Review Reports

Round 1

Reviewer 1 Report

 

In order to improve the manuscript, I suggest that minor corrections are made. As such, please revise the following errors in: lines 162-163 (for a better understanding of the presented reason), 167-168 (to improve the clarity of the idea – I suggest that a better phrase may be: The two-step procedure still has drawbacks ....), line 169 (probably, there is a missing word - ”is”), line 211 (the parameter should be named ”b”), line 247 (The Tobit regression model refers to a set of ....), line 250 (However, recently studies ....), line 482 (I suggest to delete the words ”from overuses”, as it is confusing), line 689 (Factors Influencing on Environmental...), line 705 (Moreover, the positive coefficient of the experience on EE ...), lines 859-860 (I suggest rephrasing the sentence, for a better understanding).   

 

Author Response

Thank you very much for your good comments and suggestions. All your comments are taken in the revised manuscript. Please kindly see all the responses in the attached file. 

Author Response File: Author Response.docx

Reviewer 2 Report

My comments from the previous review were not taken into account.

Comments attached.

Comments for author File: Comments.docx

Author Response

Dear Reviewer 2,

Thank you very much for your good comments and suggestions. All your comments are taken in the revised manuscript. Please kindly see all the responses in the attached file.

Author Response File: Author Response.docx

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

Referee report on manuscript agriculture-589574:

“Evaluation of Environmental Efficiency of Edible Canna Production in Vietnam”

This paper addresses the issue of technical and environmental efficiencies in edible canna production in Vietnam. The authors motivate their study by underlying the important role played by this food crop in Vietnamese agriculture. Data were collected in the main production area in Vietnam. A two-step econometric procedure is then applied. Edible canna producers' technical and environmental inefficiencies are estimated by estimating a Translog stochastic production frontier in a first step.  The estimated values of inefficiencies are regressed on a set of socio-economic characteristics of producers by using a Tobit type of modeling in a second step.

My main comment on this paper is about the methodology. I strongly encourage the authors to read recent developments in the literature on efficiency analysis (see, for instance, Parmeter and Kumbhakar, 2014). Reading the book by Coelli et al. (2005) is certainly necessary as an introduction to the literature on stochastic frontier analysis but it is not enough. For instance, two-step procedures have been used for a long time when measuring inefficiencies and assessing their determinants, but with authors generally forgetting to take into account estimation errors made in the first step when estimated inefficiencies are regressed in a second step (see below). Censored (Tobit) regression has also been criticized, truncated normal regression being a more appropriate choice (see technical appendix in Simar and Wilson, 2007). One step estimation procedures have also been proposed to take into account determinants of technical efficiency in stochastic frontier analysis a long time ago (see Chapter 7 in Kumbhakar and Lovell, 2003). They can now be easily implemented using Stata or some free packages in R. I am not aware of such developments when dealing with environmental efficiency but I think they exist. Authors should look for such developments in Journal of Productivity Analysis. The paper will gain value if its authors take into account these more or less recent developments.

My second comment is about the motivation of the paper. It would have been interesting to know if the edible canna production is really a polluting activity. A discussion with data on agricultural sources of pollution, partly due to the edible canna production, in the studied Vietnamese communes is missing in the paper.

Other comments:

Page 2, end of introduction: Here, it is usual to give an overview of the paper’s organization. Page 3, lines 116-121: the statistical distributions for u_i and v_i are assumed distributions. Page 3, line 120: one-sided instead of one-side Page 3, lines 120-121: this sentence is incomprehensible. Page 3, lines 132-133: the authors need to develop a little more why the choice of a Translog production function is more appropriate to analyze the environmental efficiency than a simple Cobb-Douglas function. Page 4, lines 139-140: Wat is the meaning of inefficiently in this sentence? Page 4, Eq. (5): Æ usually means empty space in mathematics. Use f instead of Æ. Page 4, Eq. 8: lnZ₁ + lnZ₁ = 2 lnZ₁. Same for lnZ₂ + lnZ₂. Page 4 and 5: These calculations are boring and could be given as an appendix. Pages 3-5: Lot of room is given to the calculation of environmental inefficiency but nothing is said about technical inefficiency computation. Page 6, line 188: Eq. (14) instead of Eq. (15). Pages 6 and following: socioeconomics, socioeconomic variables, socio-economic characteristics… The reader is lost. Page 8, line 252: adopted instead of adapted. Page 8, line 255: present instead of presented. Page 8, second paragraph: Measures of elasticities are local ones as they are computed as partial derivative of output (in logarithms) with respect to input (also in logarithms). So they cannot be interpreted for very large variation in input quantity such as 100%. Page 8, same paragraph: The interpretation of the negative sign of output elasticity with respect to labor must be clarified. Authors should pay attention to the fact that they have chosen labor cost instead of a physical measure of labor (number of workers or worked hours) as is usually done by works estimating production frontiers. For instance, is price of labor input the same in all the communes? Page 10, Table 7: the authors must provide standard errors instead of p-values, the interpretation of which is difficult because it is not clear whether their calculation took into account the errors of estimation made in the first step when calculating inefficiencies. Bootstrapping techniques can be used to provide corrected standard errors in the second step. See Murphy and Topel (Journal of Business and Economic Statistics, 2002). Page 10, lines 329-331: This first point is false. Only the appropriateness of the Translog function is checked but not tested as the tested null hypothesis corresponds to the choice of a Cobb-Douglas function. Existence of technical efficiency is also not tested.

References

Kumbhakar, S .C. , and C. A. K. Lovell (2003). Stochastic Frontier Analysis.  Cambridge University Press.

Murphy, K.M., Topel, R.H. (2002). Estimation and inference in two-step econometric models. Journal of Business and Economic Statistics 20: 88-97.

Parmeter, C.F., and S. C. Kumbhakar (2014). Efficiency Analysis: A Primer on recent Advances. Foundations and Trends in Econometrics, Vol. 7, n° 3-4.

Simar, L., and P. Wilson (2007). Estimation and Inference in two-stage, semiparametric Models of Production Processes. Journal of Econometrics, 136:31-64.

Author Response

We are grateful to reviewer #1 for the critical comments and useful suggestions that have helped us to improve our paper considerably. As indicated in the responses that follow, we have taken all these comments and suggestions into account in the revised version of our paper. All revisions are highlighted in red in the revised manuscript. Please see in the attached file.

Author Response File: Author Response.docx

Reviewer 2 Report

The topic of the manuscript is interesting and the authors had a good idea for a research project. The subject is relevant, aim, range and results were clearly defined and demonstrate a good scientific knowledge of the issues being discussed. The work contains appropriate testing methods and analyses of results. The authors have presented the results of original research. However, the paper contains several inconsistencies which should be explained and resolved. The comments are marked in yellow in the manuscript.

Comments for author File: Comments.pdf

Author Response

We are grateful to reviewer #2 for the critical comments and useful suggestions that have helped us to improve our paper considerably. As indicated in the responses that follow, we have taken all these comments and suggestions into account in the revised version of this manuscript. All revisions are highlighted in red in the revised manuscript. Please in the attached file.

Author Response File: Author Response.docx

Reviewer 3 Report

It's an interesting manuscript, as the authors aim to assess the environmental and technical efficiency of edible canna production in two districts of the Bac Kan province from Vietnam. As such, the article addressess the influence of some input factors on canna production in the analyzed regions, the obtained results being of use for the local authorities and communities involved in designing policies to reduce poverty.   

I suggest authors to make some minor corrections, that are present in: line 120 (the referrence to independent and identically distributed random variables – i.i.d. random variables), 124 (word ”effecting” is misleading), formula 4 (the question mark should be a minus sign), line 163 (the parameter should be named ”b”), line 179 (I suggest, for a better understanding and consequent notation, to specify that X₃ = Z₁ and X₄ = Z₂) and line 188 (I suggest ”as follows”, instead of ”as follow”).

Author Response

We are grateful to reviewer #3 for the critical comments and useful suggestions that have helped us to improve our paper considerably. As indicated in the responses that follow, we have taken all these comments and suggestions into account in the revised version of our paper. All revisions are highlighted in red in the revised manuscript. Please see in the attached file.

Author Response File: Author Response.docx

Reviewer 4 Report

Using Stochastic Frontier Analysis this paper measures the environmental efficiency for a sample of Vietnamese edible canna farms. They augment a conventional efficiency framework by incorporating nitrogen quantities and phosphorus quantities as additional inputs. The authors also perform a second stage Tobit regression to explain the differences in environmental efficiency. Although the subject is interesting, I have major concerns about this paper, which I explain below.

 

Concerns

1/ The authors only look at the nitrogen and phosphorus inputs, whereas the corresponding surpluses would be much more valuable in environmental terms. I thus suggest to focus on the surplus rather than the input. In the suggested approach, the authors should be careful in the modeling of the surpluses. Coelli et al. (2007) show that treating nitrogen surplus as an input violates the material balance condition. Murty et al. (2012) show that such an approach implies that the trade-off between a pollution-generating input (e.g., fertilizer) and a pollutant (e.g., nitrogen surplus) is unrealistically assumed to be non-positive.

2/ The regularity conditions (first- and second-order conditions) of the translog function are not checked. Are output non-decreasing in the inputs? Is there concavity in the inputs and convexity in the outputs? Very often these are violated for a translog function. See Sauer et al. (2006).

3/ The explanatory variables are regressed in a second stage. However, this method is problematic. Wang and Schmidt (2002) show that this method yields biased estimates (and provide a STATA code on their website to deal with this problem using a one-step procedure instead of a two-step procedure).

References

Coelli, T., Lauwers, L., Van Huylenbroeck, G. (2007). Environmental efficiency measurement and the materials balance condition. Journal of Productivity Analysis 28: 3-12.

Murty, S., Russell, R. R., Levkoff, S. B. (2012). On modeling pollution-generating technologies. Journal of Environmental Economics and Management 64: 117-135.

Sauer, J., Frohberg, K., Hockmann, H. (2006). Stochastic efficiency measurement: The curse of theoretical consistency. Journal of Applied Economics 9: 139-165.

Wang, H.-J. and Schmidt, P. (2002). One-Step and Two-Step Estimation of the Effects of Exogenous Variables on Technical Efficiency Levels. Journal of Productivity Analysis 18: 129-144.

Author Response

We are grateful to reviewer #4 for the critical comments and useful suggestions that have helped us to improve our paper considerably. As indicated in the responses that follow, we have taken all these comments and suggestions into account in the revised version of our paper. All revisions are highlighted in red in the revised manuscript. Please see in the attached file.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

I have read carefully the authors' responses to my comments on their paper. I do not find them satisfactory enough to change my opinion on paper. I asked them to update their approach to take into account recent contributions in the field of estimation of production frontiers and the study of the determinants of inefficiency, whether technical or environmental. This should have resulted in the in-depth review of the paper, which is not the case here. I strongly encourage the authors to read the recent literature and especially that published in specialized journals in the analysis of efficiency and productivity such as Journal of Productivity Analysis. Many recent approaches can be implemented using the R software which is free. Sampling error in two-step procedure is an important issue that cannot be ignored. A little thought is sometimes needed before applying what so many others have done, even though their work has been published. Some practices such as the use of Tobit models that have been standard for years when studying inefficiency determinants have recently been questioned. What to say when the authors do not know how to answer the question about the estimation of technical inefficiencies? No doubt Frontier's documentation gives an indication as to how these are estimated. It is then interesting to know if these estimates are unbiased, convergent and so on.

In conclusion, the authors have not achieved the requested revisions, and as a result I am not in favor of publishing the paper as it stands.

Reviewer 4 Report

My remaining remarks concern comments 1 and 2:

Comment 1: The authors argue that "Knowing that evaluating environmental efficiency based on surplus is a new measurement, yet according to Coelli et al. (2007), the environmental efficiency is calculated by the ratio of minimum nutrients over observed nutrients in Data envelopment analysis (DEA) method (cost-minimizing DEA program)." => this is misleading, as Coelli et al. (2007) do focus on the surplus. They indeed use inputs as choice variables, but this does not mean that the output side is not accounted for. Consider for instance the analogous cost-minimisation principle: it still takes into account outputs, even though inputs rather than outputs are choice variables. Comment 2: The authors claim that "The results of Monotonicity condition are presented in Table 1. Results show that all inputs satisfy Monotonicity or the first-order condition, implying that output is non-decreasing in all inputs respectively." However, looking at the corresponding table, we see that the monotonicity condition only holds on average rather than for all observations. The authors should mention the number of violations per monotonicity condition.