Submit to this Journal Review for this Journal Propose a Special Issue

Article Menu

Share Help Cite Discuss in SciProfiles

Open AccessArticle

Peer-Review Record

Smallholders’ Preferences for Different Contract Farming Models: Empirical Evidence from Sustainable Certified Coffee Production in Vietnam

Sustainability 2019, 11(14), 3799; https://doi.org/10.3390/su11143799

by Nguyen Hung Anh^1,*, Wolfgang Bokelmann¹, Ngo Thi Thuan², Do Thi Nga³ and Nguyen Van Minh⁴

Reviewer 1: Anonymous

Reviewer 2: Anonymous

Reviewer 3: Anonymous

Sustainability 2019, 11(14), 3799; https://doi.org/10.3390/su11143799

Submission received: 28 May 2019 / Revised: 30 June 2019 / Accepted: 8 July 2019 / Published: 11 July 2019

(This article belongs to the Section Economic and Business Aspects of Sustainability)

Round 1

Reviewer 1 Report

In my opinion this paper is interesting and seems quite well prepared. However after a detailed reading, I have one methodological doubt - although it may be due to my misunderstanding of certain issues. In fact I am not able to understand how a given farm has been assigned to specified contract farming model (strategy). In line 181 we can read: „Pre-test survey has revealed three different contract farming models, which are informal model (InforM), intermediary model (InterM), and nucleus estate model (NEM).” However, it is not clear to me how this was found. Initially, I thought that the basis for farm classification into 3 models were 7 attributes (including provision of inputs, technical assistance, sale volume agreement, price agreement, coffee bean qualification agreement, speed of payment, delivery, and supervision based on reveal preference technic). However, some of these attributes were then used as independent variables in the construction of the MNL model. It means that these attributes could not be the basis for identifying the strategy (model of contracting) - hence it would be necessary to explain how these strategies were identified (and then used as dependent variables in the MNL model). The question is also whether all farms in a given “contract type” represent only one type of contract farming model - is it possible that a farm represents two types of contracting ?

In line 282 there is information about 4 types of sustainability certificates - there is a question whether the relationship between the contracting model and the type of certificate was examined.
In line 201 there is information about sample size – is it possible to add information what is the share of sustainable certified coffee farmers in general population of coffee farmers?

If I could suggest something more general, I would also suggest to add an explanation why the issue of contract farming is important from sustainability point of view. It would be good to emphasize also (at least theoretically) what are social and environmental aspects of contract farming (and economic as well, however some of them have been already discussed) - what is the significance of contract farming for the social and environmental dimension of sustainable development ?.

Author Response

Dear Professor,

Thank you very much for reviewing the manuscript. I highly appreciate your useful remarks that help to improve the quality of the research paper.

The followings are my responses to each of your comments

We designed questionnaire based on the existing literature regarding contract farming model and preference. The initial steps of the pre-test survey were to identify possible contract farming models and contract attributes in the research area based on contract farming classification of Eaton and Shepherd [38], Bijman [39], and Minot [66] (in term of contract formula or specification, contract format, level of coordination, centralization of production). Figure 3 presents some of these criteria for classification. The proposed contract attributes are reviewed from several literatures that you see in the example below. The pre-test respondents revealed choice set of 3 models and 9 contract attributes (for every coffee production contract with farmer in Dak Lak), which were later incorporated into the finalized questionnaire used in 2017. Before the direct interview are the explanation of the questionnaire terminologies and concepts then the free-range conversation between interviewer and respondent.

The study areas were projected for certified production program. At the beginning, it was confirmed by local DARD that coffee farmers were promoted by more than one contract type in the study area as they are highly competitive markets for coffee beans among coffee-industrial firms.

In the MNL estimation, farmers' choice of contract farming model are dependent variables. Contract attributes are independent variables. Existing literature also considered contract attributes are decisive factors that influence farmers' choices of contract farming model or type (or participation in contract farming, or motivation towards participation)

Here are what you will find in the revised version of the paper related to your remarks (line 273 to 284). This is very useful comments you provided in term of methodological issues.

Based on classifications of Eaton and Shepherd [38], Bijman [39], and Minot [66] (Figure 3), pre-test survey results elicited three different contract farming models for certified coffee farmers in Dak Lak. Those are informal model (InforM), intermediary model (InterM), and nucleus estate model (NeM). Out of 11 proposed contract attributes based on existing literature of [6, 8, 51, 54, 60, 61, 64], the pre-test survey also confirmed 9 common contract attributes.......
In line 282 there is information about 4 types of sustainability certificates - there is a question whether the relationship between the contracting model and the type of certificate was examined.

I have moved that information to section 3.1 research area. The evaluation and re-certification process was done by authorized independent unit known as Dak Lak Cafe Control. That is a business entity specialized in agro-export products evaluation under Ministry of Agriculture and Rural Development (MARD). This information is related to another research paper of mine (about compliance). I think this is not really much of an impact on farmers choices of contract farming model, based on cross-section data approach that we utilized. However, I believe there is the relationship between them because different certified productions require different terms in production contract. This matter deserves further research especially for the panel data, which was not available in this research.
In line 201 there is information about sample size – is it possible to add information what is the share of sustainable certified coffee farmers in general population of coffee farmers?

The relevant information I provided in line 250 to 259 in the revised version.
If I could suggest something more general, I would also suggest to add an explanation why the issue of contract farming is important from sustainability point of view. It would be good to emphasize also (at least theoretically) what are social and environmental aspects of contract farming (and economic as well, however some of them have been already discussed) - what is the significance of contract farming for the social and environmental dimension of sustainable development ?.

That is interesting but this may enlarge the main objective of this paper. As you noticed, I have mentioned it in the paper and I think the typical social aspects of contract farmer are obviously in the intermediary model where the cooperative facilitates trust (reliability) between farmers and firms, and repeated transactions that are embedded in social relationships. The environmental aspects could be the stringent monitoring and evaluation of different certified production under contract.

I hope that the responses would satisfy your request of revision. Besides, I also have done other revision in the paper as you might see in the attachment of revised version (literature review, revised conclusion part, removed some unnecessary descriptive statistics that was not related to variable used in MNL model, cost-benefit analysis and some test statistics in the appendix)

Again, thank you very much for your supporting comments and remarks.

Author Response File: Author Response.pdf

Reviewer 2 Report

Inclusive value chains are very important for (small) farmers since they (inclusive VCs) enable farmers to access into markets. There are different strategies/ways for farmers to enter the market, and contract farming is one of them. Contract farming is very important having in mind farmers’ exposure to market and climate risk, the necessity to invest (in new technologies, food safety standards or simply buying more land or enlarge farm operations) a lot and often, and to keep updated with the recent farm management knowledge. Having all that said, the paper is a valuable contribution to the contemporary themes in agricultural economics.

The paper is well structured. The introduction encompasses the review of relevant literature and sets appropriate goals. The same applies to methodology, model design, and results.

What is confusing are the data in table 6. From the technical point of view, it is not clear are profits calculated from the survey data and refer to one production year only? The title of Table 6. is The relationship between profit and types of contract farming. Does the type of contract farming influence the profit and vice versa? Is profit significantly higher in the case of InterM, as you claimed in line 598? Next, does it means that the utility from choosing InterM the highest? If this is the case, shouldn’t one expect all farmers to behave as rational decision makers and switch from their current model to InterM? Or, some barriers to entry cooperatives exist? Please elaborate. Anyhow, I believe how the paper should have additional (sub)chapter on the economics of coffee production and discuss how contract farming (might) influence on revenues and its components (price, yield, government support…), costs (fixed and variable) and consequently profit.

Discussion is well written. Conclusions should provide a short, clear and systemic presentation of the main results. At the end of the Conclusion chapter, please be more specific on how this research could help in designing policy measures and development project?

Technical issues:

In the introductory part, I would suggest a table about coffee production in Vietnam and comparison with the research area. Also, the map of Vietnam and major coffee production zones (including Dak Lak) would be interesting.

Table 2.- please insert measurement unites next to every variable.

Table 2.- Please use N instead of Obs. or N of Valid Cases as from table 3. forward.

Table 9- Please insert measurement units next to category. Numbers are higher than total and percentages are above 100%. The total of Benefits for InformM is 56 and percent is 119.1 (e.g.)? Consequently, narratives (lines 547 to 573) should be and percentage should be used more carefully.

Author Response

Dear Professor,

Thank you very much for reviewing the manuscript. I highly appreciate your useful remarks that help to improve the quality of the research paper.

The followings are my responses to each of your comments

The paper is well structured. The introduction encompasses the review of relevant literature and sets appropriate goals. The same applies to methodology, model design, and results.

What is confusing are the data in table 6. From the technical point of view, it is not clear are profits calculated from the survey data and refer to one production year only?

That is correct as this study was based on cross-section data. The survey was in 2017. I provided additional table analyzing cost-benefit under different contract farming models. I have revised the term profit into gross margin for more accurate as calculation are based only on production cost.

The title of Table 6. is The relationship between profit and types of contract farming. The name of the table has changed as you see in the line 636 in the attachment of revised version.

Does the type of contract farming influence the profit and vice versa? Is profit significantly higher in the case of InterM, as you claimed in line 598? Next, does it means that the utility from choosing InterM the highest? If this is the case, shouldn’t one expect all farmers to behave as rational decision makers and switch from their current model to InterM? Or, some barriers to entry cooperatives exist? Please elaborate. Anyhow, I believe how the paper should have additional (sub)chapter on the economics of coffee production and discuss how contract farming (might) influence on revenues and its components (price, yield, government support…), costs (fixed and variable) and consequently profit.

Farm performance was analyzed based on the data of one crop year in 2017. Here, I only confirmed that gross margin and rate of return was significant higher in InterM compared to other contract farming model. I think to examine the effects of contract farming on the profit, we may need the better data (panel data). Test statistics are found in table A5 in the Appendix line 852.

In fact, coffee farmers are free to switch to other contract farming model. However, since the analysis of farmers’ gross margin and rate of return was done for cross-section data of one crop year, we do not have strong ground to explain this behavior of coffee farmer. This was already mentioned in the limitation of the study. This deserves for further research as the panel data is available. As I have mentioned in the paper, participation in cooperative is simple registration format, thus farmers are free to be members of cooperative under merely geographical constraint. For your request of economics of coffee production, I have the section 4.5 (line 592) Farm performances under different contract farming model that analyses coffee yield, gross margin, and rate of return for coffee farmer with empirical evidence from cost-benefit analysis in the table A5 in the Appendix line 852.

Thank you for these remarks. I have revised the conclusion of which recommendation were mostly for the involved coffee-industrial firms as the retreat of our government intervention that leaves the sector now is primarily operatizing by the private sectors.

Technical issues:

I have provided information regarding coffee production in Dak Lak and the map of research area in the section 3.1 Research area (line 206)

Table 2.- please insert measurement unites next to every variable.

I have removed the table 2 and incorporated descriptive statistics in table 1 for better illustration

Table 2.- Please use N instead of Obs. or N of Valid Cases as from table 3. forward.

I have changed that as your request.

Table 9 now become Table 7 after revision. The percentage are computed for each column in cross tabulation. Variables under benefits and concerns categories are not mutually exclusive. They are independent choice set from each other. For example, we asked coffee farmers whether they have better farming practices from participation in contract farming and only 6 farmers in InforM claimed that they had. So the percentage of coffee farmer who perceived improved farming practices in InforM model accounts for only 12.8% of the total farmers attributed to this contract farming model. Similarly, we asked whether coffee yield was improved from contract farming participation, 12 out of 47 farmers in InforM claimed that their coffee yield has increased.

Again, thank you very much for your supporting comments and remarks.

Author Response File: Author Response.pdf

Reviewer 3 Report

About the manuscript with the title "Smallholders’ preferences for different contract farming models: Empirical evidence from sustainable certified coffee production in Vietnam" I have the following concerns:

- I suggest the Authors create a new section for the introduction (where present and justify the main motivations, objectives, methodologies and the structure of the research) and a new section for the literature survey related with the topic of the study.

- Before the section the methodology it should be presented a section for the data analysis, including for the data obtained from the survey. This analysis will allow the readers to better understand the reality anlaysed.

- About the model specification the Authors must justify the methodology considered (multinomial logistic model) and not other approaches. For example, it could be interesting before the application of the model, to perform factor and cluster analysis to take into account the heterogeneity between the 183 coffee farmers. Or spatial approcahes (allowing for spatial autocorrelation, for example).

- The variables presented in the table 1 are not the same presented in the table 2!

- The main problem of the work is the explanation of the tables. In the present version is very hard to understand the content of the tables (including for experts in econometrics). Your tables should be understable by the general readers and stakeholders. I suggest you explain clearly the each part of the content present in the tables 3-9 including those in the appendix, as well as, the approaches considered.

- Specifically about the table 8 with the results from estimations, the Authors should test for problems related with multicollinearity, heteroscedasticity and about the adequacy of the model specified.

- The conclusions section must present something new from the study performed, avoiding vague paragraphs.

Author Response

Dear Professor,

Thank you very much for reviewing the manuscript. I highly appreciate your useful remarks that help to improve the quality of the research paper.

The followings are my responses to each of your comments

I have provided new section as your request in the introduction part (you may see in lines 113 to 129 in the attachment of revised version). The review of literature are in lines 132 to 204

The data analysis section was revised as your request in lines 307 to 325 in the section 3.3

The explanation of MNL model specification are in lines 326 to 379 as your request.

We employed between discrete choice approach (based on revealed preference) and the difference of this approach from cluster analysis were discussed in the section 2.2 in the Literature review as your request (line 159)

- The variables presented in the table 1 are not the same presented in the table 2!

I have removed the table 2 and incorporated descriptive statistics in table 1 for better illustration. I also have removed descriptive statistics of variables that are not later used in MNL model.

I think as I revised the Table 1. Explanatory variables hypothesized in the MNL model. The table illustration should be now better to understand the contents.

Multicollinearity diagnostics and heteroscedasticity test are in lines 312 to 319. Table A1 A2 in the Appendix

- The conclusions section must present something new from the study performed, avoiding vague paragraphs.

I have revised the conclusion as your request.

Again, thank you very much for your supporting comments and remarks.

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

I suggest the Authors explore better the results presented in the tables.

My suggestion is the Authors become the tables explanations simpler and more practical. For example, about the table 4 the Authors presented:

"The results from multinomial logistic estimation of farmers’ preference for different contract farming models are in Table 4. Positive estimated coefficient indicates an increase in the likelihood that a certified coffee farmer will change to the alternative option from the reference category (base category) [76]. We selected informal farming model (InforM) as a base category because it covers the smallest group of farmers in the research sample. The discussions of marginal effect Exp (B) are in term of the significance and signs of variable. The marginal effect of an estimated coefficient indicates that the probability of farmers’ preference for the InterM and NeM compared to the probability of farmers’ preference for the InforM changes with variable. The marginal effect greater than one refers to the probability of farmers’ preference for the InterM and NeM relative to the probability of farmers’ preference for the InforM increases as the variable increases. If the marginal effect is less than one, certified coffee farmers are more likely to prefer the InforM (reference contract-farming model)."

Also for me this is very confusing. It could be important become this simpler to the general readers. Another question is the practical implications of these results, namely in terms of policy framework and sector organization. The same happens for the other tables.

Author Response

Dear Reviewer,

First, I have revised the paper in accordance with your comments. To be specific, the explanation of the sign of estimated coefficient and marginal effect in the multinomial logistic model in Table 4 was bellows:

Positive sign of estimated coefficient indicates an increase in the probability of the alternative outcomes (here are InterM and NeM) relative to the changes in the probability of the reference category (InforM) response [75]. On the contrary, negative sign of estimated coefficient indicates a decrease in the probability of the alternative outcomes relative to the changes in the probability of the reference category response. The selection of reference category depends on the scientific aims of the study. Therefore, we selected informal farming model (InforM) as a base category because it covers the smallest group of farmers in the research sample. The marginal effect value Exp (B) indicates the magnitude of the changes in the probability of the alternative outcomes relative to the changes in the probability of the reference category response as one-unit increases in the explanatory variables.

In fact, this is confusing (said the 3^rd reviewer) because disagreement in the sign between the marginal effect and the coefficient comes up often in the MNL models and usually puzzle people who are not accustomed to it. In MNL model, the coefficient describes how changes (increase of decrease) in each outcome probability related to changes in the probability of the base category. However, the sum of the probabilities of all the category responses must always equal to 1. These models are difficult to wrap ones’ head round. I believed I had not-so-much confusing explanation in the previous version if the reviewer gets to the more-specific estimation results at the next paragraphs right down below. Besides, there was a citation of Greene 2010 for those who are not familiar with MNL.

Second, the discussion section already referred to several “practical implication” of the previous empirical results. I did organize the paper by providing descriptive results first then discussed each of the descriptive results in line with previous literature in regards of contract farming, contract attributes, contract preference, problems and issues in facilitating agreement of contract terms under the context of sustainable development of the coffee sector. In addition, the retreat of our government led to the fact that the sector is now primarily operating by the private sector. However, the discussion of overall institutional structure and involvement of public agencies and non-government organization was already there in section 4.2 and discussion section (lines 745-755). The second reviewer already mentioned this issue and I have explained it.

I hope this would help to satisfy the request from you.

Article Menu

Smallholders’ Preferences for Different Contract Farming Models: Empirical Evidence from Sustainable Certified Coffee Production in Vietnam

Further Information

Guidelines

MDPI Initiatives

Follow MDPI