Environmental Concern, Coal Transition, and Environmental Justice in Appalachian Communities: Evidence from Kentucky

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Overall evaluation: This paper focuses on the Appalachian community in Kentucky, USA, and examines issues related to coal decline, environmental concerns, and environmental justice. Through a statewide online questionnaire survey and ordered logistic regression, it systematically reveals the influence of social demographic attributes, residential location, and political stance on environmental risk perception, clearly presenting the differences in attitudes of local residents towards the causes of coal decline, employment transition intentions, and environmental issues. The research contributions are significant: First, it fills the gap in public perception and energy transition equity research in coal-dependent areas, integrating economic transformation, environmental justice, and public attitudes; Second, based on large sample empirical data, it clarifies the mechanism of key variables such as gender, urban-rural, and party affiliation, providing a local basis for energy transition policy design; Third, it explores the employment transition preferences and re-training intentions of former coal miners, providing practical references for regional labor force transformation. The overall topic is in line with the frontiers of sustainable development and energy equity, with solid data, clear logic, and good theoretical and application value. However, there is still room for improvement in methodological rigor, the depth of result explanation, and format standardization. Main review comments are as follows:

1. Regression results only report the statistically significant marginal effects, without presenting the results of the full model and the direction of the effects of non-significant variables. Readers cannot determine the overall picture of the variable influence. All variable coefficients, standard errors, and significance levels should be presented, distinguishing 1%, 5%, and 10% significance levels, and supplementing the overall significance test and prediction accuracy rate of the model, systematically displaying the mechanism of variable influence, and improving the completeness and readability of the results.
2. The study includes gender, age, urban-rural, and party affiliation as core independent variables, but does not include key control variables such as education level, income, and health status. These factors will significantly affect environmental perception and policy attitudes. It is recommended to expand the control variable system, through stepwise regression to screen significant variables, compare results of different model settings, and eliminate confounding factors interference, to precisely identify the independent influence of core social demographic attributes on environmental concerns.
3. The discussion part overly relies on descriptive statistics, with insufficient theoretical dialogue of regression results and failure to integrate empirical findings with environmental justice and energy transition theories. It should closely follow environmental justice theory, risk perception theory, interpret the theoretical connotations of gender and party differences, compare with international similar research conclusions, and clearly identify the theoretical contributions and innovation points of this study.
4. The conclusion part summarizes the research contributions generally, without clearly distinguishing theoretical contributions, practical contributions, and policy contributions, and does not highlight the unique value of the study in the Appalachian coal region transformation. It should condense contributions dimensionally: theoretically improve the framework of factors influencing environmental concerns, practically provide labor force transition data, and policy-wise propose fair transition paths, clearly define the research positioning and academic value.
5. In Section 2.1 of the questionnaire design of this paper, "pilot sample of 65 respondents" does not indicate the sample source, pre-research time, and feedback optimization details. It should be supplemented to the corresponding paragraph.
6. In Table 1, "Coal miners (%)", "Area of residence (Rural)" have no χ2 test values, and the "-" is used without explanation, and annotations should be added to explain.
7. In Section 3.1 of the paper, Figure 3.1 title "Survey response related to proximity to coal mines." is too brief. It should be supplemented to clarify the research question, and the format of the chart title should be standardized.
8. In Table 2 of Section 3.2, "Ordered Logistic Regression" table header is misaligned across columns, and variable names and dependent variables are misaligned. The table structure should be adjusted.

9 The citation format of the references in the reference section is not uniform, and some are missing DOI, publication year, journal volume and issue numbers. For example, the information of reference [21] is incomplete. It should be corrected according to the MDPI format.

Author Response

Reviewer 1

Overall evaluation: This paper focuses on the Appalachian community in Kentucky, USA, and examines issues related to coal decline, environmental concerns, and environmental justice. Through a statewide online questionnaire survey and ordered logistic regression, it systematically reveals the influence of social demographic attributes, residential location, and political stance on environmental risk perception, clearly presenting the differences in attitudes of local residents towards the causes of coal decline, employment transition intentions, and environmental issues. The research contributions are significant: First, it fills the gap in public perception and energy transition equity research in coal-dependent areas, integrating economic transformation, environmental justice, and public attitudes; Second, based on large sample empirical data, it clarifies the mechanism of key variables such as gender, urban-rural, and party affiliation, providing a local basis for energy transition policy design; Third, it explores the employment transition preferences and re-training intentions of former coal miners, providing practical references for regional labor force transformation. The overall topic is in line with the frontiers of sustainable development and energy equity, with solid data, clear logic, and good theoretical and application value. However, there is still room for improvement in methodological rigor, the depth of result explanation, and format standardization. Main review comments are as follows:

1. Regression results only report the statistically significant marginal effects, without presenting the results of the full model and the direction of the effects of non-significant variables. Readers cannot determine the overall picture of the variable influence. All variable coefficients, standard errors, and significance levels should be presented, distinguishing 1%, 5%, and 10% significance levels, and supplementing the overall significance test and prediction accuracy rate of the model, systematically displaying the mechanism of variable influence, and improving the completeness and readability of the results.

While diagnostic measures can enhance analytical rigor, reporting the complete set of diagnostics and marginal effects for multiple ordered logistic models would produce excessively large tables and potentially obscure the study’s substantive findings. Following the study’s exploratory objective of identifying broad socio-demographic patterns in environmental concern among Kentucky residents rather than to provide an exhaustive econometric evaluation of model performance, we prioritized concise reporting of statistically significant and policy-relevant results while explicitly acknowledging this limitation in the revised manuscript.

New addition: Line 296-306: Although model diagnostics such as proportional odds assumption tests, pseudo-R² statistics, and confidence intervals for marginal effects can strengthen the rigor and transparency of ordered logistic regression analyses, the primary objective of this study was to identify broad socio-demographic patterns in environmental concern across key environmental issues in Kentucky. Given the large number of ordered logistic regression outputs generated across multiple environmental concern categories, presenting the full set of diagnostics, marginal effects, and outcome probabilities for all response categories would substantially increase table complexity and reduce interpretability. To maintain clarity and keep the discussion concise, the analysis focuses primarily on statistically significant findings at the 1%, 5%, and 10% significance levels across the main socio-demographic variables. Particular attention is given to the two substantively informative outcome categories, “Not at all concerned” and “Extremely concerned,” which capture the strongest contrasts in respondents’ environmental concern levels. This reporting approach is intended to emphasize the study’s central findings while maintaining readability and analytical focus.

2. The study includes gender, age, urban-rural, and party affiliation as core independent variables, but does not include key control variables such as education level, income, and health status. These factors will significantly affect environmental perception and policy attitudes. It is recommended to expand the control variable system, through stepwise regression to screen significant variables, compare results of different model settings, and eliminate confounding factors interference, to precisely identify the independent influence of core social demographic attributes on environmental concerns.

Thank you for this comment. We acknowledge that additional socio-demographic variables could provide further insight into variation in environmental concern. However, the primary goal of this study was to identify broad patterns in Kentucky residents’ perceptions of coal decline, environmental justice concerns, and environmental attitudes using a focused set of theoretically relevant predictors. To maintain analytical clarity, model parsimony, and brevity, the analysis concentrated on a limited number of core socio-demographic variables (gender, age, rural residence, and political affiliation) which have been consistently identified in prior literature as important determinants of environmental concern and energy transition attitudes. Expanding the model to include a substantially larger number of socio-demographic variables would increase model complexity and reduce interpretability relative to the study’s exploratory objectives. We have clarified this rationale in the revised manuscript in lines 304 to 314. Although model diagnostics such as proportional odds assumption tests, pseudo-R² statistics, and confidence intervals for marginal effects can strengthen the rigor and transparency of ordered logistic regression analyses, the primary objective of this study was to identify broad socio-demographic patterns in environmental concern across key environmental issues in Kentucky. Given the large number of ordered logistic regression outputs generated across multiple environmental concern categories, presenting the full set of diagnostics, marginal effects, and outcome probabilities for all response categories would substantially increase table complexity and reduce interpretability. To maintain clarity and keep the discussion concise, the analysis focuses primarily on statistically significant findings at the 1%, 5%, and 10% significance levels across the main socio-demographic variables. Particular attention is given to the two substantively informative outcome categories, “Not at all concerned” and “Extremely concerned,” which capture the strongest contrasts in respondents’ environmental concern levels. This reporting approach is intended to emphasize the study’s central findings while maintaining readability and analytical focus.

3. The discussion part overly relies on descriptive statistics, with insufficient theoretical dialogue of regression results and failure to integrate empirical findings with environmental justice and energy transition theories. It should closely follow environmental justice theory, risk perception theory, interpret the theoretical connotations of gender and party differences, compare with international similar research conclusions, and clearly identify the theoretical contributions and innovation points of this study.

Thank you for this insightful comment. In response, we substantially expanded the discussion of the key socio-demographic variables, rural location, gender, age, and political affiliation to provide stronger theoretical grounding and contextual interpretation of the regression findings. The revised discussion now more explicitly connects the empirical results to literature on environmental justice, environmental risk perception, climate perception, political polarization, and energy transition scholarship as follows:

For rural location, we expanded the discussion to examine how economic dependence on extractive industries, occupational identity, cultural values, and regional economic concerns in coal-dependent communities may shape environmental concern and perceptions of environmental policy.  Line 348 to 358:  In many Appalachian communities, environmental policies and climate-related regulations are often interpreted through the lens of employment security, regional economic stability, and concerns regarding the future of coal-dependent livelihoods [2,4,5]. As a result, rural residents may perceive environmental regulations and energy transition policies differently from urban populations, particularly when such policies are viewed as potentially threatening local industries or economic opportunities. At the same time, lower concern regarding climate change does not necessarily imply lower awareness of environmental issues overall. Previous research in Appalachian and rural communities suggests that residents may express stronger concern regarding localized and immediately observable environmental challenges such as water quality, land degradation, flooding, or air pollution than broader and more politically polarized issues such as climate change [4,17]. These findings highlight the importance of recognizing geographic, economic, and cultural heterogeneity when designing environmental communication strategies and energy transition policies for coal-dependent and rural communities.

For gender, we incorporated perspectives from environmental sociology and environmental justice literature to discuss gendered differences in environmental risk perception, caregiving roles, community engagement, and participation in environmental justice advocacy within Appalachian communities.  Line 324 to 338: Gender differences in environmental concern may also reflect variations in occupational exposure, caregiving roles, community engagement, and perceptions of environmental justice issues [4]. Women in Central Appalachia account for 70% of environmental justice advocacy, often framing environmental concerns through community health and  family well-being [4,16]. Perhaps these patterns may help explain stronger concern among female respondents regarding environmental degradation, pollution, and climate-related issues. Concurently, gender differences in environmental concern within coal-producing regions are further shaped by broader social and cultural dynamics associated with coal-mining identity, labor participation, and regional economic dependence on extractive industries [5]. Because women have historically been underrepresented in coal-mining employment (10-17%) yet remain highly involved in household and community health decision-making, they may perceive environmental risks and energy transition policies differently from male respondents more directly connected to coal-related occupations [31]. These findings underscore the importance of incorporating gender perspectives into environmental communication, environmental justice initiatives, and energy transition policy development in Appalachian and coal-dependent regions. Recognizing gendered differences in environmental concern may help support more inclusive and socially responsive approaches to workforce transition, community adaptation, and environmental decision-making.

Regarding age, we expanded the interpretation to consider generational differences in climate concern, including the influence of education, media exposure, lived economic experiences, and regional coal-related identity on environmental attitudes.  Line 366 to 378: Several factors may help explain age-related differences in environmental concern within Kentucky and Appalachian contexts. Younger individuals may be more exposed to climate change discourse through educational institutions, digital media, and broader public discussions surrounding sustainability, environmental justice, and energy transition policies [4]. In contrast, older residents in coal-dependent regions may interpret environmental and energy issues through lived experiences shaped by historical economic reliance on coal production, regional labor identity, and long-standing political and cultural narratives surrounding energy development [5, 18]. At the same time, lower levels of climate concern among older respondents should not be interpreted as a general absence of environmental concern. Previous research suggests that older populations may express concern regarding environmental issues with immediate and observable local impacts such as air quality, water contamination, land degradation, or community health, while assigning lower priority to broader and longer-term climate-related concerns [4, 17, 32]. These findings highlight the importance of considering generational heterogeneity when developing environmental communication strategies and energy transition policies in coal-dependent and Appalachian communities, where age, economic history, and community identity may shape environmental attitudes in distinct ways.

For political affiliation, we strengthened the discussion by incorporating literature on climate change politicization, partisan identity, political and media messaging, and how competing narratives surrounding economic costs, scientific uncertainty, government intervention, and energy transition policies may influence environmental perceptions in coal-dependent regions. Line 386 to 394: Over the past several decades, climate change has become increasingly politicized in the United States, with political leaders, media outlets, and partisan organizations often framing climate change and environmental regulation through competing narratives regarding economic costs, scientific uncertainty, government intervention, and energy independence (McCright & Dunlap, 2011; Brulle et al., 2012). In conservative political discourse, climate policies are frequently portrayed as threats to employment, industrial competitiveness, and fossil-fuel-dependent communities, while clean energy transitions may be framed as disruptive to traditional energy sectors and regional economies. Such political messaging may influence how individuals interpret climate science, assess environmental risks, and perceive the trade-offs associated with environmental regulation and energy transition policies.

To support these revisions, we integrated additional literature from environmental justice, climate perception, environmental sociology, and energy transition research, including work by Bell and York (2010), Bell and Braun (2010), McCright and Dunlap (2011), Hamilton (2012), Brulle et al. (2012), Carley et al. (2018), Lewin (2019), Caniglia et al. (2017), and related scholarship.

4. The conclusion part summarizes the research contributions generally, without clearly distinguishing theoretical contributions, practical contributions, and policy contributions, and does not highlight the unique value of the study in the Appalachian coal region transformation. It should condense contributions dimensionally: theoretically improve the framework of factors influencing environmental concerns, practically provide labor force transition data, and policy-wise propose fair transition paths, clearly define the research positioning and academic value.

Thank you for this valuable comment. In response, we substantially revised the conclusion to provide a stronger synthesis of the study’s empirical findings and theoretical contributions. The revised conclusion now more explicitly situates the results within broader discussions of environmental justice, energy transition, environmental risk perception, and public attitudes in coal-dependent regions.

Specifically, the revised conclusion strengthens the interpretation of the socio-demographic findings by discussing how gender, rural location, age, and political affiliation shape environmental concern within Kentucky and Appalachian contexts. We incorporated discussion of gendered environmental risk perception and environmental justice perspectives, rural and geographic heterogeneity in environmental attitudes, generational differences in climate concern, and the influence of political identity, climate change politicization, and political messaging on perceptions of environmental risks and transition policies.

The revised conclusion also clarifies that lower climate concern among certain demographic groups should not necessarily be interpreted as a lack of environmental concern overall but may instead reflect stronger emphasis on localized environmental issues such as pollution, environmental degradation, and community health impacts. Additionally, we expanded the policy implications by emphasizing the importance of equitable, locally informed, and socially responsive transition strategies that incorporate workforce adaptation, economic diversification, environmental justice considerations, and recognition of social and political heterogeneity within coal-dependent communities.

5. In Section 2.1 of the questionnaire design of this paper, "pilot sample of 65 respondents" does not indicate the sample source, pre-research time, and feedback optimization details. It should be supplemented to the corresponding paragraph.

Added sentence Line 101 to 104: To ensure clarity, relevance, and appropriate survey length, the questionnaire was pretested using a pilot sample of 65 respondents recruited through Qualtrics online distribution channels to assess questionnaire clarity, wording, response consistency and completion time. Feedback from the pilot study was used to improve wording, question flow, and response consistency, following standard survey validation procedures [21, 22].

6.In Table 1, "Coal miners (%)", "Area of residence (Rural)" have no χ2 test values, and the "-" is used without explanation, and annotations should be added to explain.

Already explained in Line 132 to 136 As follows: Results indicated no statistically significant differences across most socio-demographic characteristics, except for coal mining employment and area of residence. The proportion of coal miners in the sample (1.62%) was significantly higher than the state population proportion (0.078%), reflecting the use of pre-stratification weights to ensure coal workers’ perspectives were adequately represented. Similarly, rural residents accounted for 58.51% of the sample, compared with 41.38% statewide, underscoring the need to balance representation between mining and non-mining counties.

 

7. In Section 3.1 of the paper, Figure 3.1 title "Survey response related to proximity to coal mines." is too brief. It should be supplemented to clarify the research question, and the format of the chart title should be standardized.

Rephrased the research question to: Fig. 3.1: Survey response to questions related to proximity to coal mines. Approximately how close do you live to a coal mine?

Standardized the chart titles.

 

8. In Table 2 of Section 3.2, "Ordered Logistic Regression" table header is misaligned across columns, and variable names and dependent variables are misaligned. The table structure should be adjusted.

Made the changes

9 The citation format of the references in the reference section is not uniform, and some are missing DOI, publication year, journal volume and issue numbers. For example, the information of reference [21] is incomplete. It should be corrected according to the MDPI format.

Made the adjustments

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

After carefully reviewing the author's manuscript, I believe that the chosen topic has strong practical significance, particularly in the areas of coal transition, environmental justice, and social awareness of energy transition. However, the current version of the manuscript still has significant room for improvement in terms of theoretical depth, model reporting standardization, methodological completeness, and result interpretation. I would like to offer the following suggestions for the author's consideration:

1、The manuscript does not further clarify the theoretical embedding of environmental justice. Specifically, the manuscript mentions environmental justice multiple times. However, overall, it leans more towards public environmental awareness surveys rather than establishing a truly systematic analytical framework for environmental justice. I suggest the authors add a section on theoretical frameworks, considering the systematic introduction of environmental justice theory and constructing a logical chain of coal decline—social vulnerability—environmental awareness—transitional equity. This would clarify the manuscript's theoretical positioning within environmental justice research. Otherwise, current understanding of environmental justice remains largely at the conceptual level rather than a genuine theoretical analysis.

2、I suggest adding a section on explicit research hypotheses to the manuscript. While the manuscript currently presents specific research objectives, it lacks standardized research hypotheses. These potential specific research objectives have only been explained to some extent in the discussion phase, but have not been formalized into theoretical expectations in the research design phase. I would like to suggest that the authors add a section on research hypotheses and further develop reasonable hypotheses through theoretical analysis. This would significantly enhance the theoretical logic and the standardization of the quantitative analysis in the paper.

3、The current sample representativeness interpretation is still insufficient. The manuscript clearly points out that the proportion of coal miners is significantly higher than the overall population of Kentucky, and the proportion of rural residents is also significantly higher. However, the current manuscript simply explains this as pre-stratified sampling. I believe the authors have not clearly explained the specific weighting method, whether post-stratification adjustment was performed, or whether weight correction was used in the final regression. I suggest the authors should improve on these points.

4、I would like to suggest that the authors consider adding a complete questionnaire appendix and variable coding instructions. The current manuscript only provides a general description of the questionnaire structure. However, the authors have not clearly provided detailed information such as the original items and the specific coding of the Likert scale. I suggest that the authors optimize these details, which will significantly improve the reproducibility of the research.

5、The core assumptions of Ordered Logistic Regression mentioned in the manuscript have not been verified. While the authors explicitly state the use of an ordered logit model, they also acknowledge that proportional odds assumptions are not reported. I would suggest that the authors at least supplement the Brant test and likelihood-ratio test. Otherwise, the reliability of the current model may be somewhat insufficient.

6、Table 2 in the manuscript reports multiple marginal effects, but the subsequent discussion sections do not provide rigorous explanations of these marginal effects. I suggest that the authors consider a more standardized explanation of probability marginal changes in the manuscript, avoiding the direct use of expressions such as "increase concern level."

7、I personally believe that the current manuscript's control variable system is weak, omitting many key socioeconomic variables. Perhaps the authors could consider adding variables from a reasonable perspective? I suggest they focus on the following areas for expansion: education level, income, and whether they have worked in the coal industry. Otherwise, the current model may suffer from variable omission bias.

8、The excessive mixing of Results and Discussion sections in the manuscript made it difficult for me to read. In a paper published in an academic journal, the Results and Discussion sections should theoretically be clearly distinguished, not mixed together. I would suggest the author restructure these sections. The Results section should only report: data facts, significance, and statistical findings. The Discussion section should focus on explaining: Why did this happen? Are there similarities/conflicts with existing research findings and results? What are the theoretical contributions and potential benefits of the manuscript? Perhaps the author could consult the following literature for inspiration. Specifically, the reference links are as follows: https://doi.org/10.1038/s44284-025-00303-0.； https://doi.org/10.1038/s42949-026-00377-2.

9、The academic rigor of the figures and tables in the current manuscript needs significant improvement. Several figures and tables appear to have rough compositional details, primarily manifested in low resolution, inconsistent fonts, and excessive white space. A high-quality academic journal article adheres to certain standards regarding composition. I suggest the author completely reconstruct the existing figures and tables and specifically improve the image quality. I recommend the author read the following high-quality paper and refer to its compositional details to improve the images in the manuscript accordingly. Specifically, the link to the paper is as follows: https://doi.org/10.1038/s41586-025-09922-y

10、The current manuscript only contains correlation analysis, but some statements show a potential tendency to express causality, which I believe should be approached with greater caution. This is because causal logic analysis employs professional analytical methods and models. However, the author's research in the manuscript is actually a cross-sectional survey, not a causal identification design. I would like to suggest that the author review the entire manuscript and be more cautious in expressing and elaborating on causal logic to avoid exaggeration.

11、As a scientific paper in the form of an investigative study, the manuscript currently has strong practical significance, but its international theoretical contribution is still insufficient. I suggest that the authors consider increasing theoretical contributions, such as providing new evidence on the influence of coal culture identity on environmental perception. Furthermore, the authors could perhaps further clarify the manuscript's specific contributions to policy recommendations.

Overall, the manuscript presents content with practical research significance. Furthermore, it has policy value regarding environmental transition issues. However, the author needs to fully consider the questions I raised and make significant revisions to the original manuscript to meet the journal's publication requirements. I look forward to receiving the revised version from the author.

Author Response

After carefully reviewing the author's manuscript, I believe that the chosen topic has strong practical significance, particularly in the areas of coal transition, environmental justice, and social awareness of energy transition. However, the current version of the manuscript still has significant room for improvement in terms of theoretical depth, model reporting standardization, methodological completeness, and result interpretation. I would like to offer the following suggestions for the author's consideration:

1、The manuscript does not further clarify the theoretical embedding of environmental justice. Specifically, the manuscript mentions environmental justice multiple times. However, overall, it leans more towards public environmental awareness surveys rather than establishing a truly systematic analytical framework for environmental justice. I suggest the authors add a section on theoretical frameworks, considering the systematic introduction of environmental justice theory and constructing a logical chain of coal decline—social vulnerability—environmental awareness—transitional equity. This would clarify the manuscript's theoretical positioning within environmental justice research. Otherwise, current understanding of environmental justice remains largely at the conceptual level rather than a genuine theoretical analysis.

 

Thank you for this valuable comment. We respectfully contend that environmental justice constitutes a central and recurring theme throughout the manuscript, although we have further strengthened and clarified this framing in the revised version.

Environmental justice is introduced early in the manuscript in the first instance we open the topic of discussion on the prevalence of environmental justice concerns and other contributor scholarly work in the field from  Line 55 to 63: Environmental justice concerns have also become increasingly prominent throughout Appalachian coalfields. Communities near coal mining operations frequently experience environmental degradation, including deforestation, biodiversity loss, water contamination, flooding, and air pollution, which contribute to long-term public health risks [10,12]. Mining practices such as mountaintop removal have been associated with ecological destruction and adverse health outcomes, including respiratory illness, cardiovascular disease, cancer, and reduced quality of life among nearby residents [13, 14]. Previous studies have documented elevated rates of mortality and chronic disease in Appalachian mining communities, demonstrating how environmental and health burdens are disproportionately concentrated in economically marginalized regions [9, 15]. Environmental justice scholars further argue that coal-dependent communities often experience unequal exposure to environmental hazards while possessing limited political and economic resources to address these impacts [16,17].

Secondly  through discussion of the unequal environmental, economic, and health burdens associated with coal extraction and coal decline in Appalachian communities, including issues of community vulnerability, environmental health impacts, and workforce transition challenges (Introduction, Lines 46 to 65: The decline of coal mining in Kentucky has generated significant economic, social, and environmental challenges, particularly within coal-dependent communities. Reduced coal production has contributed to widespread job losses, declining household incomes, shrinking local tax revenues, and increased outmigration from rural counties, weakening local economies and public services [1,7]. The manuscript further makes adjustments and situates the study within environmental justice scholarship through its theoretical framing, which discusses how energy transition processes generate uneven social, economic, and environmental consequences across coal-dependent populations (Theoretical Framework, objective and hypothesis, Lines 74 to 79; Our main objective is to investigate residents’ association with coal mining, perceptions of the causes of the coal industry's decline, and experiences with environmental concerns. The study also examines how socio-demographic characteristics, including gender, age, rural residence, and political affiliation, influence levels of environmental concern regarding climate change, air pollution, water pollution, deforestation, and environmental degradation. We hypothesize that, factors such as gender, age, rural location, and political affiliation are expected to shape how individuals perceive environmental risks, environmental justice issues, and coal transition pathways within coal-dependent communities.

The empirical findings revisions and interpretation also engage directly with environmental justice themes. In the Results, socio-demographic differences in environmental concern, particularly across gender, rural location, age, and political affiliation, highlight how perceptions of environmental risks and transition policies are socially heterogeneous rather than uniformly distributed. The Discussion expands this interpretation by connecting these patterns to environmental justice, environmental risk perception, political messaging, and energy transition literature, emphasizing how economic dependence, community identity, localized environmental burdens, and differential exposure to environmental risks shape perceptions within Kentucky and Appalachian coal communities (Discussion, Lines 324 to 338 on gender; Lines 348 to 358 on Rural location: lines 366 to 378 on age; Lines 386 to 394  on political affiliation

Additionally, the revised conclusion strengthens the manuscript’s environmental justice contribution by emphasizing equitable, locally informed, and socially responsive energy transition strategies that recognize workforce adaptation, economic diversification, and heterogeneous community experiences. Thus, while environmental justice may not serve as the sole organizing theory of the paper, it remains a substantive and integrated component of the manuscript’s conceptual framing, empirical interpretation, and policy implications. (See reviewer 1 comments). Lines 408 to 409 and lines 419 to 424.  

2、I suggest adding a section on explicit research hypotheses to the manuscript. While the manuscript currently presents specific research objectives, it lacks standardized research hypotheses. These potential specific research objectives have only been explained to some extent in the discussion phase but have not been formalized into theoretical expectations in the research design phase. I would like to suggest that the authors add a section on research hypotheses and further develop reasonable hypotheses through theoretical analysis. This would significantly enhance the theoretical logic and the standardization of the quantitative analysis in the paper.

Thank you for your valuable comment, we have added a reasonable hypothesis to enhance the theoretical logic and standardization of the quantitative analysis in the paper as follows.

Lines 74 to 79: Our main objective is to investigate residents’ association with coal mining, perceptions of the causes of the coal industry's decline, and experiences with environmental concerns. The study also examines how socio-demographic characteristics, including gender, age, rural residence, and political affiliation, influence levels of environmental concern regarding climate change, air pollution, water pollution, deforestation, and environmental degradation. We hypothesize that, factors such as gender, age, rural location, and political affiliation are expected to shape how individuals perceive environmental risks, environmental justice issues, and coal transition pathways within coal-dependent communities.

Removed: This study contributes to the growing body of research on environmental justice and the energy transition by examining Kentucky residents’ perceptions of coal's decline and key environmental concerns.

3、The current sample representativeness interpretation is still insufficient. The manuscript clearly points out that the proportion of coal miners is significantly higher than the overall population of Kentucky, and the proportion of rural residents is also significantly higher. However, the current manuscript simply explains this as pre-stratified sampling. I believe the authors have not clearly explained the specific weighting method, whether post-stratification adjustment was performed, or whether weight correction was used in the final regression. I suggest the authors should improve on these points.

We thank the reviewer for this important suggestion and agree that additional clarification regarding the sampling and weighting procedures improves the transparency of the manuscript. In response, we have expanded the Methods section to more clearly explain our sampling design and treatment of sample imbalance.

Specifically, our study employed pre-stratification weighting during sample recruitment, whereby coal miners and rural residents were intentionally oversampled to ensure sufficient representation of stakeholder groups central to Kentucky’s energy transition context. This approach was adopted because these populations are disproportionately affected by energy transition policies and would likely be underrepresented under proportional population sampling, consistent with practices in energy transition and just transition research (e.g., Mayer, 2018, Carley et al., 2018, Oluoch et al., 2025a).

We further clarify in the revised manuscript that no post-stratification adjustment was performed after data collection, as the primary purpose of the sampling design was analytic representation of transition-affected groups rather than strict population inference. Accordingly, the final regression analyses were conducted using the recruited analytic sample without additional post-survey weight correction. We now explicitly state this methodological choice and its rationale in the revised Methods section as follows:

Line 132 to 139: The proportion of coal miners in the sample (1.62%) exceeded their share of  the state population proportion (0.078%) due to the intentional use of purposive pre-stratification weights to ensure  adequate representation of coal workers’ perspectives, consistent with approaches used by Carley et al., [2], Oluoch et al., [4], and Mayer [27]. No post-stratification weighting adjustment was applied after data collection, and regression analyses were estimated using the analytic sample as recruited. Similarly, rural residents accounted for 58.51% of the sample, compared with 41.38% statewide, underscoring the need to balance representation between mining and non-mining counties.

4、I would like to suggest that the authors consider adding a complete questionnaire appendix and variable coding instructions. The current manuscript only provides a general description of the questionnaire structure. However, the authors have not clearly provided detailed information such as the original items and the specific coding of the Likert scale. I suggest that the authors optimize these details, which will significantly improve the reproducibility of the research.

However, due to the sensitive nature of some survey items, study context considerations, and IRB-related participant confidentiality requirements, certain materials may require controlled access. Therefore, where necessary, the questionnaire and additional supporting materials may also be made available upon reasonable request by contacting the corresponding authors, subject to institutional ethical guidelines and data protection considerations.

We have added the following statement to the manuscript:

Data Availability Statement: The data that supports the finding of this study is available on request from the corresponding author due to ethical reasons.

We appreciate the reviewer’s recommendation, which helped us improve the transparency and accessibility of the study materials.

5、The core assumptions of Ordered Logistic Regression mentioned in the manuscript have not been verified. While the authors explicitly state the use of an ordered logit model, they also acknowledge that proportional odds assumptions are not reported. I would suggest that the authors at least supplement the Brant test and likelihood-ratio test. Otherwise, the reliability of the current model may be somewhat insufficient.

We thank the reviewer for this thoughtful recommendation. We acknowledge that diagnostic tests such as the Brant test and likelihood ratio (LR) tests can be informative in assessing assumptions and model specification in ordinal regression frameworks. However, after careful consideration, we elected not to include these additional diagnostics in the main manuscript for two reasons. First, the paper’s analytical objective is primarily substantive and policy-oriented, focusing on identifying patterns of preferences and determinants of support within Kentucky’s energy transition context, rather than developing or comparing alternative econometric model specifications. The modeling approach, assumptions, variable selection, and interpretation of results were already described in sufficient detail to support the study’s analytical objectives.

We included the  Line 123 to 125: This theoretical framework follows the interpretation used by Arikawa et al. [21], Karlström and Ryghaug [23], and Oluoch et al. [22], allowing for analysis of how socio-demographic characteristics and environmental concerns influence the likelihood of stronger or weaker support for policy interventions while observing the ordinal structure of the dependent variable.

Second, given journal space and readability considerations, we sought to maintain methodological brevity by limiting the presentation to diagnostics and statistical information most directly relevant to interpreting the reported findings. Inclusion of additional diagnostic outputs such as the Brant and LR tests would considerably expand the methodological discussion and table without materially changing the substantive conclusions of the study.

To improve clarity, however, we have revised the Methods section to further explain the rationale for the selected modeling approach and the adequacy of the reported specification for addressing the study objectives. Additional technical details may be made available upon reasonable request. We respectfully believe that the current methodological presentation is appropriate for the scope, objectives, and intended audience of the paper, while remaining sufficiently transparent regarding the analytical approach employed.

6、Table 2 in the manuscript reports multiple marginal effects, but the subsequent discussion sections do not provide rigorous explanations of these marginal effects. I suggest that the authors consider a more standardized explanation of probability marginal changes in the manuscript, avoiding the direct use of expressions such as "increase concern level."

Specifically, we have reduced the use of informal expressions such as “increase concern level” and instead clarify findings in terms of changes in the predicted probability of observing higher or lower response categories, consistent with the interpretation of marginal effects in ordinal modeling frameworks. We have also expanded selected explanations to more explicitly connect coefficient direction, marginal probability changes, and substantive implications for energy transition preferences and attitudes. At the same time, we aimed to maintain balance between statistical precision and accessibility for a multidisciplinary readership. Because the primary objective of the paper is to communicate substantive insights regarding energy transition perceptions and policy implications, we avoided overly technical repetition of marginal probability interpretations for every estimated effect where the direction and substantive meaning were already adequately explained in the text.

7、I personally believe that the current manuscript's control variable system is weak, omitting many key socioeconomic variables. Perhaps the authors could consider adding variables from a reasonable perspective? I suggest they focus on the following areas for expansion: education level, income, and whether they have worked in the coal industry. Otherwise, the current model may suffer from variable omission bias.

 

We thank the reviewer for this constructive suggestion and appreciate the concern regarding model specification and potential omitted variable bias. After careful consideration, however, we believe the current set of explanatory and control variables is appropriate for the scope, theoretical framing, and analytical objectives of the study.

The variable selection was guided by the study’s conceptual framework, prior literature on energy transition attitudes and public preferences, and the objective of maintaining a parsimonious model focused on key determinants relevant to Kentucky’s energy transition context. Several of the factors suggested by the reviewer are conceptually related to variables already incorporated into the analysis. For example, occupational and energy-sector proximity effects are partially captured through variables reflecting stakeholder background, geographic context, and transition-related experiences.

With respect to coal industry employment specifically, the study design already intentionally incorporates representation from coal miners and coal-affected populations through the sampling framework, allowing the analysis to capture perspectives from groups central to the transition debate. Regarding education and income, while these variables can be informative in some energy opinion studies, they were not the primary focus of the present analytical framework, and expanding the model with additional socioeconomic controls would substantially broaden the scope of inquiry beyond the central research questions.

Importantly, the purpose of the model is not exhaustive socioeconomic prediction but rather examination of relationships among theoretically motivated factors associated with energy transition perceptions and policy preferences. The current specification produced interpretable and substantively meaningful results that are consistent with established literature and adequately support the study’s conclusions.

To improve transparency, we have further clarified the rationale underlying variable selection in the Methods section and acknowledge in the discussion that future work could explore expanded socioeconomic specifications, including education, income, and industry employment history, to assess potential heterogeneity in transition attitudes.

8、The excessive mixing of Results and Discussion sections in the manuscript made it difficult for me to read. In a paper published in an academic journal, the Results and Discussion sections should theoretically be clearly distinguished, not mixed together. I would suggest the author restructure these sections. The Results section should only report: data facts, significance, and statistical findings. The Discussion section should focus on explaining: Why did this happen? Are there similarities/conflicts with existing research findings and results? What are the theoretical contributions and potential benefits of the manuscript? Perhaps the author could consult the following literature for inspiration. Specifically, the reference links are as follows: https://doi.org/10.1038/s44284-025-00303-； https://doi.org/10.1038/s42949-026-00377-2.

 

 In response, we have revised the Results section to improve readability and strengthen the logical organization of the questionnaire findings. Specifically, we have introduced dedicated subsections corresponding to each major questionnaire group to provide a clearer thematic structure and facilitate interpretation of the results. In addition, for the section addressing respondents’ levels of concern regarding environmental issues, we have further refined the presentation by subdividing the analysis according to key sociodemographic characteristics, including gender, age, political affiliation, and geographic location. These revisions were made to improve clarity, enhance comparability across respondent groups, and provide a more systematic presentation of patterns in environmental concern and energy transition attitudes. We believe this restructuring substantially improves the accessibility and interpretability of the manuscript while maintaining the integrity of the original analytical approach.

9、The academic rigor of the figures and tables in the current manuscript needs significant improvement. Several figures and tables appear to have rough compositional details, primarily manifested in low resolution, inconsistent fonts, and excessive white space. A high-quality academic journal article adheres to certain standards regarding composition. I suggest the author completely reconstruct the existing figures and tables and specifically improve the image quality. I recommend the author read the following high-quality paper and refer to its compositional details to improve the images in the manuscript accordingly. Specifically, the link to the paper is as follows: https://doi.org/10.1038/s41586-025-09922-y

 

We thank the reviewer for this helpful observation. We acknowledge that the quality of some figures and tables in the submitted manuscript may have been compromised because, for purposes of the review draft, the figures and tables were embedded directly within the document.

For the final accepted version, the authors will submit figures and tables as separate high-resolution files, in accordance with the journal’s formatting requirements, to ensure improved clarity, readability, and visual quality. This revised submission process will address issues related to image resolution, formatting, and presentation quality identified during review.

 

 

10、The current manuscript only contains correlation analysis, but some statements show a potential tendency to express causality, which I believe should be approached with greater caution. This is because causal logic analysis employs professional analytical methods and models. However, the author's research in the manuscript is actually a cross-sectional survey, not a causal identification design. I would like to suggest that the author review the entire manuscript and be more cautious in expressing and elaborating on causal logic to avoid exaggeration.

 

We thank the reviewer for this important observation and carefully revised the manuscript to ensure findings are framed as patterns, trends, associations, or plausible contextual explanations rather than causal effects. Throughout the revised manuscript, we replaced deterministic language with more cautious formulations such as “may reflect,” “suggest,” “aligns with,” “is consistent with,” “may help explain,” and “should not be interpreted as.” Several examples include:

1. Coal mining familiarity and geographic exposure (Results Section 3.1).
   Rather than implying direct causation, the manuscript now states that respondents’ familiarity with coal mines was “an expected outcome, given the historical concentration of mining operations…” (Line 159 to 160)  and further notes that exposure “may influence not only awareness of mining activities but also perceptions of coal’s economic and environmental impacts.”  (Line 163 to 168). This wording emphasizes a plausible relationship rather than a causal mechanism.
2. Workforce transitions and coal employment decline (Section 3.1.2).
   The revised text uses cautious interpretive language such as “suggesting that economic instability and changing labor market conditions are significant drivers of workforce transitions” and “may reflect increased recognition of the long-term decline in coal employment…” rather than asserting definitive causal explanations for employment decisions.
3. Environmental concern by gender (Section 3.2.1).
   To avoid causal inference, the revised manuscript states that gendered differences in environmental concern “may reflect variations in occupational exposure, caregiving roles, community engagement, and perceptions of environmental justice issues,” and later notes “perhaps these patterns may help explain stronger concern among female respondents…” These formulations explicitly present contextual explanations rather than causal claims.
4. Environmental concern by location (Section 3.2.2).
   The manuscript avoids deterministic interpretation by stating that rural differences in environmental concern “may reflect differences in environmental attitudes…” and emphasizes that lower concern regarding climate change “does not necessarily imply lower awareness of environmental issues overall.”
5. Environmental concern by age (Section 3.2.3).
   The revised discussion frames findings as observed patterns, noting that younger individuals “may demonstrate stronger concern about climate-related issues,” while age-related differences “may be explained by exposure to educational institutions, digital media, and broader public discussions.” Importantly, the text cautions that lower concern among older respondents “should not be interpreted as a general absence of environmental concern.”
6. Climate change perceptions and coal decline (Section 3.1.3).
   Rather than asserting direct causal relationships, the revised manuscript notes that lower attribution of coal decline to climate change “may reflect political polarization surrounding climate change” and that residents “may acknowledge environmental degradation and pollution concerns yet remain less likely to directly associate climate change with coal decline.”
7. Conclusion language.
   The conclusion was revised to emphasize interpretation and contextual association rather than causality, stating that environmental attitudes are shaped not only by environmental conditions but also by “demographic characteristics, political identities, cultural narratives, and economic realities,” and that policy approaches “may therefore benefit” from being locally informed and context responsive.

We believe these revisions substantially strengthen interpretive precision and ensure that the manuscript consistently reflects the limitations and appropriate interpretation of correlation-based, cross-sectional analysis.

11、As a scientific paper in the form of an investigative study, the manuscript currently has strong practical significance, but its international theoretical contribution is still insufficient. I suggest that the authors consider increasing theoretical contributions, such as providing new evidence on the influence of coal culture identity on environmental perception. Furthermore, the authors could perhaps further clarify the manuscript's specific contributions to policy recommendations.

We thank the reviewer for this thoughtful recommendation. In response, we strengthened the Conclusion to more explicitly articulate the manuscript’s theoretical contribution and policy relevance. Specifically, we added language clarifying how the findings contribute to broader scholarship on energy transition, environmental perception, and community identity, while further specifying the implications for context-sensitive policy design in fossil-fuel-dependent regions. We believe this revision improves the manuscript’s articulation of both its conceptual contribution and practical significance.

Added new line 443 to 448:  Overall, the manuscript presents content with practical research significance. Furthermore, it has policy value regarding environmental transition issues. However, the author needs to fully consider the questions I raised and make significant revisions to the original manuscript to meet the journal's publication requirements. I look forward to receiving the revised version from the author.

 

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

The author has written very well; I have no questions.

Author Response

No comments or reviews by the third reveiwer

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

I would like to thank the author for adopting and revising my suggestions, and I recommend that this manuscript be accepted.