The Role of Education and Income Inequality on Environmental Quality: A Panel Data Analysis of the EKC Hypothesis on OECD Countries
Round 1
Reviewer 1 Report
The authors positively responded to minor issues noted in my previous review and I think the paper is now acceptable for publication. I think Appendix I is a good addition to the paper.
Author Response
Dear Reviewer
thank you very much for your support and for giving us the opportunity to revise the manuscript. The new version really improved from the original one.
Best regards
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The article may be published in the present form
Author Response
Dear Reviewer
thank you very much for your support and for giving us the opportunity to revise the manuscript. The new version really improved from the original one.
Best regards
Author Response File: Author Response.pdf
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
Round 1
Reviewer 1 Report
Review of:
The Role of Education and Income Inequality on Environmental Quality. A Panel Data Analysis of the EKC Hypothesis on OECD Countries
The paper is definitely suitable for the special issue of Sustainability on the environmental Kuznets curve conjecture. The authors seem to have a good grasp of appropriate data sets and research methods, as indicated also by a substantial set of references. The paper is well organized and, aside from its technical content, well written and easy to comprehend. The detailed and knowledgeable description of the EKC models, the data, and the statistical methodologies should give readers confidence in the authors’ expertise. For readers who are not familiar with econometrics, the authors should explain what they mean by “panel data” and “panel data techniques.” I’m familiar with the concept from another disciplinary area but I had to look up the term “panel data.”
My one critique of the model and methods, as an economics outsider, is the use of a quadratic function to indicate the “inverted U shape.” There is nothing holy about the “inverted U,” “concave down”, or the quadratic function as a mathematical model. To me, a quadratic overly restricts possible relationships (e.g., schooling & per capita CO2 emissions). Would it not be fruitful to use a local smoother, spline, or kernel model to open up the flexibility of the model? Concavity could still be assessed but it wouldn’t be as strongly forced on the outcome.
The empirical results in the Tables and Figures are well presented for the most part. I have just a couple recommendations:
• In Tables 8 & 9, provide some kind of symbol, such as a dash, to indicate a lack of statistical significance. Alternatively, use gridlines in the tables so that variables which are not significant (e.g., “Trade openness” in Table 8) are easily identified.
• In Figure 4, though the analyses are carried out in log scale, linear scale values should be used to convey the results clearly. For example, instead of asking readers to comprehend what e1.8 years is, make the scale go from 6 years to 13 years, which is about the range shown.
Figure 4 calls attention to the fact that a Kuznets curve is being simultaneously fit to data from several countries. The smoothness and simplicity of the model curve contrasts with the scatter and apparent dispersion of the data. I suggest identifying the data symbols for one or two of the countries to indicate how closely (or not) the model follows the pattern of change in the data. For example, one country could be selected because it follows the model closely, another because it has the least similarity to the model.
The manuscript could be improved throughout in terms of English grammar and vocabulary but most instances are minor and don’t cause confusion. Here are some places where I recommend changes:
Line 27: taking → including
Line 72: remaining → remainder
Line 109: present → contains
Line 129: product → produces
Line 131: about → from supports →support
Line 200: logistic → logistics
Lines 225 & 241: contributes → contributions
Line 280: contribute → inclusion
Line 489: sever → severe
Line 503: 10.83$ years → 10.83 years
Line 520: consolidating → confirming
Reviewer 2 Report
Notwithstanding the fact that the topic is interesting, I do have substantial and fundamental problem with the paper in regard to the "wild and inconsistent" mix of statististical and econometric methods used.
Just as one example let me allude to the - widely-discussed by now in the EKC literature - problems in relation to using (and if so how) unit root and cointegration based methods. The issues are: (i) if GDP per capita is an integrated process, then its square or cube are not - and hence if one talks about cointegration and all that one has to take this nonlinearity into account. (ii) cross-sectional independence assumptions as necessary for the tests applied - so-called first generation tests - make usefulness of these tests quite limited in some sense; in addition to the fact that these are not applicable really because of the "polynomial nonlinearity" issues.
Given these issues - and some others - I do not see the value added of the paper in the sense that I do not see trustworthy (logically consistent) empirical analysis.
Reviewer 3 Report
The content may be improved with respect to present background.
The research design is not explained in a coherent manner.
The results and discussions are not always consistent with hypotheses (eg. the energy discussion).
Line 136
Please specify the reason you select CO2 as environmental degradation indicator among a series of other possible ones.
Line 430
From where results that in both FE and RE is about renewable sources of energy? What about non-renewable sources?
