Submitted to EEPSEA
Awareness of consumers and their willingness to pay for green agricultural products: the case of
VietGAP dragon fruit in Vietnam
Principal researcher: Hoang Trieu Huy
Current affiliation: College of Economics, Hue University 100 Phung Hung Street, Hue City, Vietnam Tel: 84 54 3529435
Email address: email@example.com
Section 1: Researcher’s details
1. Name of principal researcher: Hoang Trieu Huy 2. Citizenship/Nationality: Vietnamese
3. Current Affiliation: College of Economics, Hue University 4. Current Designation: Lecturer
5. Mailing Address: College of Economics, Hue University 100 Phung Hung Street, Hue City, Vietnam 6. Contact Number: 84 54 3529435
7. Email address: firstname.lastname@example.org
8. Previous attendance to an EEPSEA meeting, training, or workshop - EEPSEA biannual meeting
- Small grant research meetings 9. Name(s) of co-researcher(s):
Dr. Bui Dung The
College of Economics, Hue University, Vietnam
MSc. Truong Quang Dung
College of Economics, Hue University, Vietnam
Section 2: Full proposal information
1. Proposed title
Awareness of consumers and their willingness to pay for green agricultural products: the case of VietGAP dragon fruit in Vietnam
2. Type of grant requested: Individual Research Grant
3. Research problems
Consumers’ concern with environmental and health protection issues have just become popular in Vietnam recently (Dam, 2010), while such matters had been worrying consumers in the USA since the 1960s (Klonsky and Tourte, 1998) and in Europe since the mid-1980s (Greenan et al., 1997). Environment and health address the question of
‘consumerism’, its influence on human health and on the long-term maintenance of the country's resources (Silverstone, 1993). One type of environmental and health consciousness is the safety of food consumption, especially, the consumption of fresh fruit and vegetable products. Consumers who claim to be environmentally and healthily conscious wish to place stricter quality monitoring mechanism on producers (Krystallis and Chryssohoidis, 2005). Since the early 2000s, there have been regulations related to food safety and quality in Viet Nam. One of the important events was the issuing of the National Assembly Standing Committee Ordinance on Food Hygiene and Safety in 2003, which were followed by several decrees and directives, including the approval of the National Action Plan for Food Safety in 2006.
The fruit and vegetable industry in Vietnam, while regulated, still struggles to meet national and international food standards (Dam, 2010). Small scale farmers are often blamed to pump their crops full of chemicals to increase yields or prolong the shelf life of the produce (Almvik et al., 2012). For example, farmers were found injecting their fruit with dubious chemicals (e.g. a yellow substance imported in unmarked containers from China that were injected or sprayed onto fruit and vegetables) to make them ripen quickly and stay fresh longer (Mueller, 2013). More than 5000 incidents of food poisoning from pesticide residues were reported in 2008 (MOH, 2008). Little is known about the extent and amounts of pesticide residues in fruit and vegetables produced by farmers. It is alleged that farmers often use agrichemicals at higher doses than recommended.
On the market, there is not yet an appropriate and reliable quality control system to guarantee that certified fruit and vegetables on sale do not exceed Maximum Residual Levels (MLRs) of chemicals and contaminants. In addition, product differentiation between certified/safe and ‘unsafe’ fruit and vegetables is often not distinctive for consumers. In many cases, only verbal explanations of the shop owners act as the guarantee that they are indeed selling ‘safe’ products (Trudel and Nam, 2009).
Accordingly, many owners of fruit and vegetable shops at local markets can easily claim the safety of their products and no one knows for sure if the claim is true. Such the claim
can be made as health problems arising from eating unsafe fruit and vegetables manifest slowly and at a latter time, unless there is acute poisoning. Regarding supermarkets, they insist that farmer associations and cooperatives take responsibility for safety assurance and take care of auditing for compliance with regulations for safe food production (Nicetic et al., 2010). This makes some consumers view ‘safe food’ more as a marketing hoax, rather than a real healthy and environmentally friendly choice (Mueller, 2013).
Consumers’ recent demand for environmentally friendly and safe food has again gained significant government support, especially resulting in the introduction of the Food Safety Law and the Consumer Rights Protection Law in 2011, and the establishment of Vietnam Good Agricultural Practice (VietGAP) in 2008. While the Food Safety Law provides guidelines on food safety and defines rights and obligations of organizations, producers and consumers in general; VietGAP is a government decree laying out the main standards and guidelines for production of safe fruit and vegetables supported by official certification and auditing systems. VietGAP is closely inspired by AseanGAP, EurepGAP/GlobalGAP, and Freshcare. It is aimed at preventing or minimizing the risk of hazards which occur during the production, harvesting and postharvest handling of fresh fruit and vegetables (VietGAP, 2008). The hazards covered in VietGAP include food safety, produce quality, environmental impacts and health, safety and welfare for workers.
As of now, however, it is not clear whether most farmers and enterprises will benefit from the adoption of VietGAP. On the one hand, in many cases, VietGAP is still perceived by farmers and enterprises as a procedure increasing the cost of production without a corresponding increase in the price paid by consumers. On the other hand, there are uncertainties in consumer awareness of, and attitude towards, VietGAP-certified agricultural products. These uncertainties may hinder farmers in their efforts to implement VietGAP. In this regard, one of critical issues relates to consumer demand or willingness to pay for VietGAP-certified fruit and vegetables which farmers can offer.
While a number of studies relating to willingness to pay (WTP) for different types of GAPs have been carried out recently (e.g. Florax et al., 2005), most of them focus on industrialized countries. Since the results cannot simply be transferred to Vietnam, specific research in Vietnam is needed, particularly in metropolitan areas, which usually have a leading role in the food system transformation for the rest of the country (Pingali, 2007). The literature on consumer willingness to pay (WTP) for safe fruit and vegetables in Vietnam, however, is rather sketchy. In addition, most of the studies are descriptive in nature. To our knowledge, there have been only two studies conducted to estimate consumer WTP for safe vegetables in Hanoi and Ho Chi Minh City. However, there are not yet any studies conducting to estimate consumer WTP for VietGAP-certified fruit and vegetable products (for more, see Literature reviews section).
Given the above background, this study attempts to investigate awareness of consumers and their willingness to pay for fresh VietGAP dragon fruit using a stated choice model.
The choice to focus only on fresh fruits is driven by the fact that, as Gil et al. (2000) suggested, environmental attributes are more important in fresh and perishable products,
(or at least it is easier to directly identify them in such products), and also, consumers are willing to pay a higher premium for certified fruits. In addition, the dragon fruit is chosen because the fruit is very popular in Vietnam. The dragon fruit is considered as a high value commodity and the tree is able to produce fruits for fresh eating year round.
Furthermore, nearly 7,000 hectares growing dragon fruit (in about 25,000 hectares in total) are applying the VietGAP procedures, ranking in the top three of the country’s certified fruits with the VietGAP certificate based on cultivated area. Clearly, understanding consumers’ valuation of VietGAP product attributes should be essential for designing effective food policies. The findings are expected to be useful for Vietnam fruit and vegetable industry, including fruit growers/farmers, warehouse and retail store managers, marketing managers and food policy makers.
4. Gender aspect
The importance of fruit to the Vietnamese rural economy is evidenced by the considerable increase in the total area for fruit production and its important role in improving rural farm income. Gross revenues from fruit are very much higher than that from rice (Lan, 2010). Many growers involved in the cultivation, harvesting and sale of fruit are women. Therefore, understanding consumers’ valuation of certified agricultural products, and VietGAP fruit in particular, are expected to be useful for designing effective food policies that can enhance the important role of women in the safe production, promotion, and utilization of VietGAP products, and hence increase their incomes. For instance, the empowerment of women in fruit and vegetable industry can be achieved through, education and skills training in VietGAP-oriented safe production techniques, marketing and utilization of safe fruit and vegetables.
Furthermore, women are in most cases are the buyer of the dragon fruit. They will be respondents of the study. Their perception and attitude toward food safety will be analyzed. Comparative analysis between man and woman will be undertaken. Policy implications of the study will pay attention to the gender aspect of the production and consumption of the dragon fruit.
5. Policy context
Since the VietGAP norms are still quite new and complex to be implemented by small scale growers/farmers, the adoption of VietGAP will require further testing, establishing of demonstration sites, development of manuals, training, and dissemination. In this regards, this study should provide important insights for public and private policy- makers. Choice experiment studies should provide relevant demand-side information to fruit agribusiness managers and food policy makers in various ways. First, it should be an opportunity for fruit producers to add value, improve their products and find new ways to possess the attributes most valued by consumers. This adds to the concept of market driven fruit and vegetables, a strategy in which managerial decisions are based on information elicited from consumers rather than from the inside business only (Jaeger and Harker 2005). Second, it should benefit warehouse and retail store managers to apply appropriate techniques to enhance fruit quality characteristics most appealing to consumers. Third, it should be valuable for marketing managers to base their strategies
on research-based information about consumers’ preferences. Finally, it should be of interest to VietGAP regulators (e.g. the Ministry of Agriculture and Rural Development) to evaluate current VietGAP promotion programs and facilitate the work of organizations in disseminating the new practices to farmers and enterprises. An important contribution of this study is to minimize the problem of mismatch in terms of attributes that can be supplied by fruit producers and what the fruit consumers really want and are willing to pay for. To date, no such study has been conducted in Vietnam.
6. Research objectives
6.1. General Objectives
In view of the research problem identified, the overarching objective of this proposed study is to enhance the understanding of research-based information on consumers’
preferences for fresh fruit in Vietnam.
6.2. Specific Objectives
Specific objectives are proposed
(i) to document the characteristics and trends, including constraints and opportunities, for the development of the VietGAP dragon fruit markets in Vietnam;
(ii) to explore consumers’ awareness of VietGap-certified dragon fruit;
(iii) to empirically elicit consumers’ willingness to pay for different attributes of dragon fruit;
(iv) to estimate the implicit price for each attribute of dragon fruit and the tradeoffs among the attributes;
(v) to inform public and private policy-makers in Vietnam; and
(vi) to enrich the empirical literature on GAPs in developing countries.
7. Literature reviews
Consumer demand in Vietnam for environmentally friendly and safe fruit and vegetables, as measured by consumers’ willingness to pay (WTP), has been documented by several studies. Hai et al. (2013) employed contingent valuation (CV) study to estimate consumer WTP for organic vegetables in Hanoi. They found that the average price paid for organic vegetables was about 70% higher than that of conventional ones. Besides, the consumers with higher income and organic-consumed experience in the past would be likely to pay more for organic vegetables. In another study, Mergenthaler et al. (2009) also used CV techniques to estimate consumer WTP for safe vegetables in both Hanoi and Ho Chi Minh City. Their results showed that consumers are willing to pay an average price premium of 60% for chemical free mustard and of 19% for different convenience attributes of potatoes. Income levels and media also have positive impacts on the WTP.
To our knowledge, however, there have been no studies in Vietnam estimating consumer WTP for fruits, including VietGAP-certified fruit products using a stated choice model.
Internationally, efforts to understand consumer attitudes and the relative importance of various attributes in purchasing fruits have been widely explored, primarily with stated preference techniques such as Contingent Valuation (CV) and Choice Experiments (CE).
For example, Zhang et al. (2010) used the CV method to evaluate consumers’ WTP for Anjou pears with different levels of ethylene treatments. They found that consumers prefer pears with a six-day ethylene treatment and are willing to pay a premium of
$0.25/pound compared to the market price. Choosing the same fruit and applying CE, Gallardo (2011) found that consumers were willing to pay $0.19, $0.16, $0.16, and $0.06 for a one unit increase in Anjou pears’ sweetness, texture, juiciness, and firmness, respectively. In general, the literature reviews suggest that fruit attributes can be classified into three main groups. The first group includes flavor - defined as being made up of taste (sweetness, acidity, astringency, bitterness) and aroma - texture (defined as firmness, juiciness, succulence) and color and shape (Tan, 2000). These visual, smell and taste components are often top rated among attributes listed since they represent the basic components of eating pleasure (Ernst et al. 2006).
The second group is credence attributes, including health related components; production methods related attributes; environmental and socially oriented attributes; local and origin related attributes; and certification and other labels (Moser et al., 2011). Perceived personal health related differences in fruits can be linked to specific food components (artificial additives, genetically modified organisms), to the presence of nutritional components (rich in vitamins), and to the perceived risk associated with the use of agrochemicals (e.g. whether a specific fruit is pesticide free). Production methods evoke a bundle of attributes related to environment, risk concerns, and certification criteria, and in many cases, related to other attributes. The relevance of environmentally related attributes include increased biodiversity, ecosystem protection and natural system conservation; while socially oriented attributes can be job creation or support for farms.
The attribute local generally seems to be relevant to the decision to buy fresh fruits, while attributes referring to the products’ origin was found to be either important, or somewhat important in a majority of the studies (Moser et al., 2011). Also, consumers often use third party certification and labels as safety and quality cues for attributes that require oversight by knowledgeable experts (Lohr 2000). The third, and final, group includes branding and packaging attributes. Price attributes also belong to this group.
8. Research Methods
8.1. Research questions
Given the low level of environmental and health consciousness by fruit growers/farmers in Vietnam and the constant battle between supporters of VietGAP and those of conventional growing techniques, it is reasonable to ask:
(1) What are the characteristics and trends, including constraints and opportunities, for the development of the VietGAP dragon fruit market in Vietnam?
(2) What is willingness to pay (WTP) for different options of dragon fruit attributes?
(3) How are the attributes of dragon fruit important to consumers? and
(4) What is the implicit price for each attribute of dragon fruit?
8.2. Analytical framework/economic models
Willingness to pay with the use of a choice experiment
Consumer decisions about choices are the result of a complex relationship between personal preferences, socio demographics, psychosocial, and environmental factors (Trudeau et al., 1998). Consumer preferences for certain food attributes are important for food producers and processors as well as policy makers (Gao & Schroeder, 2009).
Different preference elicitation methods have been used by economists and market researchers to obtain the WTP for certain product attributes. Some of those methods, such as conjoint valuation and choice experiments are defined as stated preference methods because participants are asked to make hypothetical choices rather than real market choices. In consumer WTP studies, non-hypothetical incentive compatible elicitation mechanisms can be used as well, for example, experimental auctions, where real money is exchanged for actual products. For new product ideas, non-hypothetical studies often cannot be done since the product does not exist yet. If the product is available, non- hypothetical preference methods such as experimental auctions are preferred since those methods are theoretically incentive compatible. However, the cost of conducting experimental auctions is significantly higher since the participants are involved for a longer period of time. Oftentimes, because of budget constraints for this type of research, hypothetical methods are chosen to calculate WTP.
Lusk and Schroeder (2004) compared two different hypothetical methods, conventional contingent valuation (CV) and choice experiment and preferred choice experiments for three reasons: (1) with choice experiments, multiple attributes can be valued simultaneous, (2) they are consistent with the random utility theory and Lancaster’s theory of consumer demand, and (3) a choice experiment is more similar to the actual consumer purchasing decision because a choice needs to be made among alternative products and therefore might be less prone to hypothetical bias compared to CV. As a result, we use a choice experiment to elicit consumers’ WTP for VietGAP dragon fruits.
In choice experiments participants are asked to make repeated choices between alternatives described by varying attributes. It is an established approach for understanding and predicting consumer trade-offs and choices in marketing research based on a household survey (Gracia et al., 2009). The choice experiment is consistent with the random utility theory (McFadden, 1974) which assumes that decision makers are rational and individuals make choices to maximize their utility subject to their budget constraints. Choice experiments are also based on the Lancasterian consumer theory (Lancaster, 1966) which assumes that utility of a good can be segregated in utilities of different attributes of a product and proposes that consumers make choices based on preferences for attributes of these goods. Goods are made up of attributes and the total utility gained from a product or service is the sum of the individual utilities provided by the attributes of that good. As a result, choices are determined by particular combinations of product attributes. Utility is derived from the attributes and attribute levels of product.
In making choices, participants make trade-offs between different attributes and attribute levels (James & Burton, 2003).
Individuals are asked to choose their preferred alternative amongst hypothetically constructed scenarios, where each scenario is a function of different attributes of product (including price) and each attribute varies at different levels. The response data are modeled within a utility function which provides information on: whether the given attributes chosen are important; the relative utility of different attributes and combination of it; the rate at which individuals are willing to trade between attributes (trade-off) and the total satisfaction or utility that participants derive from the product.
The main purpose of this study is determining the importance of fruit attributes and the willingness-to-pay for those attributes, the random utility theory and a conditional logit model will be utilized as it allows the estimation of both mentioned interests. Following McFadden (1974), the econometric models based on the random utility theory are called random utility models and are derived as follows. A decision maker i faces Jalternatives and the utility derived from choice alternatives j is Uij, j=1,...,J. Since there are aspects of utility that the researcher does not or cannot observe, the decision maker’s utility Uij differs from the observed utility Vij. Therefore, the ith individual’s utility of choosing option j consists of an observed (deterministic) component Vij and a random stochastic component εij:
(1) Uij =Vij+εij
The key assumption is that each individual i will choose the alternative in the choice set which will provide him/her the highest utility (utility maximizing behavior). When an individual i is facing a choice set Ci , consisting of J options, the choice probability of choosing alternative j is equal to the probability that the utility of alternative j, Uij, is greater than or equal to the utilities of all other alternatives in the choice set, i.e.
(2) P ri
(Uij≥Uik, for all k∈Ci withk≠ j
(Vij+εij ≥Vik +εij, for all k∈Ci withk ≠ j
Assuming that the random errors in equation (1) are i.i.d. (independently and identically distributed) across the J alternatives and N individuals with a Type I extreme value distribution and scale parameter equal to 1, the probability that an individual i chooses alternative j of a choice set Ci can be given by:
Pr chosen ij ik with
V j V
j e e k C
The observed (deterministic) utility component, Vij, may consist of different types of determinants. Alternative-specific constants for all but one (the reference) alternative, if enter the model, capture choice probabilities relative to the reference alternative that cannot be attributed to the other explanatory variables. Also, the underlying utility may depend on individual-specific and/or alternative-specific variables. Individual-specific variables describe characteristics of the decision maker and may influence the relative attractiveness of the alternatives. Alternative-specific variables vary both over individuals
and alternatives. They may enter the utilities in two different ways. Since the variation over alternatives provides additional ground for identification, a separate parameter for each alternative is statistically identified. On the other hand, a joint coefficient for all alternatives can also be estimated. This is possible because of the variation of alternative- specific variables over the alternatives.
Including all these variables, the deterministic part of the utility Vij can, in general, be written as:
(5) Vij =αj +xij′βj+zi′γj
where individual-specific variables are collected in the vector zi, alternative-specific variables are collected in the vector xij for each individual, and αj is an alternative- specific constant.
The relative importance weights indicate which attributes are more important in influencing consumer choice. The relative importance can be calculated by using the following formula:
= UR ×
where RIl is the relative importance of the attribute lth and URl is the utility range for the attribute lth (Harrison et al., 2002). The utility range is the difference between the highest and lowest part-worth value of the attributes.
The willingness-to-pay is used to estimate the amount of money an individual is willing to pay to obtain a specific attribute. In this study willingness to pay is interpreted as a dollar increase that consumers are willing to pay to obtain an attribute. Willingness-to- pay for attribute l will be calculated as the negative ratio of the coefficient for attribute l and the price premium coefficient as:
(7) l l
where βl is the coefficient of attribute l and
βM is the price premium coefficient.
The conditional logit model (5) can be estimated by the maximum likelihood technique.
However, the model has some limitations (Phanikumar & Maitra, 2007) such as the Independent of Irrelevant Alternatives (IIA) property (implying proportional substitution across alternatives), or the assumption of preference homogeneity in the sample (implying that all coefficients of all attributes in the utility function are assumed to be the same across all participants), or the assumption of independent errors over time.
8.3. Variables or factors to be measured or otherwise addressed by the research
Consumer preferences for certain fruit attributes are important for fruit producers and processors as well as policy makers. Consumer decisions about choices are the result of a complex relationship between personal preferences, socio demographics, psychosocial,
and environmental factors. In this study, the psychographic and behavioral variables are collected to test which participants are information seekers, price conscious, health conscious, weight conscious, environmental responsibility, time pressure (busy life), and which participants choose products for convenience. The health and weight conscious variables are included in the study because fresh dragon fruit are often considered a healthy product. It is hypothesized that people who are interested in having a healthy life will choose a product with attributes that involve a health improvement. Demographic variables (e.g. age, gender, education, place of residence, income, marital status), frequency of fruit consumption and the participant’s opinion regarding price issues are also collected. The participant’s opinion regarding price issues is used to determine if consumers are price sensitive and compare these results with the selection of the fruit product. For certain fruit attributes and levels used in the choice experiment, the following section will describe the selection procedure in details.
8.4. Methods to be applied in collecting primary data
Conjoint analysis is a multivariate technique developed specially to understand how consumers develop preferences for different products or services. It is widely used in marketing research because it allows estimating consumer’s preferences of a product by combining part-worth utilities for each attribute. Conjoint analysis is not a model or an estimation technique but rather a methodology for constructing the data collection instrument when the final objective is choice modeling. In the first step, the product attributes and their levels are defined. In most situations consumers know which attributes are more important when they look to purchase a product. Consumer perceptions of a product are based on more than one attribute, so it is fundamental to identify which attributes influence their intention to purchase. A series of focus groups and other qualitative technique could be used to obtain information about the attributes and levels. In the second step, an experimental design and a choice of data collection method are constructed.
Selection of product attributes and their levels
A bundle of specific attributes and levels are required that are directed to satisfy consumers’ needs. A series of focus groups and other qualitative technique should be the usual means to determine the attributes and their levels, including definition of attributes, number of levels for each attribute, wordings, and the effect of photos/visual aids or real products. Also, some rules of thumb should be followed, including:
- No more than 6 or 7 attributes employed with 2 or 3 attribute levels. Many attributes and levels may cause participants to simplify, looking only at 2-3 most important.
- Attributes should be independent, mutually exclusive.
- The number of scenarios (or sets) should be around 20 and 40 at the maximum.
As for now, the attributes used in this proposal are only based on literature review and will be modified based on the results of focus groups that will be taken. Two issues in particular need to be considered when deciding which attributes will be considered. First, the attributes should be relevant to the requirements of the policy makers. Second, the attributes need to be meaningful and important to the respondents. Accordingly, six
attributes of dragon fruits are identified and examined. Since this study focuses on the role of VietGAP on consumer’s choices for fruit, VietGAP-related attributes are firstly identified. In the beginning, two important attributes relating to VietGAP, the production method and the certification, are identified. However, we suspect that ‘production method’ attribute and ‘certification’ attribute are not completely independent. Therefore, the ‘production method’ attribute is dropped.
In reality, the VietGAP certification can be obtained from certification bodies which are accredited by Department of Crop Production to operate over the national territory, and by Provincial Department of Agriculture and Rural Developments to operate at the province level. We also aware that there are farmers/enterprises already certified for EurepGAP/GlobalGAP, for example the Ham Minh Dragon Fruit Cooperatives located in Binh Thuan Province. Accordingly, three levels of the certification attribute are defined:
the no-certificate, the VietGAP certificate and the GlobalGAP certificate. However, this attribute is presumably quite abstract and difficult for respondents to understand. In order to make the attribute meaningful and important to the respondents, clear explanations of VietGAP are needed. As an example, information related to VietGAP presenting to respondents may be as follows.
VietGAP - Good Agricultural Practices for production of dragon fruits in Vietnam
What is the VietGAP?
- is a government decree laying out the main standards and guidelines for production of safe dragon fruits supported by official certification and auditing systems.
What are the purposes of implementation of VietGAP?
- to enhance the responsibility in production and management of food safety of dragon fruit growers, and
- to enhance the quality of dragon fruit
What are the differences among VietGAP, GlobalGAP and conventional production methods?
- (Show respondents Table 1)
- VietGAP was developed based on GlobalGAP with slightly lower criteria for compliance in the areas of worker protection and environmental issues but not in areas that directly affect food safety including compliance with pesticide and fertilizer use or microbiological contamination.
VietGAP provides standards for: a) Site assessment and selection, b) Planting material, c) Soil and substrate management, d) Fertilizers and soil additives, e) Water and irrigation, f) Crop protection and use of chemicals, g) Harvesting and post harvest handling, h) Waste management and treatment, i) Worker health and welfare, and j) Record keeping, traceability and recall
Who are eligible for VietGAP certification?
- all growers of dragon fruits who wish to follow VietGAP production procedure can apply for VietGAP certification, either in individuals or in groups.
Where to apply for VietGAP certification?
- The VIETGAP certification is obtained from certification bodies which are accredited by Department of Crop Production to operate over the national territory, and by Department of Agriculture and Rural Developments to operate at the province level. As of now, there are 17 of certification bodies in Vietnam.
Table 1. Differences among VietGAP, GlobalGAP and conventional production methods Production methods
VietGAP GlobalGAP Conventional method 1.Managing body
Accredited bodies of Government
Company representing retailers and suppliers
2. Hazards covered √ √ x
2.1 Food safety √ √ x
2.2 Environmental management √ √ x
2.3 Worker health and social welfare √ √ x
- suitability of production site √ √ x
- quality of soil and substrates √ √ x
- source and quality of seeds and
seedlings √ √ x
- source and quantity of fertilizer used √ √ x
- source and quantity of agrichemicals
used √ √ x
- quality of water used √ √ x
- suitability of harvesting and handling
produce √ √ x
- fruit treatment √ √ x
- fruit storage and transport √ √ x
- cleaning and sanitation √ √ x
- waste management and treatment √ √ x
Next, the attributes related to dragon fruit size, variety and flavor (ripeness) are identified for the visual, smell and taste components. Based on the common classification of dragon fruit sizes, three levels of the size attribute are considered: the small type if the fruit weighs less than 300 grams, the medium type if the fruit weighs about 300-500 grams, and the big type if the fruit weighs more than 500 grams. Regarding variety attribute, as of now, only two existing varieties of the dragon fruit are grown in Vietnam. They are white and red flesh. The red flesh is often considered as giving a better taste; however, they both have red skins. The flavor (ripeness) attribute is indentified as being made up of taste (sweetness, sourness, saltiness, bitterness). One of the first steps is to determine the way in which levels are to be presented. From own experience of eating dragon fruits and common sense, three levels of flavor (ripeness) attribute are defined: the semi-sourness, the semi-sweetness and the sweetness. Packaging is one of the important attributes to consumer choice is also considered. We only distinguish between non-packaged and plastic-packaged products. Price is required as part of the experiment to determine willingness to pay. Three price levels for fresh dragon fruit available in the market are included in the analysis. The attributes and their levels are described in Table 2.
Table 2. Attributes and their levels employed in the experiment
No-certificate (base) VietGAP certificate GlobalGAP certificate
Packaging Non- packaged product (base)
Plastic--packaged product Size
Small type (weight ≤ 300g) (base) Medium type (weight 300g-500g) Big type (weight ≥500g)
Variety White flesh (base)
Red flesh Flavor (ripeness)
Semi-sourness (base) Semi-sweetness Sweetness Price
(i.e. monetary attribute)
14.000 VND/kg (Base price)
21.000 VND/kg (50% above the base price) 28.000VND/kg (100% above the base price) 35.000 VND/kg (150% above the base price) (1USD ≈ 21.300 VND)
With six attributes associated with (3 x 2 x 3 x 2 x 3 x 4) levels respectively, there are 432 possible product combinations. Because the complexity associated with a larger number of choice sets in the design could affect participant decisions, we propose minimizing the number of choices using an orthogonal fractional factorial design. The experimental design, which was developed in accordance with Bliemer and Rose (2005), will be generated using Microsoft Excel. It is expected to formulate 30 orthogonal attribute combinations. The 30 alternatives will be randomly ordered to create 15 pairs of alternatives. Hence, a total of 15 choice sets will be selected for estimation. Because only the differences in attributes levels matter in logit models, the random order alternatives have the maximum differences with the original alternatives (Louviere et al., 2000).
These 15 choice sets will be randomly assigned into three versions. Every participant will receive one of the three versions, which have 5 choice sets. In each choice set, the participant will be presented with three alternatives, two alternatives refer to product profiles, and the third option refers to an opt-out alternative, which states “I would not buy either product indicated”. Table 3 presents an example of a choice set.
15 Table 3. An example of a choice set
Please indicate the product that you would buy if these products were made available to you in the marketplace. (Select only one)
Attributes Product 1 Product 2
Sweetness (fully ripe)
Price 21.000 VND/kg 28.000 VND/kg
I would BUY (Check √)
I would NOT BUY either product indicated
If you chose ‘would NOT BUY’, please tell us why:……… ………
No packaged Plastic-Packaged
White flesh Red flesh
Medium (300-500 grams)
Big (> 500 grams)
FRIENDLY Safe Fruit
1 kg 1 kg
Safe working conditions
To minimize the possible hypothetical bias, we also propose to use a cheap talk script to reduce hypothetical bias right before the choice questions.
A proposed cheap talk script
As you prepare to answer the next questions of your choices of dragon fruit, please keep in mind the following four things. First, keep in mind your household budget, you would be able to buy dragon fruit if your budget allows. Second, your family members may have different preferences to different attributes of dragon fruit. Third, aside from dragon fruit there are other alternative fruits that you can buy. Fourth, in previous surveys we have found that the choice of dragon fruit people made when shopping may be not the same as the choice they stated in a survey.
For this reason, could you please consider the dragon fruit attributes carefully before you make choice and please imagine you are ACTUALLY buying dragon fruit at a shop.
8.5. A description of the population and samples
Survey areas will be Hanoi, Ho Chi Minh and Danang cities. As mentioned earlier, metropolitan areas usually have a leading role in the food system transformation for the rest of the country. Accordingly, Hanoi in the north, Danang in the central, and Ho Chi Minh in the south of the country are chosen as these three cities are the largest cities in each region and presumably demand a very large amount of fresh dragon fruit.
Understanding consumers’ valuation of VietGAP product attributes in each region should be essential for designing effective food policies at the regional level. Hence, the sample design should be large enough to account for the geographical and climatic differences among the regions.
The choice modeling literature recommends at least 500 choices to allow for valid maximum likelihood estimations (Long, 1997). Since each participant will presumably be given 5 choice sets, there should be at least 100 participants as the minimum required for valid maximum likelihood. Taking analysis of regional differences into account, a sample of 300 participants at the least per study site will be required in this study. Therefore a total of 900 respondents/consumers will be interviewed. At least, 50% of the respondents are women.
Data will be gathered using face-to-face interview surveys. Food shoppers will be approached during their food shopping in supermarkets, fruit shops, fruit stalls or fruit hawkers. To explore consumers’ awareness of VietGap-certified products, food shoppers will first be presented with a card with some simplified possible definitions, ranging from an accurate definition to one which is totally irrelevant, and asked to indicate which definition, in their opinion, best explains the ‘VietGAP’ term. Then, all food shoppers will be given clear explanations of what they are paying for, including the VietGAP and the process of issuing VietGAP certification, the attributes and the levels of these attributes so that they could understand the choice set. To minimize the possible hypothetical bias, a cheap talk script will be used in reducing hypothetical bias right before the choice questions.
In addition, the study will undertake key informant interviews and focus group discussions. The key informant would be representatives of dragon fruit producers, sellers, officers of VietGAP program, and VietGAP certification institutions. Several focus group discussions will be organized: focus group discussions of producers, focus group discussions of sellers, and focus group discussions of consumers.
8.6. Procedures and techniques for processing and analysis of information
In the final step in implementing a choice experiment, a data set from valid samples will be prepared and a statistical analysis will be conducted on the basis of the data set by using discrete choice models. A conditional logit model (see section 8.2) will be used to determine the importance of fruit attributes and the willingness-to-pay for those attributes.
Regarding data preparation, the response variable, whether or not an alternative is chosen will be coded with 1 when chosen, and 0 otherwise. For each of the product attribute categories, one of the attribute levels is omitted and the others are assigned a value of 1 if chosen, and 0 otherwise. An alternative-specific constant (ASC), c, will also be created.
This will be coded as 1 if the participants choose one of the two product alternatives, and 0 if they choose the “I would not buy either product to the left”. In other words, ASC is a dummy variable that equals one when the opt-out option is not chosen. Price attribute will be entered as a continuous variable. The rest of the attributes will be dummy codes, with a base level of ‘no-certificate’ for the certification, ‘non- packaged product’ for the packaging, ‘small type’ for the size, ‘white flesh’ for the variety, and ‘semi-sour’ for the flavor (ripeness). Other socio demographic variables will also be included as explanatory variables.
It is quite common that the majority of fruit shoppers in Vietnam are female. Therefore, understanding gender differences in evaluating VietGAP product attributes should be useful for fruit producers, warehouse and retail store managers, marketing managers as well as food policy makers. This study intends to fully exploit this aspect.
9. Expected outputs and dissemination
There are the journal papers that could be produced by this study:
- Vietnamese Consumers’ Perception of Fresh Fruit safety in Vietnam: the Case of Dragon fruit
- WTP for Environmentally-friendly attributes of Dragon Fruit in Vietnam We will also produce at least two policy briefs
- How to Expand Outlets for VietGAP produces: The Case of Dragon Fruit - Consumers’ Attitude toward Food Certification
To disseminate the findings of the study, the research team will organize workshops with the participation of academe, policy makers, VietGAP institutions and other players in
the dragon fruit supply chain. The research team will also send its reports, policy briefs to relevant agencies.
10. Institutions and personnel
The researchers involved in this proposed project include Dr. Hoang Trieu Huy (appendix 1), the principal researcher, and Dr. Bui Dung The, co-researchers, and other researchers from Hue College of Economics.
Hue College of Economics (HCE) belongs to Hue University in Central Vietnam. HCE so far has received more than 7 individual and small research grants from EEPSEA. A number of lecturers have participated in different workshops, training courses organized by EEPSEA.
The duration of the proposed research project is 24 months. The time table below is based on the expectation that the project could start in July 2014
Research Activities Starting
month Note 1. Additional literature review of the research
methodology and the research problem and study sites
Jul. 2014 Aug. 2104 2 months 2. Survey instrument development and testing
and finalization Sep. 2014 Oct. 2014 2 months
3. Data collection Nov. 2014 Mar. 2015 4 months
4. Data entry and cleaning Apr. 2015 May. 2015 2 months 5. Data analysis and prepare interim report May. 2015 July. 2015 3 months 6. Data analysis and preparing final report Aug. 2015 Dec. 2015 5 months 7. Disseminations (workshops, writing papers,
policy brief) Jan. 2016 Jun 2016 6 months
19 12. Budget
The funding requested from EEPSEA for this research is estimated at VND 764,390,000 or about USD 35,886. The breakdown of expenses proposed for this research is given below (in Vietnamese Dong).
1. Direct Research Costs
a. Research Expenses
Research assistant (1 RA*VND2,500,000 /month*24 months) 60,000,000 Group discussions (3 groups * 10 persons/group * 3 sites) 27,000,000 Training of 10 enumerators (venue, food, resource person,
5,000,000 Enumerators (10 persons *7 days* VND400,000/person/day * 3
84,000,000 Travel costs of the research team and enumerators (tickets, hiring
60,000,000 Accommodation for enumerators and research team in 3 survey
sites (3sites * 7 nights*13 persons*VND300,000/person)
81,900,000 Gifts for respondents (key informants, focus groups, fruit shoppers)
in 3 sites (3 sites* 400 samples/site* VND50,000/sample)
60,000,000 Data entry, cleaning (1200 samples* VND10,000/sample) 12,000,000 Obtaining secondary data from related provincial departments 10,000,000 Obtaining journal articles, reports, internet access 5,000,000 Subtotal 404,900,000
Three seminars (venue, food, materials, accommodation, hiring
Publication of reports 5,000,000
c. Support Services
Photocopying, stationary, and other office consumables 10,000,000
Communication expenses and others 5,000,000
Subtotal 15,000,000 Total for Direct Research Costs 454,900,000
Lead Researcher (1LR*VND 4,000,000/month*24 months) 96,000,000 Research Members (2RM*VND3,000,000/ month*24 months) 144,000,000 Total for Remuneration 240,000,000 3. Administrative Cost (10% of project cost for HCE) 69,490,000 TOTAL (1+2+3) 764,390,000 Budget requested from EEPSEA: VND 764,390,000 or about USD 35,886
Conversion rate: USD1= VND 21,300
APPENDIX 1. CURRICULUM VITAE OF THE PRINCIPAL RESEARCHER
Full name: HOANG TRIEU HUY
Nationality: Vietnamese Sex: Male
Date of birth: 22 June 1974 Marital status: Married Languages: Vietnamese (native), English (fluent)
PRESENT OCCUPATION University lecturer
Department of Economics and Development Studies Hue College of Economics, Hue University
100 Phung Hung Street, Hue City, Vietnam
Phone: +84 54 3538332; Fax: +84 54 3529491 Mobile: +84 915 19 19 59
Email: email@example.com EDUCATION
Ph.D, Agricultural economics and International rural Development, Lincoln University, New Zealand, 2013.
M.Sc, International and Development Economics, the Australian National University, Australian, 2006.
B.Sc, Agricultural Economics, Hue University of Agriculture and Forestry, Hue City, Vietnam, 1996.
RECENT PROJECTS INVOLVED
• “Solutions to develop agricultural insurance in Mekong delta, Vietnam”, research member, funded by The Ministry of Education and Training of Vietnam, 2014-2016.
• “Agriculture Growth, Livelihood Diversification and Poverty in the Rural Central Vietnam”, research member, managed by VERN, funded by IDRC, Vietnam, 2007- 2009.
• “Early Warning System for Storm Management in A Coastal Community: Role of Information and Local Institutions”, project holder, funded by EEPSEA, 2007- 2008.
• “Study On the Options for Financing Conservation in Bach Ma National Park and Improving Sustainable Livelihoods in Critical Areas Within the Buffer Zone”, consultant, the project funded by WWF Vietnam, 2006-2007.
RECENT CONFERENCES AND WORKSHOPS ATTENDED
• “On-Farm Impacts of Environmental Policy”, The NZAES conference held by New Zealand Agricultural and Resource Economics Society, Christchurch, New Zealand, 28 – 30 August 2013.
• The VERN II workshop on “Vietnam Economic Research Network: The 2nd Networking Workshop”, held by International Development Research Center and Center for Analysis and Forecasting, Sa Pa, Vietnam, 14-17 April 2008.
• 28th EEPSEA Biannual Workshop on “Water and Sanitation Planning for Developing Countries” held by EEPSEA at Kuala Lumpur, Malaysia, 13-19 November 2007.
• The “Research and Policy in Development” workshop held by CAF & Oversea Development Institute of London at Hue City, Vietnam, 12- 13 July 2007.
• The “Poverty and Poverty Reduction in Vietnam During Period of 1993-2004”
workshop held by Vietnam Academy of Social Science at Danang City, Vietnam, 21 March 2007.
RESEARCH PAPERS In English
Huy, H.T, Lyne, M., Ratna, N.N. and Nuthall, P., (2014). The rental market for cropland in Vietnam: Efficiency, equity and transaction costs. Forthcoming in Australian Journal for Agricultural and Resource Economics.
Huy, H.T, Lyne, M., Ratna, N.N. and Nuthall, P., (2013). Efficiency and Equity Impacts of the rental market for cropland in Vietnam and sources of transaction costs impeding the market. Conference Papers, New Zealand Agricultural and Resource Economics Society Conference, Lincoln University, Christchurch.
Huy, H.T, (2009). Early Warning System for Storm Management in the Coastal Commune of Hai An: Role of Information and Local Institutions. Technical Reports. EEPSEA.
Tam, B.T, Huy, H.T, (2009). Agriculture Growth, Livelihood Diversification and Poverty in The Rural Central Vietnam. Research Reports. VERN-IDRC.
Othman, M.S, Nga, D.T, Huy, H.T, (2007). Opportunities for sustainable financing in Bach Ma National Park and to assess some possibilities for livelihood improvements within the buffer zone of the park. Report Paper. WWF Vietnam.
Huy, H., Khoi P. D., (2014). Factors influencing the accessibility of farming households to formal credit in Hau Giang Province, Vietnam. Forthcoming in Hue University Journal of Science.
Huy, H., Khoi P. D., Nguyet, P.T.A, (2014). Factors affecting participation of rice farmers in rice crop insurance program at Dong Thap. Forthcoming in Can Tho University Science Journal.
Khoi P. D., Huy, H., Duong, T.Q, (2014). Determinants of credit rationing of urban households in Can Tho City. Forthcoming in Hue University Journal of Science.
Huy, H., Khoi P. D., (2014). Agricultural Risks and Insurance: The Implementation of the Pilot Program on Agricultural Insurance in Vietnam. Forthcoming in Journal of Development Economics.
Hoa, H.H, Huy, H.T, (1999). Agricultural Land Tax and Development of Farm Households in Thua Thien Hue Province. In “The 1995-1999 Record of Scientific Research on Economics and Business Administration”. Agriculture House Publisher.