Unveiling the Catalytic Role of Digital Trade in China’s Carbon Emission Reduction under the Dual Carbon Policy
Round 1
Reviewer 1 Report
Comments and Suggestions for Authors
REVIEW OF SUBMISSION TO SUSTAINABILITY –– 2993318– “UNVEILING THE CATALYTIC ROLE OF DIGITAL TRADE IN CHINA’S CARBON EMISSION REDUCTION UNDER THE DUAL CARBON POLICY”
Summary of the paper
This paper investigates the impact of digital trading of carbon emissions reduction in the People’s Republic of China. Naturally, there is a maintained assumption that this association is negative. Six hypotheses are mounted. The first hypothesis conjectures that the scale effect, as a mediator of this association, is uncertain. The next two hypotheses relate to production-side variables, as mechanisms via which this association occurs. The second and third hypotheses respectively postulate that industrial structure optimization and technology have a positive mediating effect on this association. The fourth and fifth hypotheses relate to mediating effects of consumption-side variables. The fourth hypothesis conjectures that consumption upgrade has a positive mediating effect. The fifth hypothesis articulates that industry agglomeration has an uncertain effect. The final hypothesis relates to a moderating effect (i.e., variables upon which the strength of the negative association depends). Hypothesis Six articulates that the negative association between carbon emissions and digital trade is stronger in provinces that participated in the low-carbon pilot program.
It appears that Sobel’s (1982) methodology was used to test H1-H5. Generalised Method of Moments (GMM) regressions were used to estimate the baseline and augmented models. The dependent variable was the natural logarithm of carbon emissions (total emissions from consumption of ten types of fossil fuels). For all tests, the independent variable of interest was an index of the level of digital trade, comprising measures of digital innovation, infrastructure environment, technological innovation, innovation environment, digital trade capacity and trade potential. For tests of H1-H3, the mediating variables were stock output and advancement of industrial structure (the ratio of value added of tertiary industry to value added of secondary industry). For tests of H4 and H5, the mediating variables were rural consumer upgrading (rural consumer spending per 10,000 people) and urban consumer upgrading (measured equivalently). One-year lagged natural logarithm of carbon emissions was included as an independent variable, to capture its temporal evolution.
H6 was tested via Ordinary Least Squares (OLS) regressions. The dependent variable was the same as for tests of H1- H5. The digital trading index was an independent variable of interest. There were also independent variables measuring the concentration of employment in the highest technology-related industry, the concentration of employment in the highest productive services industry and a dummy variable flagging provinces that participated in China’s pilot low-carbon policy. Each of these independent variables was interacted with the index of digital technology. H6 would be supported by positive coefficients attaching to the interaction terms.
Data were collected for 30 provinces, over the investigation period was 2013-2021. There were 270 province-year observations. Data were sourced principally from statistical yearbooks.
The estimates of the GMM regressions uphold the maintained assumption of a negative association between carbon emissions and digital trading. The evidence also indicates that all hypothesised mediators (consumption- and production-side) cause the coefficient of digital trade to change. (However, in each set of tests, the coefficient is still negative and significant.) The OLS regressions report significant coefficients attached to all four interaction variables.
The authors conclude that all hypotheses are accepted. Their results indicate that in the People’s Republic of China, digital trade reduces carbon emissions, via mechanisms related to both production and consumption. They also conclude that the strength of this relation depends of industry agglomeration.
Critical review
Literature review
A coherent literature review is contained within the introduction.
Hypothesis development
I have some concerns about the hypothesis development.
The first hypothesis is framed in the null. Limitations of research design can cause a null result. Hence, it would be impossible to interpret results from tests of the first hypothesis as supporting the hypothesis versus being due to research design limitations. My suggestion is simple. The authors should amend the hypothesis to be non-directional. The revised H1 would articulate that there is a scale effect, mediating the association between carbon emissions and digital trading. This criticism applies equally to H5.
H5 should be preceded by a formal hypothesis that there is regional heterogeneity in the impact of digital finance, in reducing carbon emissions.
Methodology
The methodology is sound and befits the hypotheses. I have some criticisms.
Clarify whether Sobel’s (1982) test was used in tests of H1-H3. If so, insert the Z-statistic into the relevant tables.
What are the units of measurement of fixed capital stock, proxying regional output?
Sample selection
The total sample size is 270 province-year observations. (i.e., 30 provinces * nine years). A strength of the paper is that no observations were discarded due to data unavailability. This should be mentioned in the paper.
Data collection
Were the data from the yearbooks hand-collected?
Descriptive statistics
Table 2 is very unclear. What is the variable reported? It appears that this variable has been calculated on a province-year basis. The “mean” for each province seems to refer to the time-series mean, calculated for the relevant province, over the investigation period. These features should be clarified, in the captions.
Figure 2 and the accompanying discussion do not relate to the statistics in Table 2. I suggest that Figure 2 should be deleted. The discussion should be replaced with discussion of the content of Table 2.
Discuss the univariate statistics reported in Table 4.
Present a bivariate correlation matrix and discuss the statistics therein.
Empirical results
My key concern relates to Table 5. The table contains an OLS estimation of a model where an independent variable is the one-year lag of the dependent variable. This is econometrically impossible. This would violate the assumption that the independent variables are non-stochastic. I suggest deleting the first column of results in Table 5.
The results in Table 6 should be discussed (but not reported) as sensitivity analyses.
Why are there 240 observations for each stratum in Table 7? There are 270 observations for the entire sample. Hence, surely the observations for the strata should sum to 270.
Presentation
The presentation generally reflects a high standard.
The authors should reduce usage of colloquial language. Examples include “hot research direction”, “on the one hand”, “on the other hand”, “among them”, “at the same time”, “in line with”, “rolled out” and “to sum up”.
The abbreviation “DID”, on p.6 should be written in full. This comment applies to the abbreviation “IPCC” on p.8.
Reference in this report, not in the paper
Sobel, M., 1982, “Asymptotic Confidence Intervals for Indirect Effects in Structural Equation Models”, Sociological Methodology 13, 290-312. http://www.jstor.org/stable/270723
Author Response
Please see the attachment. Thanks.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsGeneral issues:
The authors pose two research questions:
(1) Is China's digital trade level conducive to carbon emission reduction? What is the specific mechanism of action? [lines: 117-118]
(2) Do different types of industrial agglomeration and low-carbon pilot policies have a moderating effect on the impact of digital trade on carbon emission reduction? [lines: 118-120]
The research attempts to comprehensively explore the carbon emission reduction mechanism of China's digital trade level by combining the analysis framework of China's production side and consumption side [lines: 121-123].
The authors have created a system of indicators to measure the level of regional digital trade from the aspects of digital innovation, factor endowment, infrastructure, innovation environment, trade potential [lines 124-126]. The research sample covered 30 Chinese provinces in 2013-2021 [lines: 126-127]. The mechanism of digital trade on carbon emission reduction was analyzed [lines: 129-133]. An empirical analysis was made of the impact of China's digital trade level on carbon emission reduction and the regulatory role of industrial agglomeration and low-carbon pilot policy through [lines: 135-137].
The following hypotheses were put forward:
H1: In the process of the impact of digital trade on carbon emission reduction, the scale effect is uncertain [lines: 169-170].
H2: In the process of the influence of digital trade on carbon emission reduction, the structural effect is positive [lines: 190-192].
H3: In the process of the impact of digital trade on carbon emission reduction, the technology effect is positive [lines: 208-210].
H4: In the process of the impact of digital trade on carbon emission reduction, the effect of consumption upgrading is positive [lines: 233-235].
H5: In the process of the impact of digital trade on carbon emission reduction, the regulatory effect of industrial agglomeration is uncertain [lines: 256-258].
H6: In the process of the impact of digital trade on carbon emission reduction, the regulatory effect of low-carbon pilot policies is positive [lines: 274-276].
The research adopts systematic GMM method to conduct empirical research, which can effectively solve the endogenous problem of digital trade level [lines: 429-430].
Quantitative regression and variable substitution were also used to test the robustness of measurement results [lines 459-460].
Detailed issues:
Line 281: Shouldn't it be "Baseline Model"?
Line 291: Be consistent and start subsection titles with a capital letter.
Line 311: I think it's "it" and not "It".
Lines 322-323: CE is “energy consumption carbon emission” or “carbon emissions”?
Line 390: No space between et.'s and research, why the capital letter "The"?
Lines 330; 425: Is the 'Digital Trade Level' variable DT or DTL? It is marked differently in the text. Needs to be standardised.
Summary:
The article is interesting, but it only refers to the perspective of China, limited to 4 provinces. There is also no reference to the situation in other countries or regions. There is no mention of the European Union, whose policy is aimed at significantly reducing the carbon footprint and supporting digital trade. The authors put forward hypotheses and do not indicate whether they have been verified. The discussion part of the article is very poor and there is no reference between the obtained results and the results of other authors. No limitations of the study were indicated. The article uses literature mainly by Chinese authors. The question is whether measuring carbon emissions based only on the consumption of energy sources is sufficient. Wouldn't a better solution be to measure your carbon footprint? Another issue is CO2 emissions. It is assumed that the entire consumption of energy sources causes CO2 emissions into the atmosphere? The article needs refinement.
Author Response
Please see the attachment. Thanks.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsDear authors,
The topic discussed in your manuscript is relevant and in line with the current trends in both the global and Chinese economies. The manuscript contains a fairly comprehensive descriptive section that details the theoretical and overall aspects of the topic. However, we have some comments:
1. We recommend that you present the research methodology more explicitly, showing the connections between formulas. At present, the methodology and calculations presented in the manuscript lack a clear logical structure.
2. The presented calculations and analysis based on statistical data have a wide scope and are indirect. A clear relationship between the development of the digital economy and the carbon footprint has not been demonstrated. Factors influencing the carbon footprint, such as the creation of microchips, processors, and other necessary infrastructure to support digital services as well as the generation of electricity for these services, are not taken into account.
3. The conclusions about a clear correlation between digitalization and the carbon footprint presented in the report are based on statistical analysis using indirect data. To verify this hypothesis, a specific example would need to be provided based on calculations of the carbon footprint before and after digitalization. Consider, for example, a case study based on water transport, which has a wide range of research available and ambiguous conclusions.
4. There are spelling and punctuation errors.
Author Response
Please see the attachment. Thanks.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsREVIEW OF SUBMISSION TO SUSTAINABILITY –– 2993318R1– “UNVEILING THE CATALYTIC ROLE OF DIGITAL TRADE IN CHINA’S CARBON EMISSION REDUCTION UNDER THE DUAL CARBON POLICY”
Preliminary notes
1. I have not reproduced my summary of this paper. There have not been sufficient changes to warrant this.
2. The authors have attended to some of the concerns broached in my previous report. There are other concerns that they have not addressed. I have repeated the latter concerns in the current report.
Critical review
Hypothesis development
1. The wording of H7, in its current form, is confusing. I suggest changing the verbiage to the following.
“H7 – Companies from provinces that participated in the pilot low-carbon policies program experienced a greater reduction in carbon emissions, due to digital trading, than companies in other provinces.”
Methodology
2. It seems that haggi,t is defined as Maxj [Sj,I,t / Si,t], analogously to caggi,t. Is this correct? If so, the definition of haggi,t should be amended accordingly.
Data collection
3. In their rejoinder to my previous report, the authors explained that data were hand-collected from the yearbooks. This explanation should be inserted into the section of the paper that covers data collection.
Descriptive statistics
4. I regard this dimension of the paper as lacking. The authors do not provide discussion of the univariate statistics in Tables 2 and 3. Similarly, they do not present a correlation matrix. In their rejoinder, they support these decisions by claiming that they are space restricted. I do not accept this explanation.
My suggestion is to insert discussion of the univariate descriptive statistics and a bivariate correlation matrix (including discussion of the significant correlations). To reduce space, it would be suitable to delete Table 6. These results are sensitivity analyses. Hence, it would be feasible to summarily discuss them, without reporting them. Since I am in disagreement with the authors, I request that the editors adjudicate, regarding this issue.
Empirical results
5. I do not accept that the results in Table 9 support H5. The coefficient of DT in the augmented model exceeds (i.e., is less negative than) the counterpart coefficient in the baseline model (reported in Table 5). This suggests that consumption upgrading attenuates (rather than enhances) the impact of digital trading on reduction of carbon emissions.
I suggest the authors modify the write-up accordingly. The results suggest that production factors (but not consumption upgrading) enhance the positive impact of digital trading on reducing carbon emissions.
6. I do not accept the claim that the results in Table 10 support H6. To support H6, the coefficients of hagg*DT, paggDT and cagg*DT should all be negative. These coefficients are positive.
I suggest the authors modify the write-up accordingly. The results suggest that the impact of digital trading on reducing carbon emissions is greater for companies from provinces involved in the pilot carbon emissions reduction program. However, the results to do not support the claim that this impact of digital trading is positively moderated by industry agglomeration.
Conclusions
7. If the authors agree with my revised interpretation of the results in Tables 9 and 10, they should revise the conclusions accordingly.
Presentation
8. Delete the word “studying” on p.2. (i.e. the sentence should read “So does DT help reduce CE?”)
9. There are two examples of shorthand. The word “etc.” should be deleted on p.3). “Research and development” should be written in full.
Author Response
Please see the attachment. Thank you.
Author Response File: Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsGeneral issues:
The authors introduced most of the changes proposed in the first review and also addressed questionable issues.
The hypotheses were reformulated:
Hypothesis 1 (H1): There is a scale effect in the impact of DT on CE reduction, which plays an intermediary role in CE and may have a positive or negative impact [lines 160-162].
Hypothesis 4 (H4): DT exhibits regional heterogeneity in its impact on CE, with different regions showing varying impacts due to regional differences [lines 217-219].
Hypothesis 5 (H5): In the process of the influence of DT on CE reduction, the effect of consumption upgrading is positive [lines 242-244].
Hypothesis 6 (H6): In the process where DT impacts CE, industrial agglomeration has a positive moderating effect [lines 263-265].
Hypothesis 7 (H7): During the DT's impact on CE reduction process, the regulatory effect of low-carbon pilot policies is positive [lines 281-283].
Detailed issues:
Line 120: Missing period at the end of the sentence.
Line 153: It seems that "two paths" should be followed by a dot, not a colon.
Line 193: Correct subsection number to 2.1.2.
Line 422: Correct punctuation in the case of OT, lowercase the expression after the bracket, similarly after the comma, remove the dot before the bracket.
Line 431: Correct formatting in table 5.
Line 492: Improve formatting, unify (different 4.1 and 4.2 formatting).
Line 497: No call to table 7.
Line 565: Correct the beginning of the paragraph.
Line 554: Make a clear reference to table 9. Without the reference, it is not known which table columns are in question.
Line 780: Standardize the literature, other items do not mark pages as "pp".
Line 794-795: "pp 363-+"??
Line 797: "China?s" or "China's".
Summary:
The authors introduced most of the changes proposed in the first review and also addressed questionable issues. Some changes have been made to the hypotheses, which now appear to be more clear and transparent. Reference was made to the European Union's ecological policy, although to a very limited extent, even though the European Union and European countries are pioneers in implementing solutions that reduce the negative impact of human activity on the environment. The verification of individual hypotheses was clearly indicated, which was missing in the original version.
Author Response
Please see the attachment. Thank you.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsCorrections accepted
Author Response
Please see the attachment. Thank you.
Author Response File: Author Response.pdf
Round 3
Reviewer 1 Report
Comments and Suggestions for AuthorsREVIEW OF SUBMISSION TO SUSTAINABILITY –– 2993318R2– “UNVEILING THE CATALYTIC ROLE OF DIGITAL TRADE IN CHINA’S CARBON EMISSION REDUCTION UNDER THE DUAL CARBON POLICY”
Preliminary notes
1. I have not reproduced my summary of this paper. There have not been sufficient changes to warrant this.
2. The authors have attended to some of the concerns broached in my previous report. There are other concerns that they have not addressed. I have repeated the latter concerns in the current report.
Critical review
Introduction
1. A within-text reference should be inserted into the first paragraph.
Methodology
2. Discussion of Sobel’s test (1982) should be in the methodology section, within the discussion of the mediation tests, rather than in the hypothesis development section.
3. There should be a within-text reference in the first paragraph of S3.2.1.
4. It seems that haggi,t is defined as Maxj [Sj,I,t / Si,t], analogously to caggi,t. Is this correct? If so, the definition of haggi,t should be amended accordingly.
5. The authors claim that the purpose of including ln (CEi,t-1) as an independent variable is to control for endogeneity. I do not accept this explanation. The purpose of including lagged observation of the dependent variable, as an independent variable, is to capture the temporal evolution of the dependent variable. Hence, Ordinary Least Squares is not appropriate due to violation of the assumption that the independent variable is non-stochastic (Akbar, Poletti-Hughes, El-Faitouri and Shah, 2016). I suggest the authors amend the discussion accordingly.
Descriptive statistics
6. There is no discussion of the univariate descriptive statistics reported in Table 4. I suggest the authors insert a discussion of the nature, focusing on economic interpretation of the statistics.
7. The bivariate correlations reported in Table 5 are most appropriate. However, the accompanying discussion merely re-states the significant correlations. I suggest that the authors delete the present discussion. They should insert a discussion with economic interpretation of the significant correlations.
8. The discussion of the VIF factors (unreported) should be relegated to a footnote. The coefficients of the independent variables of interest are generally significant (although not always in the direction anticipated). Hence, it seems unlikely that multicollinearity is a concern.
Empirical results
9. Discussion of the sensitivity analyses (quantile regressions) belongs in the results section, rather than the conclusions.
Presentation
10. There are still some instances of colloquial language. Examples include “along with” (rather than “in conjunction with”), “brought about” (rather than “caused”), “greener purchasing decisions” (rather than “more environmentally responsible purchasing decisions”) and “alongside” (rather than “in conjunction with”)
11. There two examples of incorrect usage. The expression “formed the theory” on p.1 should be replaced with “based on the theory”. There is no word “researches”. The word “research” is both singular and plural.
Reference, cited in my report but not the paper
Akbar, S., J. Poletti-Hughes, R. El-Faitouri and Z. Shah, 2016, “More on the Relationship between Corporate Governance and Firm Performance in the UK: Evidence from the Application of Generalised Method of Moments Estimation”, Research in International Business and Finance 38, 417-429. https://doi.org/10.1016/j.ribaf.2016.03.009
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
