CBAM’, ‘carbon leakage’, and ‘trade emissions’ was conducted…”
• Expand captions slightly to explain what is being shown and its importance. E.g., for Figure 3, clarify that it shows forecasted marginal effects over a 24-month period
• Indicate sample size and specify whether the comparison was done in-sample or out-of-sample.
• Add a short subsection on limitations within the Discussion:

• Possibility of omitted variable bias or reverse causality.
• Limitations of monthly data in fully capturing lagged environmental impacts.
• Sensitivity to variable definitions (e.g., sectoral import classifications).

• Briefly note future research opportunities, such as including exports or using firm-level carbon intensity data.
• Expand on how developing countries might be affected by CBAM (beyond China and US), this would increase the paper’s global relevance.
• Briefly link the findings to SDGs or climate transition pathways.
• Correct repetitive phrases (e.g., “Manufactured Goods from the US” repeated heavily).Fix grammatical slips like "the carbon emission" to "carbon emissions".

 

Comments on the Quality of English Language

Correct repetitive phrases (e.g., “Manufactured Goods from the US” repeated heavily). Fix grammatical slips like "the carbon emission" to "carbon emissions".

Author Response

We sincerely thank the reviewers for their time, constructive feedback, and valuable suggestions. Their comments have significantly improved the quality and clarity of our manuscript. In response, we have carefully revised the paper and addressed all comments point by point. Revisions include clarifying prior assumptions and estimation strategies in our Bayesian framework (BART, BASAD, TVP-SVAR), improving transparency regarding convergence diagnostics and variable ordering, expanding our literature review to include legal-economic perspectives on CBAM, and enhancing the discussion with explicit limitations and future research directions.

All revisions are clearly marked in the updated manuscript using tracked changes. Page and line numbers are provided in each response for ease of reference. We hope the revised manuscript now meets the expectations of the reviewers and the editorial team.

Point-by-point response to Comments and Suggestions for Authors.

Comment	Revise
1. The Methodology part
Comments 1: The Bayesian methods (BART and BASAD, TV-SVAR) are well chosen but would benefit from: Clarification of prior assumptions.	- We have clarified the prior distributions used in both the BART and BASAD frameworks in the TV-SVAR model. - Please see [1] page 6, lines 260-272. [BART] [2] page 7, lines 299–301. [BASAD] [3] page 9, lines 344–362. [TV-SVAR]
Comments 2: Details on convergence diagnostics (e.g., trace plots, R-hat values).	-We have addressed this point in the revised methodology. - Please see page 9, lines 344–362.
Comments 3: Explanation of how hyperparameters were tuned or chosen. Mention the software/packages used (e.g., R, Python) and the availability of code or data for reproducibility.	- We now provide additional detail regarding model implementation. [1] BART hyperparameters were selected based on prior literature and tuning via 10-fold cross-validation. [ Please see page 6, lines 260-272.] [2] BASAD, hyperparameter settings followed recommendations from Narisetty & He (2014). [page 7, lines 293–301. [BASAD] [3] TVP-SVAR used automatic bandwidth selection. [Please see page 9, lines 344–362]
Comments 4: Explain how contemporaneous relationships were identified (e.g., Cholesky decomposition) and justify the ordering of variables.	- We have clarified that contemporaneous identification in the TVP-SVAR model follows a recursive identification scheme using the Cholesky decomposition of the residual covariance matrix. - Please see page 9, lines 344–362.
2. Literature reviews part
Comments 1: Add 1–2 references on the legal implications of CBAM, such as compatibility with WTO rules or likely trade partner retaliation (e.g., China, US concerns).	- We have added 1–2 references discussing its compatibility. - The revision can be found on page 4, lines 142–154.
Comments 2: Briefly explain how literature was selected. For example, “A targeted search of Web of Science and Scopus databases using keywords like ‘CBAM’, ‘carbon leakage’, and ‘trade emissions’ was conducted…”	- We have included a brief explanation of our literature search strategy. - This addition appears on page 3, lines 109–112.
3. Result part
Comments 1: Expand captions slightly to explain what is being shown and its importance. E.g., for Figure 3, clarify that it shows forecasted marginal effects over a 24-month period	Agree, we have revised it at the Figure 3.
Comments 2: Indicate sample size and specify whether the comparison was done in-sample or out-of-sample.	- We have added clarification on the sample size and confirmed that the evaluation was conducted in-sample. - The details can be found on page 14, lines 522–529.
4. Discussion Part
Comments 1: Add a short subsection on limitations within the Discussion: Possibility of omitted variable bias or reverse causality.	- We appreciate the reviewer’s insightful comments regarding the Discussion section. In response, we have added a new subsection titled “Limitations and Future Directions” to address the concerns raised.   - However, the comment “Comments 5: Expand on how developing countries might be affected by CBAM (beyond China and US), this would increase the paper’s global relevance.”. While our empirical results focus on China and the US due to data coverage, we have added a brief discussion noting. However, we do not extend the empirical claims beyond our dataset to avoid unsupported generalization.   - Mention exactly where in the revised manuscript this change can be found – page number [16-17] and line [604-622].
Comments 2: Limitations of monthly data in fully capturing lagged environmental impacts.
Comments 3: Sensitivity to variable definitions (e.g., sectoral import classifications).
Comments 4: Briefly note future research opportunities, such as including exports or using firm-level carbon intensity data.
Comments 5: Expand on how developing countries might be affected by CBAM (beyond China and US), this would increase the paper’s global relevance.
Comments 6: Briefly link the findings to SDGs or climate transition pathways.

 

Response to Comments on the Quality of English Language Point 1: Correct repetitive phrases (e.g., “Manufactured Goods from the US” repeated heavily).

Response 1: Thank you for the comment. To maintain consistency with the defined variables in our analysis, we have retained the term “Manufactured Goods from the US” where necessary. However, we have revised surrounding text to reduce redundancy and improve readability.

 

Point 2:  Fix grammatical slips like "the carbon emission" to "carbon emissions".

Resoponse 2: Minor grammatical errors have also been corrected throughout the manuscript. For instance, “the carbon emission” has been revised to “carbon emissions” where appropriate.

 

Note: We have also received a notification regarding a 4% textual similarity with a previously published. In response, we have carefully reviewed the repetition report and revised to ensure originality while preserving the technical accuracy of the content.

 

Reviewer 2 Report

Comments and Suggestions for Authors

The given paper explores the impacts of carbon mitigation policies, especially the Carbon Border Adjustment Mechanism (CBAM) on trade relationships between the Eurozone and its top partners, the USA and China. The study uses strong econometric techniques, namely Bayesian variable selection (BART and BASAD) and Time-Varying Structural Vector Autoregressions (TV-SVAR) to determine the import categories that have the most impact on EU carbon emissions. The authors employ monthly data in 2011-2023, and they concentrate on 15 main categories of imported goods and use control variables, including GDP and exchange rates.

The research is timely, addressing the emerging intersection of trade policy and environmental regulation. Its originality lies in the methodological approach—combining advanced Bayesian variable selection and dynamic modeling to disentangle time-varying relationships. Its focus on granular import categories from the US and China, instead of sector-aggregated data, provides new insights for policymakers and supports export strategy adaptation among EU trading partners.

The paper employs BART and BASAD for variable selection, followed by TV-SVAR modeling to capture the evolving impacts. The approach is well-explained and appropriately justified, with model selection criteria (AIC/BIC) favoring TV-SVAR over alternatives. Predictive accuracy is pursued using ten different forecasting models, with ARIMA models chosen for their superior performance. The data are described comprehensively, and controls are included to enhance robustness.

The results reveal that the US Manufactured Products already have a declining beneficial effect on the EU carbon emissions and, probably, will not be subject to the more stringent regulations. US Manufactured Goods and Chinese Petroleum Products are set to contribute to the rise in emissions and so there is a need to implement specific policy measures. Results are presented in a clear manner and discussion is well based on the available literature, which is an added value to the policymakers who have an interest to balance the trade and environmental objectives.

The article is well-written and has a logical structure in general. The findings are supported by the use of descriptive tables and visualizations. The translation to the English language with minor edits and explanation of certain graph legends and terms can enhance accessibility to a general audience.

Author Response

We are grateful to Reviewer for their constructive and encouraging feedback. We are pleased that the reviewer found the methodological design, use of granular trade data, and clarity of presentation to be strengths of the manuscript. In response to the minor suggestions regarding English language and accessibility, we have carefully reviewed the manuscript and made editorial improvements, including the clarification of certain graph legends and technical terms to ensure accessibility for a broader audience.

Point-by-point response to Comments and Suggestions for Authors

Comment	Revise
1. The Methodology part
Comments 1: The methodology is strong, combining BART, BASAD, and TVP-SVAR, and model selection is justified using AIC/BIC.	Thank you. We are glad the methodological approach was well received. No further changes were made here.
Comments 2: The use of ten forecasting models, with ARIMA selected, is appropriate.	Thank you for this observation. We retained this modeling strategy and clarified our model selection process in the methodology section.
Comments 3: The analysis uses disaggregated trade data rather than broad sectors, adding value.	We appreciate this recognition. No revision was required.
2. Result part
Comments 1: The results on US and Chinese imports are clearly presented and policy relevant.	Thank you. We maintained the structure and enhanced some descriptions in the Discussion section to improve clarity and relevance

Response to Comments on the Quality of English Language Point 1: improve translation and clarify graph legends and technical terms for general readers.

Response 1: We have reviewed the manuscript and implemented minor language edits to improve fluency. Graph legends and technical terminology were also revised for accessibility, especially in Figures 2 and 3.

Note: We have also received a notification regarding a 4% textual similarity with a previously published. In response, we have carefully reviewed the repetition report and revised to ensure originality while preserving the technical accuracy of the content.