Smallholders Are Not the Same: Under the Hood of Kosovo Agriculture

Round 1

Reviewer 1 Report

Authors need to be more concise and detailed in their presentation, especially in the specification of the models. 

The grammar needs to be checked.

Comments for author File: Comments.pdf

Author Response

Introduction:

1. Most claims by the authors are not supported by the literature and also weak in some instances. E.g., Lines 26-27, “ assumption that all observed units have homogenous production function leads to expectation of homogenous response”.

Added reference to supports this.

Also, check lines 33-40,

Weak claim: In lines 57-58, the authors stated that Kosovo is one of the poorest countries in Europe because the per capita GDP was only Euro 3,746 [3]. However, this claim cannot be entirely true unless it is compared with other nations in Europe. It is therefore suggested to be reframed this statement.

We have added table 1 which compares Kosovo GDP to that of neighbouring countries as well against the EU average, which clearly illustrates this fact.

The claim in lines 58-59 is weak and questionable. Why is agriculture an important sector in Kosovo? Is it because of the 7.2 percent it contributed to GDP in 2018? 7.2% out of 100% in just 2018 is far too small and cannot, therefore, make agriculture an important sector in the economy of Kosovo. The authors can just acknowledge the contribution of agriculture to the economy of Kosovo instead of making a such questionable claim.

The new table 2 compares the contribution of agriculture to GDP to that of neighbouring countries and the EU. Although this contribution is higher in Albania and similar is some other Western Balkans states, it is several times smaller (almost 5) that of the EU as a whole.

Materials and Methods

equations are not completely numbered. See lines 212 and 229.

Corrected

All the variables must be defined. What is g(.) in equation 1, n in the BIC equation (LINES 212)?

g() is a probablility distribution as stated immediately below equation 1.

n is the sample size and this has bee added in.

The sentence (lines 210-211) preceding the BIC equation should be linked to the BIC equation to

read as “A popular criterion in model selection problems is the Bayesian Information Criterion

(BIC) defined as:”

Lines 222-223 is not clear. “The BIC is furthermore approximately equivalent to the popular in information theory Minimum Description Length (MDL) criterion”.

One could support this by an explicit reference:

“Hence the BIC criterion, derived as approximation to log-posterior probability, can also be viewed as a device for (approximate) model choice by minimum description length.” p 236

Hastie, T., Tibshirani, R., & Friedman, J. (2016). The elements of statistical learning: data mining, inference, and prediction. Springer series in statistics. Second Edition. Springer Series in Statistics

It would however be too much , particularly since this is basically a textbook knowledge and it is cursory to what we do. We did nevertheless include a reference in the text

The information on the data is inadequate. What criteria is employed in the data collection?

FADN methodology is well established . It is the standard in European (actually EU) agricultural statistics. We have added a description of the overall approach as well as the Kosovo adaptations to it.

Line 225, “ while all other variables are in monetary terms”. In what currency?

It is in Euros, but it does not really matter in what currency they are expressed since this will not change any results We have clarified this I the text.

What is IC in table 2. Table 2 need to be clearly presented. It is difficult to understand ints current form.

IC is intermediate consumption and is defined in the data description just above table1. As stated in the text,table 2 follows the format of the table 1 (which is the summary statistics for all data, while here these are by cluster, followed by tests of difference in clusters (mean sand distributions).

Typos and Grammar

The sentence in lines 64-65 should be rephrased probably to read as: “In terms of land area, the largest proportion (i.e., 74%) constitutes smaller holdings less than 2 ha, while only 1.7% constitute holdings larger than 10 65 ha [5]. OK

Lines 227-228 should read as “According to [16], the ICL can be expressed as BIC with an additional entropy penalty term as follows:”

There should be a full stop after some statements. Examples are listed below:

Lines 62-64, “According to the Agricultural 62 Household Survey, carried out by the Kosovo Agency of Statistics, in 2018 there were 63 108,108 agricultural holdings with a total land area of 185,130 ha”.

Lines 220-221, “It has been proven that the BIC is consistent and efficient on practical grounds [e.g. 14]”.

Lines 78-86 should be deleted. I guess they are probably some reviewer comments and suggestions.

Not sure about that. It is unclear what these lines are. References to lines in this review are always somewhat off the mark (they are greater in my copy, before editing)

The citations in the text should not be highlight.

Reviewer 2 Report

This study aims to propose the use of underlying functional relationship in order to homogenous groupings be derived. There are some comments that are recommended to be addressed prior to consideration of publication.

1.       The article is too brief and lacks literature references. Please support with the help of evidence from the literature the Lines: 23-28, 33-54, 88-102, 109-169, 172-177, 181-185, 189-218, 225-240, 262-268, 282-290, 294-305,

2.       Please justify with the help of evidence from literature why did you choose the examined variables (Lines: 252-255).

3.       I recommend you not use the First Person Singular or Plural. Usually, the third one is appropriate for a research paper. Example: Present study’s authors merged… (Not: We propose, we need, we merge etc.). Please re-examine the whole range of the text.

4.       Please quote a research background, research significance, purpose, and contribution in the section of introduction. Make clear why the literature needs your study in the field of Smallholder Farming, agriculture, and policy-making through the pandemic and war period.

5.       Regarding the previous remark,  it is also recommended to re-examine the whole range of your paper, taking into account the current situation (pandemic, Russia-Ukraine war) since the data you are using is old (2016). You should, essentially, adapt (with the help of evidence from literature) your paper on how your research –in which the analysis is supported by old data- will contribute to the new situation.

6.       Why the second section is named “Motivation” and not “Literature review”?

7.       Why did you use the finite regression mixture model in order to specify and estimate farm groups with regard to the pre-specified production function? This choice must be supported by the literature and specifically, is an extracted conclusion after an extended literature review (2nd Section) conduction.

8.       After the results section and before the Conclusions one should be added a discussion section where you have to compare the current research results with previous studies' results.

9.       Please add a source for the FADN database and cite it where needed throughout the text. 

10.   In the Conclusions section should be referred the paper’s contribution to the literature, farmers, policies, economy (e.t.c.), the research paper’s limitations (for example you could refer to some of the research difficulties-old data?) and specific suggestions for further research approaches (for example, how future researchers can use your paper in order to conduct further research?). 

Author Response

This study aims to propose the use of underlying functional relationship in order to homogenous groupings be derived. There are some comments that are recommended to be addressed prior to consideration of publication.

1.       The article is too brief and lacks literature references. Please support with the help of evidence from the literature the Lines: 23-28, 33-54, 88-102, 109-169, 172-177, 181-185, 189-218, 225-240, 262-268, 282-290, 294-305,

We have added reference throughout the paper

2.       Please justify with the help of evidence from literature why did you choose the examined variables (Lines: 252-255).

This is standard production function specification in agricultural economics specification. We have added reference use uses the same variables.

3.       I recommend you not use the First Person Singular or Plural. Usually, the third one is appropriate for a research paper. Example: Present study’s authors merged… (Not: We propose, we need, we merge etc.). Please re-examine the whole range of the text.

This is a matter of style. We will leave this judgement to the editors. It is however widely used convention, particularly in the agricultural economics literature, where this paper sits.

Nonetheless we have changed all text into third person

4.       Please quote a research background, research significance, purpose, and contribution in the section of introduction. Make clear why the literature needs your study in the field of Smallholder Farming, agriculture, and policy-making through the pandemic and war period.

This has been better explained

5.       Regarding the previous remark,  it is also recommended to re-examine the whole range of your paper, taking into account the current situation (pandemic, Russia-Ukraine war) since the data you are using is old (2016). You should, essentially, adapt (with the help of evidence from literature) your paper on how your research –in which the analysis is supported by old data- will contribute to the new situation.

There is now discussion about the significance of these findings with regard to external shocks

6.       Why the second section is named “Motivation” and not “Literature review”?

7.       Why did you use the finite regression mixture model in order to specify and estimate farm groups with regard to the pre-specified production function? This choice must be supported by the literature and specifically, is an extracted conclusion after an extended literature review (2nd Section) conduction.

We propose a novel methodology. The reasons for this are presented in the section Motivation, which is in our view appropriately named. It cannot be called literature review, since the relevant literature is essentially methodological, while the previous literature which is in some way cursory to the issue is essentially asking a very different question and making questionable assumptions. Going too deep into these would however distract from the empirical application.

Still we have renamed it to ‘Literature review’ (which in our view is the wrong heading))

8.       After the results section and before the Conclusions one should be added a discussion section where you have to compare the current research results with previous studies' results.

Unfortunately there is no such previous studies . This on ts own is an extremely dangerous statement to make, but we are confident in its veracity. Kosovo is a small country (and even not universally recognised) so there are very few studies regarding its agriculture. The issue in hand here is identification of (homogeneous) groups, which is even less prominent in the literature. As we argue throughout the paper citing literature we object does not add much to the argument we are making. Furthermore the reference style in this particular journal (numbering unlike the author/year format commonly used is social sciences) prioritises quantity over quality in referencing since once cannot easily ‘read’ the text. While this may may sense in technical subjects, it is unhelpful is social sciences.

9.       Please add a source for the FADN database and cite it where needed throughout the text. 

It has been explained and referenced.

10.   In the Conclusions section should be referred the paper’s contribution to the literature, farmers, policies, economy (e.t.c.), the research paper’s limitations (for example you could refer to some of the research difficulties-old data?) and specific suggestions for further research approaches (for example, how future researchers can use your paper in order to conduct further research?). 

Additions to the introduction and conclusions have been made to address this

Round 2

Reviewer 2 Report

It is clear that the authors made a great effort but still there are some comments that are recommended to be addressed.

1. The article is too brief and lacks literature references. Please support with the help of evidence from the literature the Lines: 23-28, 33-54, 88-102, 109-169, 172-177, 181-185, 189-218, 225-240, 262-268, 282-290, 294-305,

We have added reference throughout the paper

Reviewer’s Answer: Still there are some lines that are not supported by the literature. Please re-examine.

2. Why did you use the finite regression mixture model in order to specify and estimate farm groups with regard to the pre-specified production function? This choice must be supported by the literature and specifically, is an extracted conclusion after an extended literature review (2nd Section) conduction.

We propose a novel methodology. The reasons for this are presented in the section Motivation, which is in our view appropriately named. It cannot be called literature review, since the relevant literature is essentially methodological, while the previous literature which is in some way cursory to the issue is essentially asking a very different question and making questionable assumptions. Going too deep into these would however distract from the empirical application.

Still we have renamed it to ‘Literature review’ (which in our view is the wrong heading))

Reviewer’s Answer: If you believe that the title “Literature review'' is wrongly given then you should provide relevant arguments -as in the answer you give above- in the introduction of this section. For example: “In this section the authors are trying (or are motivated) to propose a novel methodology. The reasons for this choice are presented through their point of view”.

3. After the results section and before the Conclusions one should be added a discussion section where you have to compare the current research results with previous studies' results.

Unfortunately there is no such previous studies . This on its own is an extremely dangerous statement to make, but we are confident in its veracity. Kosovo is a small country (and even not universally recognised) so there are very few studies regarding its agriculture. The issue in hand here is identification of (homogeneous) groups, which is even less prominent in the literature. As we argue throughout the paper citing literature we object does not add much to the argument we are making. Furthermore the reference style in this particular journal (numbering unlike the author/year format commonly used is social sciences) prioritises quantity over quality in referencing since once cannot easily ‘read’ the text. While this may may sense in technical subjects, it is unhelpful is social sciences.

Reviewer’s Answer: The effort to provide relevant literature studies concerning other countries (except Kosovo) is recommended.

4. Please add a source for the FADN database and cite it where needed throughout the text.

Reviewer’s Answer: It is recommended to add a relative link.

5. Line: 503. What means “his paper”? Do you mean “This paper”? Line: 517. Please change «yo» to «to». It is recommended to re-examine the whole range of the text because maybe there are more spelling mistakes (for example check again the lines: 530-535).

6. In the Conclusions section should be referred the paper’s contribution to the literature, farmers, policies, economy (e.t.c.), the research paper’s limitations (for example you could refer to some of the research difficulties-old data?) and specific suggestions for further research approaches (for example, how future researchers can use your paper in order to conduct further research?).

Additions to the introduction and conclusions have been made to address this

Reviewer’s Answer: The limitations could be clearer. The same applies to the proposals for further research approaches.

Author Response

Still there are some lines that are not supported by the literature. Please re-examine.

We have re-examined the text and have added additional references

If you believe that the title “Literature review'' is wrongly given then you should provide relevant arguments -as in the answer you give above- in the introduction of this section. For example: “In this section the authors are trying (or are motivated) to propose a novel methodology. The reasons for this choice are presented through their point of view”.

We have now reverted to the motivation heading and have amended in line with the above suggestion.

We have also linked the proposal to the existing literature , in particular the one that have argued for ‘functional’ characterization of farm typologies, which provides a more natural framework for our proposal

The effort to provide relevant literature studies concerning other countries (except Kosovo) is recommended.

The amended paper contains two separate strands of relevant literature: the first is the methodological implementations, with references to small farm clustering applications or applications in European context. The second (as per the answer to the previous comment) refers to the calls for functional justification of farming typologies with regard to the purpose to which they are used. The latter provides natural justification of our approach

Please add a source for the FADN database and cite it where needed throughout the text.

We have now added a link to a source that describes the particula implementation of the FADN database in Kosovo, which although following the EU methodology makes a number of adaptations due to the particular characteristics of Kosovo farms.

The limitations could be clearer. The same applies to the proposals for further research approaches.

We have added another paragraph extending the discussion on the paper limitations and further research