Elsevier

Economic Analysis and Policy

Volume 56, December 2017, Pages 114-125
Economic Analysis and Policy

Full length article
Which farming systems are efficient for Vietnamese coffee farmers?

https://doi.org/10.1016/j.eap.2017.09.002Get rights and content

Abstract

This paper provides a comparative assessment of the productive efficiency of three common coffee growing systems in Vietnam: mono-cropping, synchronization and segregation. Results from an input distance function approach deliver several important findings. First, the average inefficiency level is estimated to be around 18% although inefficiency varies significantly between the three farming systems. Second, the synchronized system of growing coffee and the other industry crops is found to be the most efficient farming system. Third, coffee mono-cropping is less efficient than synchronized systems due to the presence of economies of scope between coffee and industrial crops. Fourth, the least efficient system is segregated cultivation of coffee and rice. Food insecurity is seen as a primary reason for coffee farmers diversifying into rice. These findings provide empirical evidence of agronomic benefits being derived from synchronized systems, and which are translated into higher productive efficiency. Policy options promoting synchronized farming systems may therefore enhance both economic and agronomic benefits.

Introduction

Coffee production is one of the primary economic sectors in the Central Highlands region of Vietnam with nearly 96% of Vietnam’s export of coffee coming from this region. Due to significant price increases in the early 1990s, the area used for coffee cultivation increased by approximately 400% from 1999 to 2000. This expansion appears to be a natural adaptation of farmers in response to past increases in prices. However in subsequent periods the resulting increases in market supply caused prices to drop to a level which, by 2001, was lower than the production cost (Marsh, 2007). This forced many coffee farmers into bankruptcy (Wollni and Zeller, 2007) and is seen as one of the reasons motivating coffee farmers to diversify their business and for the presence of several distinct farming systems in Vietnam.

In this study, we consider three typical coffee growing systems in Vietnam: mono-cropping, synchronization and segregation. The nature of specialization and diversification vary significantly across these three distinct systems. Mono-cropping farms have only one land plot and grows only coffee. Segregated farming systems have more than one plot of land with each plot growing one primary type of crop. For example, where farms have two plots, one plot grows coffee and another plot grows rice. Synchronized farming systems grow coffee together with other industrial crops in one plot and rice in a separate plot(s). There is an obvious need to know which farming system are most efficient for coffee farmers in Vietnam — an issue which the existing literature provides no empirical evidence. This literature gap is, therefore, the primary motivation of the present paper.

It is noted that diversified systems, particularly through crops diversification, may obtain higher yields and/or cause less environmental damage (Letourneau et al., 2011). This is known as complementary or synergy effects among crops sharing the same environment. However, there may be diseconomies of scope or negative effects of synergy as empirically observed in, for example, Coelli and Fleming (2004). Since appropriate crop diversification strategies can deliver positive effects of synthesis, it could be expected that coffee farmers would be motivated to diversify by growing industrial crops such as pepper or durian which may also mitigate market risk. However, it becomes less clear why coffee farmers have chosen the segregated system in which rice and coffee are grown in separate land plots given it does not deliver positive synthesis effects and could produce diseconomies of scope (Villano et al., 2010). More particularly, some studies have hypothesized that Vietnamese coffee farmers diversify to rice because of the insecurity created by low incomes and volatile market conditions (i.e., Dang, 2003). If this is true, segregated farms face a trade-off between productive efficiency and income or poverty risks. In this paper, we aim to provide empirical evidence on this trade-off hypothesis. Our empirical results are therefore designed to indicate the rationale for Vietnamese farmers’ decisions over which crops to grow rather than accepting that they may be made on ad hoc or irrational basis (Dang and Shively, 2008). As such, this study can provide a useful guide for policy makers in raising the productivity of Vietnamese coffee farmers.

We utilize several techniques to examine differences in the level of productive efficiency among the three farming systems using a dataset of 167 farms surveyed in five Central Highlands communes in 2012. The input distance function is used to estimate efficiency scores for each farm. Parametric and non-parametric tests are then applied to assess if these differences are statistically different across the farming systems of the three districts. Additionally, the input distance function allows us to examine economies of diversification which is based on the concept of economies of scope in diversified farms Baumol et al. (1982), Willig (1979). While there are several approaches to measuring scope economies (Chavas and Di Falco, 2012; Chavas (2008), Ofori-Bah and Asafu-Adjaye (2011), Chavas and Kim (2010), Hajargasht et al. (2006)), we use Coelli and Fleming’s (2004) model as it does not require price information and provides a more straightforward interpretation of both efficiency results and diversification economies of each pair of crops.

The remaining part of this paper is set out in eight sections. Section 2 provides a literature review. Section 3 provides a measure of economies of diversification using the distance function. Section 4 sets out the empirical models, data sources and the use of relevant variables. Survey and descriptive statistics are presented in Section 5. Section 6 provides the empirical results. Section 7 discusses the presence of agronomic benefits and the way in which they are translated into efficiency improvements and provides an explanation of why coffee farmers still choose rice. Section 8 sets out the conclusions, policy recommendations and avenues for further study.

Section snippets

Literature review

The various dimensions of farming management practices are well captured in the literature (Bell and Moore, 2012). In particular the farming system in which crops use the same resources, i.e., water and nutrients simultaneously, is known as an intercropping system or synchronization (van Asten et al., 2011). Another common farming integration system is crop rotation: however it is not applicable to perennials such as coffee and other industrial crops. In addition, segregated systems are known

Input distance function

The measure of diversification proposed by Coelli and Fleming (2004) in the framework of the input distance function has several advantages. First, it is applicable to cases where data on input prices are not available. Second, this measure allows the derivation of complementary effects directly from the distance function. Third, an inefficiency model can also be estimated to examine factors that can explain variations in the technical efficiency across farms. Following Coelli and Fleming (2004)

Empirical model specifications

A translog input distance function is used in this study to characterize the production technology as it is more flexible than the Cobb–Douglasfunctional form and does not require restrictions of rectangular hyperbola isoquants (Morrison-Paul et al., 2000). The fully specified translog input distance function is specified as: lnDiI=α0+k=1qβklnyk+12k=1qk=1qβkklnyklnyk+m=1pαmlnxm+12m=1pm=1pαmmlnxmlnxm+12m=1pk=1qτkmlnxmlnykwhere i=1,,N refers to the number of firms, q and p

Farming systems and data collection

As depicted in Fig. 1, the sample is categorized into coffee specialized farms and diversified farms with more than one farming enterprise. There are two main types of diversified farms. The first are farms where different crops share or co-locate in the same plot of land. This is called a synchronized farming system, i.e., coffee — other industrial crops. The second type are farms are characterized by different crops being planted in separated plots of land. This is called a segregated farming

Maximum likelihood estimates

The result’s maximum likelihood estimation (MLE) of the stochastic input distance function is presented in Table 2. The statistically significant negative signs of the first order coefficients of outputs imply an inverse relationship between the input distance and output quantities; that is, more outputs result in a smaller distance or a higher efficiency level. The estimated coefficients, β 1= -0.4956 and β 2 =0.2875, indicate that an

Evidence of agronomic benefits between coffee and other industrial crops

The results from both the diversification measure and the inefficiency model consistently confirm that growing other industrial crops in the same land plot of coffee provide efficiency benefits. It should be noted that our efficiency measures are not impacted by prices of inputs, and we argue that this observation can be due to at least three possibilities. First, the positive complementary effects of growing other industrial crops in the coffee plot is due to lower overall consumption of

Conclusions

This paper adds to the literature on crop farming diversification by coffee farmers by providing additional empirical evidence derived from estimations of the input distance function based on a new dataset of Vietnam’s coffee farmers. Empirical results provide several important findings. First, the average inefficiency level is estimated to be 18%. Second, there is evidence of positive economic benefits in the form of economies of scope and diversification efficiencies flowing from the

Acknowledgements

The authors would like to gratefully acknowledge two anonymous reviewers for their valuable comments and suggestions. We are responsible for any remaining errors.

References (48)

  • IllukpitiyaP. et al.

    Farming vs forests: Trade-off between agriculture and the extraction of non-timber forest products

    Ecol. Econom.

    (2010)
  • MandalR.

    Flood, cropping pattern choice and returns in agriculture: A study of Assam plains, India

    Econ. Anal. Policy

    (2014)
  • O’DonnellC.J. et al.

    A Bayesian approach to imposing curvature on distance functions

    J. Econometrics

    (2005)
  • Ofori-BahA. et al.

    Scope economies and technical efficiency of cocoa agroforesty systems in Ghana

    Ecol. Econom.

    (2011)
  • PadrónB.R. et al.

    Diversification and labor market effects of the Mexican coffee crisis

    World Dev.

    (2015)
  • Picazo-TadeoA.J. et al.

    Assessing farming eco-efficiency: a data envelopment analysis approach

    J. Environ. Manag.

    (2011)
  • RahmanS.

    Whether crop diversification is a desired strategy for agricultural growth in Bangladesh?

    Food Policy

    (2009)
  • van AstenP.J.A. et al.

    Agronomic and economic benefits of coffee-banana intercropping in Uganda’s smallholder farming systems

    Agric. Syst.

    (2011)
  • van de WalleD. et al.

    Sources of ethnic inequality in Viet Nam

    J. Dev. Econ.

    (2001)
  • VillanoR. et al.

    Evidence of farm-level synergies in mixed-farming systems in the Australian Wheat-Sheep Zone

    Agric. Syst.

    (2010)
  • AhmedN. et al.

    Socioeconomic aspects of rice-fish farming in Bangladesh: Opportunities, challenges and production efficiency

    Aust. J. Agric. Resour. Econ.

    (2011)
  • AthukoralaW. et al.

    Impact of agricultural chemicals on the environment and human health in Asia

  • BatteseG.E. et al.

    Estimation of a production function model: with application to the pastoral zone of Eastern Australia

    Aust. J. Agric. Econ.

    (1977)
  • BaumolW.J. et al.

    Contestable Markets and the Theory of Industry Structure

    (1982)
  • Cited by (0)

    View full text