Research on the Impact of the Industrial Cluster Effect on the Profits of New Energy Enterprises in China: Based on the Moran’s I Index and the Fixed-Effect Panel Stochastic Frontier Model

Round 1

Reviewer 1 Report

Dear Authors,

I found your paper interesting and novel in methods. In the next paragraphs, I list some shortcomings that authors should amend in order to publish the paper.
1. The main shortcoming - you use a sample of 31 enterprises. Can you explain in the paper why did you take the exact sample? Or, if the reason is not good enough, I suggest taking a bigger sample, as the population is much bigger (you state there are nearly 200 listed companies in the new energy concept), or even the whole population.

2. As there is no explanation about sample size, I doubt your conclusion. After dealing with the first comment please revise the conclusion. If there is a good reason for this exact sample, please state in the conclusion that they are a generalization.

3. In my opinion, the paper will be more interesting if you make the literature review a bit shorter. 

4. In many places you cite authors by their first name, and that is not in line with the references guide and common practice where the last name should be cited. Please check all the citations in the paper.

5. The sentence in lines 232-234 is unclear (first sentence in the last paragraph of the literature review).

6. You have two tables 1.

7. In part 4.3. you start with the sentence “Observing the WALD test values…” but they are nowhere in the paper. This whole part is filled with such ambiguities (e.g. the coefficient of sale cost, it is unclear). I suggest you emphasize when something is in one of the tables, and explain in more detail when it is not.

8. The first sentence of the conclusion should be deleted.

9. Add in the conclusion what are the limitations of the study.

10. References are incomplete, the author's initials are mostly missing (2, 3, 5, 6, 7, 9, 16, 17, 20, 23, 24, 28, 29, 30, 31, 32, 33, 35, 36).

Due to all of the above mentioned, I recommend your paper for publishing after correction for the above comments.

Author Response

We do appreciate your work and treasured comments on our manuscript. We have edited our paper point by point according to your review report.

Please see the attachment for more details, thanks.

Author Response File: Author Response.pdf

Reviewer 2 Report

The article tries to find evidence of the existence of some kind of positive (or negative) effect between the formation of Energy Industrial Clusters on firm performance (total profit).

To achieve this objective, an interesting and novel contribution is made, introducing in a Fixed-effect Panel Stochastic Frontier Model (which tries to analyze the existence of heterogeneous factors that reduce the efficiency of individual companies see Green 2005) as a possible factor of efficiency (or positive externalities, spillover effect, or economies of scales) Local Moran'I (as an indicator of local spatial dependence between the profits of a company and the weighted mean profit of the rest of the companies analyzed).

Although the idea is original and interesting, I have a number of doubts that prevent me from recommending the publication of this article in its present form.

First, total profit is used as an indicator of output. Perhaps if production were measured directly (e.g. through operating income) the effects would be different. In any case, at no point in the article is the choice of profit as an indicator of output justified.

Secondly, my main reservation concerns the use of the Local  Moran’ I as an indicator of efficiency factor due to belonging to an industrial cluster. The paper identifies industrial cluster membership with the existence of a local spatial dependence (measured by the local Moran's I). It seems that the authors identify regional industrial cluster membership with the existence of a positive correlation between the total profits of a firm and the profits of firms in its immediate vicinity. And for this they use a weight matrix where the spatial importance matrix is calculated as the inverse of the distance between firm locations. I think that the membership to a regional cluster must not be measured indirectly by the spatial correlation in profits but rather (by definition of geographical cluster) by the proximity between the firms.

Third, in any case, the calculation of Moran's "I" depends crucially on the choice of the spatial weights matrix. The main weakness of spatial dependence models lies in the fact that there is no single way to construct these spatial weights matrices

Third, if we are talking about geographic clusters (as the authors seem to be using), this spatial dependence on benefits should be limited to a maximum distance to be considered as originating from cluster membership. For example, a spatial dependence limited to around 5 km or 10 km ..... So that dependence between two widely separated firms could not be considered to have any spatial effect, since they do not belong to the same cluster.

Fourth, in my opinion, the cluster concept is a concept of local grouping of companies, and not so much of positive dependence between the business benefits of "nearby" companies.  The authors show the results of 12 companies that seem to belong to the same cluster (and they also find no spatial relationship in their profits). How do they determine that these companies do in fact belong to the same cluster?, it is by geographical proximity, isn't it? Does it mean that the remaining 20 companies are part of another cluster?, are they disseminated throughout the territory?, what is the clustering of the 30 sampled companies?

Perhaps a map of the spatial distribution of firms (not of firms' profits, but of the spatial distribution of firms on a map) could help to understand whether or not there are clusters or concentration of firms (or if, on the contrary, industrial energy firms are scattered throughout the geography without forming clusters). In fact, in the regression model it is said that a locator variable (locator factors) is included, but nothing is said about what exactly is this measure of location. Possibly these locator factors are related to the fact of belonging to a regional cluster. Then this variable should be the one that would measure the effect on the efficiency of belonging to a cluster, and not the local Moran's I.

In summary, without spatial concentration of companies, it is not possible to speak of a regional cluster of companies. That is why I insist that I am not entirely sure that spatial dependence on profits is capturing the effect of cluster membership. Thus, I believe that the use of spatial dependence in profits as an indicator of cluster membership should be better justified. Furthermore, it should also be better explained and substantiated why the mere existence of negative local effects when using distances to "all firms" is considered to be a good measure of cluster effects. I can not see the connection between local dependence in profits among the 32 companies analyzed with the existence of inefficiencies as a consequence of clusters. These are the points that in my opinion need to be improved before recommending the paper for its publication.

 

Minor comments:

 

A revision of the English language should be made as far as possible (sometimes the word effect is repeated twice in the same sentence, line 58 or line 66).

 

The introduction and the literature review can be rewritten to summarize the contribution and the real objective of the article, to test empirically whether or not there are positive effects of Industrial cluster formation.

I say this because much emphasis is placed on highlighting the positive effects of the formation of industrial clusters (by positive externalities, spillover effect, or economies of scales ), and it is even recommended that cluster formation policies be strengthened, but subsequently the effects that are found are negative. This would recommend the suppression of this type of cluster formation policies for regional development since the evidence shows that the effects of these clusters are negative. The approach in the introduction and the literature review section bets on a high relevance of industrial cluster formation, but the empirical evidence found actually shows that clusters actually harm the profit of firms.

The first paragraph of the conclusions section is unnecessary (given the discussion of the model, I do believe the conclusions section is relevant).

I end with a question that may be outside the possibilities of the paper (and is only a suggestion). Considering that the authors are using concepts of spatial dependency, why don't they directly use some sort of spatio-temporal model to analyze the effects of nearby firm profits on firm's profits (introducing Wyit as an additional inefficiency factor in the model), controlling for another series of factors (for example belonging to the same cluster)?

Author Response

My appreciaion is beyond description for your comprehensive and valueble comments on our manuscript. We have edited our paper according to your review report point by point.

Please see the attachment for more details and don't hesitate to contact us.

Thanks.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

I would like to congratulate the authors for the revision made to the first version of their paper. In fact, the decision to limit the analysis only to companies belonging to a well know cluster has radically changed the results. Now, the spatial correlation is positive in almost all years, and the proximity to other companies in the same cluster does have a positive effect on profits. This result seems to be in line with the logic of the contributions that support the formation of industrial clusters (positive externalities, spillover effect, cooperation, and economies of scales).

My congratulations, I consider that all my comments to the first version of the paper have been addressed, and therefore, in my opinion, the article should be published in Sustainability