Sustainable Development of Cassava Value Chain through the Promotion of Locally Sourced Chips

Abstract

The benefits of using homegrown inputs in the production of processed agricultural products have been recognized, for example, in protecting the environment and generating local income. In Thailand, local fresh cassava can replace imported potatoes in the production of chips. However, chips made from local cassava are not widely available on the market. This market access difficulty could be because of insufficient information about consumers’ preferences. This study thus aims to address the factors determining people’s consumption decisions on locally sourced snacks in the case of chips made from Thai cassava. These factors include, for example, price, gender, generation, education, occupation, purchasing frequency, product attributes, nutrition, food safety, and an understanding of cassava chips. The questionnaire used was developed to elicit data related to these factors. The statistical analysis is undertaken by the probit model and marginal effect. The results of three estimated models with 19 independent variables show that the effects of generation and occupation significantly indicate the probability of preferring cassava chips. Having a higher level of education and coming from the northeastern part of the country seem to determine consumers’ preference for local cassava content. Moreover, a better understanding of the related issues implies a higher chance of favoring chips from the local cassava. Those findings would provide useful information for entrepreneurs and government agencies in promoting locally sourced chips, and further develop the higher value of the cassava supply chain.

1. Introduction

Recently, the promotion of food products made from local produce has acquired the public’s attention as a means of sustainable development through at least two channels. First, using local inputs in the manufacturing of food products can generate income for farmers, create local businesses, and, at the same time, save the country’s foreign currency by substituting imported raw materials. Moreover, it can also help improve the environment by cutting down on international transportation.

Thailand, as an agricultural-abundant country, can practically encourage the use of local inputs in the manufacturing of processed food products. Among the country’s major crops, cassava has a high potential because it can be used as a substitute in the production of chips, one of the most popular snacks, usually made from potatoes. There are two types of cassava. The first variety is mainly used in industrial production owing to its bitter taste, and is normally processed into pellets, starch, glucose syrup, and alcohol, for example. The second one is suitable for direct human consumption, cassava chips included, and has been planted on a large scale in some countries, where the products are commercialized. Several studies have been carried out to study how to further the development of cassava chips, especially in the remote villages of Indonesia and Malaysia, for example, Sutiknjo et al., Khaeruman and Hanafiah, Fitriani et al., Panggabean et al., Amir et al., and Rani et al. [1,2,3,4,5,6]. This simply shows the economic opportunity of cassava chips in this region to generate income for people in rural areas. However, in Thailand, cassava chips are not widely known, nor have they been widely produced and commercialized in the local market. At the same time, to the best of our knowledge, there is no study focusing on the commercial opportunities of these products. Further, the possibility of producing chips from local fresh cassava is visibly rising, as Thai growers, in 2016, successfully improved the cassava varieties. A new breed, the Piroon 4, has a sweeter taste, a softer texture, and reduced splintering, making it more appropriate as a raw material to process into food products [7]. This improvement in the cassava varieties can increase the chance of Thai cassava being processed into a varied type of snack food, especially chips.

As mentioned above, the market for cassava chips in Thailand is small, which consequently suggests production on a local business or small enterprise scale. On the contrary, potato chips, with their much wider market, are produced on a large scale. The price of fresh cassava tubers (THB 2.7–3.5 per kg) is lower than the price of fresh potatoes (THB 12–15 per kg.) by about 4.5 times. However, the price of standard certified cassava chips is about THB 30 per 40 g (or THB 52.50 per 75 g), while that of potato chips is about 29–30 per 75 g. The price difference between the two chips is largely because of the economies of scale in the production of potato chips. This also highlights the potential competition for cassava chips, at least in the Thai market.

More interestingly, the snacks market in Thailand has grown quite consistently, with a 4.55% increase over a three-year average, from 2017–2019. Potato chips take around 32.20% of the market share with a value of THB 11,992 billion (USD 1 = THB 38.12, as of 19 October 2022) in 2019. In terms of the potato chip market, Lay’s, an American brand, manufactured by Frito-Lay Thailand Co., Ltd. (Bangkok, Thailand), accounts for around 75.00% [8]. The size of the chip market and its expansion indicate the opportunity for cassava chips to become a potential alternative to potato chips. Furthermore, this implies a large demand for potatoes, which is not a plant native to Thailand. As a result, a substantial amount of imported potatoes is required. Currently, potato cultivation in the country requires the importation of seed potatoes from Australia, the United Kingdom, and the Netherlands, whose value has grown 23.26% over three years, from 2017–2019. Moreover, the amount of fresh potatoes imported is still not sufficient for use within the country, which is, therefore, left to depend on imported fresh and chilled products from other countries, such as Canada, China, and Germany, at an increasing rate of 19.93%. The value of imports of both seed potatoes and fresh and chilled potatoes has risen steadily from about USD 28.03 million in 2017, to USD 31.82 million in 2019. The difficult situation of heavy dependence on imported potatoes results in the elevation of the price of potato-related goods and the unnecessary outflow of foreign currency. This emphasizes the importance of the introduction of local fresh cassava in the manufacturing of snacks, especially chips. It can help decrease the avoidable importation of fresh potatoes. In addition, the replacement of imported potatoes with local cassava can improve the food safety of chips which can contain gluten and a carcinogen, so-called acrylamide, if made from potatoes [9,10].

Acrylamide is a carcinogenic substance that arises from a chemical reaction during the high-temperature processing of potatoes [11]. Potato chips, along with French fries, other potato-based products, cereal-based products, as well as coffee, are among the foods containing a high level of acrylamide [12,13]. Comparing these two types of chips, those from cassava and those from potatoes, it appears that, generally, cassava chips enjoy a lower level of acrylamide formation than potato chips do. Arisseto et al. [14,15] investigated the level of acrylamide contamination in different foods in Brazil. They found that the highest level can be found in potato chips. In contrast, the lowest levels are found in cassava chips. Similar results can also be found in Egypt [16]. In addition, the level of acrylamide in cassava chips can be minimized, if not eliminated, by using the appropriate processing techniques. Fadilah et al. [17] found that longer heating time and higher cooking temperature are more likely to lead to a higher level of Acrylamide formation. They present that the heating time of 15 min and the cooking temperature of 210 °C produce the highest level of acrylamide in cassava chips. However, the most suitable production technique which can result in the best chemical, physical and sensory qualities of cassava chips is frying at 160 °C for 2 min [18]. Another cooking method found to produce the best cassava chips is vacuum frying of blanched cassava at 130 °C [19]. Thus, it is unlikely for acrylamide to be intensively formed during the manufacturing of cassava chips. To a large extent, consumers are less likely to encounter contamination with a massive amount of acrylamide when consuming cassava chips.

All in all, having local content such as fresh cassava in the manufacturing of snack food products can help improve the value added to their supply chain, the country’s international finance position, farmers’ welfare, individuals’ consumption choices, and consumers’ health by lowering the risk of exposure to cancer. Other good means causes supporting locally sourced inputs, especially cassava, can also be promoted. Such causes included, for example, being healthier, being fresh, being more environmentally friendly, and accommodating local farming entrepreneurs [20]. Moreover, promoting locally produced products can be considered home-bias or hometown pride [21].

In spite of the gains from using local inputs, cassava chips are not yet widely manufactured and consumed in the country. Thus, their marketability, especially price determination, can be difficult. Determining the factors affecting consumers’ behavior toward these products should increase the market opportunity for cassava chips. Moreover, the knowledge learned should help to create a sustainable cassava supply chain in Thailand. Cassava chips can be crucial to the country, both as an employment generator and foreign exchange earner, and provide stable ways of life for farmers. What is revealed here could fill the research and market gaps and complete the pathway between different stakeholders on the production front in Thailand’s cassava supply chain from backward through forward linkages. The increased processing of fresh cassava could add value to the produce and improve farmers’ quality of life. This could offer more opportunities for cassava farmers to diversify their sources of income, from such a traditional one as selling their crop to a tapioca starch factory to processing it into chips at the higher market prices. Putting it in a broader picture, this could be healthy for the agricultural sector in general as farmers’ income will be driven more by market-based mechanisms and less reliance on the government’s market intervention schemes such as the Cassava Income Insurance Program.

In terms of forward linkage, with the processing of cassava into chips in place, small local entrepreneurs can take part in this venture, generating more income for the community. At the same time, the manufacturing of chips can switch from using potatoes as raw materials to cassava, depending less on imported inputs, and helping save the country’s foreign currency in the process. On the consumption front, having a clearer idea about individuals’ preference for chips made from local cassava could permit producers to respond better to customers’ demand and attract them to buy more of these products. Consequently, the decrease in consumers’ health risk and increased fresh cassava prices eventually lessen the weight on the government’s related budget. In the long run, promoting cassava chips as a locally sourced product could lead to sustainability in terms of both the Thai people’s quality of life and government stability.

The reasons as to why consumers prefer locally sourced cassava chips can be explained by their consumption and buying behaviors. These behaviors are determined by passion, habit, and individual traits. Investigating consumers’ behaviors often involves looking into personal, emotional, and social factors. These factors can, explicitly and implicitly, affect consumers’ purchase decisions for certain products. Kotler and Keller [22] identify four marketing attributes, namely product, price, place, and promotion, the so-called four Ps, together with economic, political, and cultural factors, as the determinanst of consumers’ decisions. Furthermore, consumers’ consumption and buying decisions have been empirically studied on different occasions. Susilawati et al. [23] study the case of Indonesian consumers’ attitude toward snack buying. Their result shows the factors that, according to the marketing mix approach, can stimulate consumers’ behavior. The factors include product quality, price, brand, service, and warranties.

Nørgaard et al. [24] explore the plans of 600 teenagers, aged between 9 to 16 years old in Denmark, to purchase healthy snacks or not. Concerns about price and health, snack anxiety, peer pressure, social activities, and word of mouth are the potential determinants of their demand for healthy snack products. The causal linkage assessments reveal the dominance of social factors as the determinant of which new snack is favored, while health awareness has a limited and implicit effect on the purchase intent. Meanwhile, Damen et al. [25] find that, despite being in the same nation, geographic differences in maternal discretion arise. Mothers’ consideration for snacks for children aged 2–7 is divided into four main groups relating to different issues: health, children, time, and products. Mothers from Northern Italy seem to be more concerned about health than their southern counterparts. The main theme associated with the product shows that mothers in the South pay more attention to snack brands than those from other regions. These works lead to an open range of investigations on people’s concerns for food safety, standard, and quality before deciding to buy. Shin et al. [26] point out the issues shaping persons’ willingness to pay a higher price for dining at a locally sourced restaurant. This work also studies whether health consciousness and community attachment influence willingness to pay. The findings indicate that they positively affect mindset, subjective norm, and behavior toward visiting a locally sourced restaurant, which eventually motivates a desire to pay extra.

Additionally, Dinushika and De Silva [27] investigate consumers’ features, their consumption preferences, and their beliefs about the chips market. Most of them give importance to the taste and brand of chips. Forty-six percent know about the variety of chip brands on the market. The Wilcoxon label rankings test reveals that product characteristics such as trusted brand names, air tightness, price, good quality, overall consumption, taste and new products are significant. Meanwhile, Kongstad and Giacalone [28] study the importance of labeling and how it can influence consumers’ attitudes toward salt reduction patterns for processed potato products. The results show the significance of clear labeling on consumers’ behavior. Their finding inclines to the work of Dinushika and De Silva [27]. Nevertheless, it was found that promotional programs, wide availability, and attractive appearance of products do not matter to consumers, unlike the results of the work of Susilawati et al. [23]. This leads to inconclusive results, demonstrating the need to conduct further analysis on marketing mix factors.

To measure the preference of consumers for locally sourced cassava chips, the contingent valuation method can be applied by asking surveyed individuals to show their willingness to pay more for chips made from fresh Thai cassava. Numerous works were carried out to analyze the willingness to pay for locally produced products. For instance, Darby et al. [20] identify the reason for buying locally grown products as their freshness and home-bias consumption, and suggest that consumers primarily pay more for the freshness. This study shows a strong foundation for the possibility of a niche market for locally grown product producers. It also finds that consumers are willing to pay a premium price for locally produced foods, and that the size of premium price differs based on socio-economic factors.

Some studies also focus on consumers’ preferences and willingness to pay for specific products produced in communities in the United States. Brown [29] examines consumers’ preferences for locally grown food in Southeast Missouri. The study finds that people pay attention to quality and freshness when buying fresh produce and believe that local produce is of a higher quality and cheaper. Consumers involved in environmental conservation activities have higher education and income levels, and are more willing to pay extra for local products. Individuals whose backgrounds are from the farming sector tend to have a higher willingness to pay for locally produced items. Taking this into account, to market local products efficiently, the issue of quality, freshness, and price competitiveness should be emphasized. A more recent work centers on the effect of knowledge about local products on consumers’ willingness to pay and their views about the quality of one type of broccoli cultivated in California, and two recently developed homegrown varieties. The work shows the importance of having access to information. Without information, respondents rank the look and flavor of the Californian type higher than the two local varieties. However, once the origin of the two broccoli types is revealed, consumers’ evaluation of these varieties increases markedly. The results indicate consumers’ willingness to pay an additional price for local broccoli. They also provide evidence that information about local products positively influences consumers’ willingness to pay and their views on the quality of local foods [30].

Apart from the case of fresh produce discussed earlier, Jensen et al. [31] study the issue of using local inputs in processed food products and find similar results. They examine factors affecting people’s willingness to pay an additional price for locally produced hard apple cider and the price level. They look into various aspects of hard apple cider and consumer characteristics, as well as their relationship with consumers’ behavior. This study finds that consumers are ready to pay an additional price of USD 3.22 for locally produced hard apple cider on top of the USD 6.99 paid for standard products. Such factors as a favor for homegrown goods, suburb expansion, and shopping places can determine the amount of extra price. Furthermore, products’ qualities, for instance, sweetness and dryness, play a role in consumers’ buying decisions. The studies of Brown, Fan et al., and Jensen et al. [29,30,31] indicate consumers’ steady support for food products involved with local inputs through their clear preferences and willingness to pay a higher price.

From the evidence shown by earlier works, chips made from fresh local cassava stand a chance of succeeding in the local market. Moreover, the expansion in the size of the chip market with an increasing trend provides a high potential for cassava chips to thrive in the market, being a substitute for potato chips. Nonetheless, because of their innovative nature, information necessary for their marketing is limited, making it difficult for the product to be promoted to wider consumers. Hence, it is worth focusing on these products.

To realize the marketing opportunities, factors affecting consumers’ behavior toward local-made cassava chips need to be investigated. Therefore, the research question posed here is what the necessary conditions for promoting chips made from local cassava are. If cassava chips are successfully and commercially promoted, they can improve Thailand’s foreign exchange and provide more opportunities for farmers as well as develop the higher value cassava supply chain. Therefore, this study aims to examine consumers’ behavior toward food products made from local produce, with a special emphasis on cassava chips. The central hypothesis aims to prove whether behavioral bias strongly explains consumers’ buying decision of snacks with local ingredients. The differences in experience, attitude, perception, and personal characteristics are the key sources of this bias, which drives individuals to disclose their satisfaction. Thus, the hypotheses on the explanatory power of those related variables on surveyed individuals’ preference for chips made from fresh Thai cassava are tested.

The rest of the paper is organized as follows. Section 2 explains the materials and methods employed in this study. Section 3 presents its findings. Section 4 discusses the results from different perspectives. Additionally, the last section concludes this paper.

2. Materials and Methods

The data used here relate to the decision making of consumers and preference to pay extra for processed snacks using local ingredients, in the case of cassava chips, based on the price of potato chips generally available in the market. Because the price for the feature “made from local raw materials” of these products is not available in the market, it is then appropriate to apply the concept of non-marketable goods in this case of study. The contingent valuation method (CVM), frequently employed in environmental valuation, is reasonably applied to provide a framework for collecting data [32,33] by letting the respondents assess the value of their satisfaction on locally sourced cassava chips in terms of money.

Technically, the number of samples surveyed is based on Yamane’s criteria [34] that developed a statistical guideline. The method for determining the statistically appropriate sample size in this study is based on the table of Taro Yamane at a conventional 95% confidence level. The formula for calculating a sample size of an unknown population uses no fewer than 385 samples. Moreover, Calia and Strazzera [35] investigate the efficacy and bias resulting from the CVM technique, which shows no significant bias in the case of 250 or more samples. In this study, the surveyed data are hence collected from a sample of 1000 Thai respondents aged 18 years or older as a selective group. They are required to have experiences of purchasing/consuming chips as they could truly reveal their preference in monetary terms.

Moreover, the sampling quota is also adopted by equal dividing of the total respondents, resulting in 250 samples for each initial bid price. This is to reduce the probability of bias caused by the dominating effect of a large number of one particular group of samples. To elicit information from those 1000 surveyed individuals as the representative of Thai consumers aged 18 years and over, the questionnaire with four sections is developed based on Dukeshire et al. [36]. These questionnaire surveys are strictly conducted under the endorsement of the Kasetsart University Research Ethics Committee (Certificate of Approval COE No. COE65/037). The survey took place in April–May 2022, covering five country regions. Participants were randomly asked to voluntarily complete the questionnaire to obtain relevant information, including socio-economic characteristics, consumption behavior, related attitude and perception, and preference for local content in the case of chips made from fresh Thai cassava. Regarding the sampling method, the demographic information of samples should be appropriately distributed according to the statistical principle, which will be discussed later.

Respondents are asked to answer a yes–no question to indicate their willingness to pay extra for cassava chips. Surveyed individuals’ preference is a self-measurement, assessed under the concept of contingent method application, showing their marginal willingness to pay extra in addition to the market price of potato chips (THB 29 per 75 g). Before the survey, the questionnaire was pretested to ensure its reliability. An analysis of Cronbach’s Alpha coefficients is used with an acceptable level of 0.7 for primary research [37]. Cronbach’s Alpha indicates that the overall test result of all questions was 0.87, which implies the questionnaire’s reliability. Besides, the pretest also collects a range of marginal values of willingness to pay to set the initial market pricing. According to the pretest, these initial market prices are divided into four price levels, called initial bid prices. The questionnaire is thus separated into four sets based on initial prices. If surveyed individuals first showed their willingness to pay more for locally sourced products, they would be further asked to answer the yes–no question upon the initial price, as shown in

Table 1. Probability of additionnal willingness to pay categorized by four starting prices in the case of closed-ended questions.

Initial Price (THB) *		Probability of Preference	Dichotomous Choices
2		Pr(Yes)	1
		Pr(No)	0
5		Pr(Yes)	1
		Pr(No)	0
8		Pr(Yes)	1
		Pr(No)	0
10		Pr(Yes)	1
		Pr(No)	0

Note: * is the default currency used in the actual survey, which is based on the results of the willingness to pay from the questionnaire pretest.

. This would enable individuals to monetarize their preference for locally sourced products in the case of chips made from local cassava.

There are two possible answers for a total of two situations, which are “willing” (Yes) and “not willing” (No). The possibility of additional willingness to pay is stated as Pr(Yes) and Pr(No) with a mark of one and zero, respectively. Therefore, willgness is expressed as nothing but a dichotomous variable.

The binary analysis is employed to analyze factors determining the chance that Thai consumers are willing to pay more for the use of local raw materials in the case of chips made from fresh Thai cassava root. Because of the non-linear form of logit or probit equations, a generalized estimation is needed to transform to the logarithmic linear equation form, which is expressed as follows:

$$\ln \left(\frac{\mathrm{Pr}\left({\mathrm{y}}_{\mathrm{i}}=1\right)}{\mathrm{Pr}\left({\mathrm{y}}_{\mathrm{i}}=0\right)}\right)=\mathrm{a}+{\mathrm{b}}_1{\mathrm{x}}_{1,\mathrm{i}}+{\mathrm{b}}_2{\mathrm{x}}_{2,\mathrm{i}}\dots +{\mathrm{b}}_{\mathrm{k}}{\mathrm{x}}_{\mathrm{k},\mathrm{i}}+{\mathsf{\epsilon }}_{\mathrm{i}}$$

(1)

or

$${\mathrm{y}}_{\mathrm{i}}=\mathrm{a}+{\mathrm{b}}_1{\mathrm{x}}_{1,\mathrm{i}}+{\mathrm{b}}_2{\mathrm{x}}_{2,\mathrm{i}}\dots +{\mathrm{b}}_{\mathrm{k}}{\mathrm{x}}_{\mathrm{k},\mathrm{i}}+{\mathsf{\epsilon }}_{\mathrm{i}}$$

(2)

The variable y is a dichotomous variable. If respondents i express their willingness to pay more for locally sourced products in this particular case, the variable would then be marked as “1”; otherwise “0”. x_k represents the set of relevant factors at k = 1, 2, …, 19, whose definitions are elaborated in Table 2. a and b_k are constant terms and the interested coefficients concerning 19 relevant variables. The symbol ε denotes the disturbance term. The estimation of b_k is undertaken by the maximum likelihood estimation (MLE) method. It must be mentioned that the estimated coefficients obtained from Equation (2) cannot be directly interpreted. The calculation of marginal effect, which is the outcome of the partial derivatives of the variable on the likelihood of an event occurrence, is required to circumvent this drawback.

Moreover, the marginal effects provide the result in terms of probability improvement, which is supplementary information for policy recommendations in this aspect of economics and business. The change in the value of some characteristic variables for one unit will affect the creati opportunities in which Thai consumers would be willing to pay more for locally sourced products, at least in this case. It must be noted that all those 19 variables are operationalized based on the information retrieved from the questionnaire. For the variables whose information is in the qualitative form, they are converted to quantitative variables by marking zero or one, a so-called dichotomous variable. Meanwhile, the questions of some specific issues with rating scale answers are calculated in the natural logarithmic form of the grand mean. The yes–no questions used to measure understanding of related issues are in the ratio of the number of questions with correct answers to the total number of questions and the percentage form. The detail of the description is again shown in Table 2.

Table 2. Description of relevant variables.

Variable	Description
y	This dependent variable is the dichotomous variable and indicates the respondent i’s preference for chips made from the local raw material, fresh cassava. If, upon the proposed four different initial prices (THB), the respondent shows preference to pay an additional price for cassava chips, it is marked as “1”; otherwise “0”.
Relevant variables (xk) are categorized into four groups regarding price, personal characteristics, attitude, consumption behavior, buying decisions, and perception and understanding.
PRS	The initial price (THB) of 4 values (2, 5, 8, and 10) is used to query the respondents. According to the law of demand, the initial price should negatively relate to a willingness to pay extra for using local content in the case of chips made from fresh Thai cassava root.
Item of Socio-economic Characteristics
GEN	Despite being considered a ponderable, fearful attitude and not an adventurous trait, females generally take responsibility for house matters, including how the family consumes. They are mostly interested in surroundings, health, and sensitive issues. They would possibly tend to favor this local product with supply chain and health advantages. The gender of the sample is thus divided into two groups: female = 1, others = 0.
LNUM	The number of people in the household in the natural logarithmic form is expected to have a positive relationship with a willingness to pay more for chips made from fresh Thai cassava, according to the work of Muhammada et al. and Chancharoenchai and Saraithong [38,39].
AGEZY	The age grouping is based on generations. This variable is designed to capture the generation effect that may be influenced by the shared social, economic, and historical context. Those differences in conditions and situations may affect lifestyle, and influence personality traits, attitudes, and opinions of those individuals [40]. Surveyed individuals whose age 18–20 years (GENZ) and 21–37 years (GENY) = 1, others = 0.
EDUC	This variable reflects how education could explain the preference for local produce in this particular case. Higher education is commonly expected to provoke awareness of the advantages of using domestic raw materials in production. The correlation between this variable and willingness to pay extra for chips made from fresh Thai cassava root should be positive. To address the explanatory power of the level of education on the opportunity of willingness to pay more, a bachelor’s degree/vocational certificate and higher = 1, others educational levels = 0.
OCC	Respondents’ occupations are divided into six categories: student, private company employee, civil servants/state enterprises/other government agencies, business owner, freelance, and others. To assess the impact on the likelihood that respondents with such occupations would be willing to pay, the occupation is divided into three groups to capture their reflection of attitude, work environment, and related knowledge, as follows: - OCCG: government officials, state enterprises, and other government agencies = 1; otherwise = 0. - OCCPB: private company employee and business owner = 1; otherwise = 0. - OCCS: student = 1; otherwise = 0.
INCL	The interest in testing the hypothesis of purchasing power is taken by how respondents’ monthly income determines their preference for paying more for chips made from fresh Thai cassava. Monthly income is classified into two groups based on data from the National Economic and Social Development Council (NESDC). It reports a per capita income of Thai people at THB 248,468 in 2022, which has an average monthly income of THB 20,705.67. Regarding the NESDC database, this variable is therefore set to equal to 1 when their income was THB 20,000 per month or more, otherwise is 0.
RACET	Respondents’ race was divided into four ethnic groups, Thai-Thai, Thai-Chinese, Thai-Muslim, and others, to analyze the impact of ethnicity. The differences in prospective nationality can reflect the identity and attitude of respondents on their standpoint about locally sourced products. The nationality is divided into two groups: Thai-Thai = 1, others = 0.
HOMNE	The respondents’ domiciles are divided into five regions (Northern, Northeastern, Central, Southern, and Bangkok and Metropolitan). Because Thailand’s Northeastern region has the largest area of cassava farming, and surveyed individuals who domiciled in that region are more likely to be willing to pay more in the case of this study. Therefore, the hometown is a dichotomous variable: Northeastern region = 1, others = 0.
Item of perception and understanding, and Thai national sentiment
UND	Awareness and understanding about local products, breeding and processing of fresh cassava, and the standard and safety of chips are measured by yes–no questions of six issues. Correct answers are marked 1 point, while wrong answers are marked 0 point. This variable is presented in percentage form, where a higher percentage implies better awareness and understanding. Realizing the advantages of using local ingredients and cassava should encourage more chances to show a willingness to pay more.
LNNPRI	The national pride is measured by 14 questions about being Thai, environmental positions, food safety and standard, agricultural and processed products, and innovation and agricultural technology. The survey respondents are asked to rate five-level of opinions based on the Likert scale, from strongly agree = 4 to cannot decide or not sure = 0. The higher mean score of those 14 questions illustrates more sentimental value in the Thai nation. The higher level of this sentiment would hypothetically show support for locally sourced products at least in the case of cassava chips.
Item of consumers’ behavior in buying snacks
FREQ	The frequency of buying snacks for own self and household consumption is a dichotomous variable. It is marked as 1 when purchased daily or at least twice a week; otherwise it is given as 0.
ISHOP	The place where respondents buy snacks is a dichotomous variable. If respondents always buy from local grocery stores, they may be highly likely to pay more for products made from local ingredients and, thus, Thai cassava chips. Respondents’ buying often from local grocery stores is marked as 1; otherwise = 0
Item of shopping behavior is influenced by product composition and attributes, awareness of nutritional value, standards and safety, and external factors and components.
LNBSIGN	Six questions related to opinions about the importance of the factors affecting the type of snacks respondents choose to buy are asked. Additionally, surveyed individuals are to indicate a 5-point Likert scale rating, cannot decide or not sure = 0, strongly disagree = 1, disagree = 2, agree = 3, and strongly agree = 4. This variable is the natural logarithmic of an average value of these six questions.
LNBBUY	Similar concept to BSIGN, surveyed individuals express their opinions on eight questions based on a 5-level Likert scale. Those questions reveal how product attributes affect purchasing and/or consuming behavior toward snacks. To analyze the explanatory power, the rating scores are recalculated into the natural logarithmic average.
LNBNUR	It is the natural logarithmic mean scores of opinions about the importance of nutrition. To assess those of opinions, surveyed individuals are asked to identify five rating scores on the Likert scale of six related questions. The higher the score they rate, the more importance they attach to these issues.
LNBSAF	It represents the natural logarithmic mean of seven questions that correspond to the opinion level on food safety. Respondents are asked to give a score based on the 5-point Likert scale. This variable captures how confident Thai consumers feel about Thai raw material products. The relationship between willingness to pay and this variable can be positive or negative, which could lead to the belief about locally produced goods as being normal or inferior products, respectively.
LNBFAC	This variable is the natural logarithmic value of the grand mean of four questions about how surveyed individuals rate the importance of taste, packaging appearance, manufacturer/brand, and advertisement as influencing their buying decisions.
LNBCONT	This is the grand mean of seven questions related to opinions about the importance of product components, given based on a 5-level Likert scale from least to most agreement. It is in the natural logarithmic form.

Table 2. Description of relevant variables.

Variable	Description
y	This dependent variable is the dichotomous variable and indicates the respondent i’s preference for chips made from the local raw material, fresh cassava. If, upon the proposed four different initial prices (THB), the respondent shows preference to pay an additional price for cassava chips, it is marked as “1”; otherwise “0”.
Relevant variables (xk) are categorized into four groups regarding price, personal characteristics, attitude, consumption behavior, buying decisions, and perception and understanding.
PRS	The initial price (THB) of 4 values (2, 5, 8, and 10) is used to query the respondents. According to the law of demand, the initial price should negatively relate to a willingness to pay extra for using local content in the case of chips made from fresh Thai cassava root.
Item of Socio-economic Characteristics
GEN	Despite being considered a ponderable, fearful attitude and not an adventurous trait, females generally take responsibility for house matters, including how the family consumes. They are mostly interested in surroundings, health, and sensitive issues. They would possibly tend to favor this local product with supply chain and health advantages. The gender of the sample is thus divided into two groups: female = 1, others = 0.
LNUM	The number of people in the household in the natural logarithmic form is expected to have a positive relationship with a willingness to pay more for chips made from fresh Thai cassava, according to the work of Muhammada et al. and Chancharoenchai and Saraithong [38,39].
AGEZY	The age grouping is based on generations. This variable is designed to capture the generation effect that may be influenced by the shared social, economic, and historical context. Those differences in conditions and situations may affect lifestyle, and influence personality traits, attitudes, and opinions of those individuals [40]. Surveyed individuals whose age 18–20 years (GENZ) and 21–37 years (GENY) = 1, others = 0.
EDUC	This variable reflects how education could explain the preference for local produce in this particular case. Higher education is commonly expected to provoke awareness of the advantages of using domestic raw materials in production. The correlation between this variable and willingness to pay extra for chips made from fresh Thai cassava root should be positive. To address the explanatory power of the level of education on the opportunity of willingness to pay more, a bachelor’s degree/vocational certificate and higher = 1, others educational levels = 0.
OCC	Respondents’ occupations are divided into six categories: student, private company employee, civil servants/state enterprises/other government agencies, business owner, freelance, and others. To assess the impact on the likelihood that respondents with such occupations would be willing to pay, the occupation is divided into three groups to capture their reflection of attitude, work environment, and related knowledge, as follows: - OCCG: government officials, state enterprises, and other government agencies = 1; otherwise = 0. - OCCPB: private company employee and business owner = 1; otherwise = 0. - OCCS: student = 1; otherwise = 0.
INCL	The interest in testing the hypothesis of purchasing power is taken by how respondents’ monthly income determines their preference for paying more for chips made from fresh Thai cassava. Monthly income is classified into two groups based on data from the National Economic and Social Development Council (NESDC). It reports a per capita income of Thai people at THB 248,468 in 2022, which has an average monthly income of THB 20,705.67. Regarding the NESDC database, this variable is therefore set to equal to 1 when their income was THB 20,000 per month or more, otherwise is 0.
RACET	Respondents’ race was divided into four ethnic groups, Thai-Thai, Thai-Chinese, Thai-Muslim, and others, to analyze the impact of ethnicity. The differences in prospective nationality can reflect the identity and attitude of respondents on their standpoint about locally sourced products. The nationality is divided into two groups: Thai-Thai = 1, others = 0.
HOMNE	The respondents’ domiciles are divided into five regions (Northern, Northeastern, Central, Southern, and Bangkok and Metropolitan). Because Thailand’s Northeastern region has the largest area of cassava farming, and surveyed individuals who domiciled in that region are more likely to be willing to pay more in the case of this study. Therefore, the hometown is a dichotomous variable: Northeastern region = 1, others = 0.
Item of perception and understanding, and Thai national sentiment
UND	Awareness and understanding about local products, breeding and processing of fresh cassava, and the standard and safety of chips are measured by yes–no questions of six issues. Correct answers are marked 1 point, while wrong answers are marked 0 point. This variable is presented in percentage form, where a higher percentage implies better awareness and understanding. Realizing the advantages of using local ingredients and cassava should encourage more chances to show a willingness to pay more.
LNNPRI	The national pride is measured by 14 questions about being Thai, environmental positions, food safety and standard, agricultural and processed products, and innovation and agricultural technology. The survey respondents are asked to rate five-level of opinions based on the Likert scale, from strongly agree = 4 to cannot decide or not sure = 0. The higher mean score of those 14 questions illustrates more sentimental value in the Thai nation. The higher level of this sentiment would hypothetically show support for locally sourced products at least in the case of cassava chips.
Item of consumers’ behavior in buying snacks
FREQ	The frequency of buying snacks for own self and household consumption is a dichotomous variable. It is marked as 1 when purchased daily or at least twice a week; otherwise it is given as 0.
ISHOP	The place where respondents buy snacks is a dichotomous variable. If respondents always buy from local grocery stores, they may be highly likely to pay more for products made from local ingredients and, thus, Thai cassava chips. Respondents’ buying often from local grocery stores is marked as 1; otherwise = 0
Item of shopping behavior is influenced by product composition and attributes, awareness of nutritional value, standards and safety, and external factors and components.
LNBSIGN	Six questions related to opinions about the importance of the factors affecting the type of snacks respondents choose to buy are asked. Additionally, surveyed individuals are to indicate a 5-point Likert scale rating, cannot decide or not sure = 0, strongly disagree = 1, disagree = 2, agree = 3, and strongly agree = 4. This variable is the natural logarithmic of an average value of these six questions.
LNBBUY	Similar concept to BSIGN, surveyed individuals express their opinions on eight questions based on a 5-level Likert scale. Those questions reveal how product attributes affect purchasing and/or consuming behavior toward snacks. To analyze the explanatory power, the rating scores are recalculated into the natural logarithmic average.
LNBNUR	It is the natural logarithmic mean scores of opinions about the importance of nutrition. To assess those of opinions, surveyed individuals are asked to identify five rating scores on the Likert scale of six related questions. The higher the score they rate, the more importance they attach to these issues.
LNBSAF	It represents the natural logarithmic mean of seven questions that correspond to the opinion level on food safety. Respondents are asked to give a score based on the 5-point Likert scale. This variable captures how confident Thai consumers feel about Thai raw material products. The relationship between willingness to pay and this variable can be positive or negative, which could lead to the belief about locally produced goods as being normal or inferior products, respectively.
LNBFAC	This variable is the natural logarithmic value of the grand mean of four questions about how surveyed individuals rate the importance of taste, packaging appearance, manufacturer/brand, and advertisement as influencing their buying decisions.
LNBCONT	This is the grand mean of seven questions related to opinions about the importance of product components, given based on a 5-level Likert scale from least to most agreement. It is in the natural logarithmic form.

It is worth mentioning that the questionnaire is designed to sort out samples to avoid behavioral bias. The screening questions are thus created to categorize surveyed individuals who expressed their displeasure to pay more into two groups, protest and true preference, by asking why they were reluctant to pay more [41,42]. If their reason is that the government should be responsible for the incremental price of locally sourced products, they protest by showing their unwillingness to pay. Hence, they will be excluded from the estimated population since they actually realize the advantage of using local raw materials, but conceal it.

3. Results

This section presents the results of this study. The first part describes the preliminary statistics of respondents’ characteristics. This is to provide background information of the samples. The next part shows respondents’ preference for locally sourced cassava chips through their willingness to pay for the products. The last part presents the findings on factors determining consumers’ preference for cassava chips.

3.1. Descriptive Statistics of Sample

The personal information of 1000 surveyed individuals, shown in

Table 3. Demographic, social and economic characteristics of surveyed individuals.

Characteristics	Number (%)	Characteristics	Number (%)
Gender:		Occupation:
Male	532 (53.20)	Student	88 (8.8)
Female	421 (42.10)	Private company employee	225 (22.5)
Alternative	47 (4.70)	Civil servants/state enterprises/other government agencies	473 (47.3)
Age:		Business owner	107 (10.7)
18–20 years old (Generation Z)	23 (2.3)	Freelance and others	107 (10.7)
21–37 years old (Generation Y)	492 (49.2)	National Identification:
38–53 years old (Generation X)	407 (40.7)	Thai-Thai	876 (87.6)
54–72 years old (Baby Boomer)	78 (7.8)	Thai-Chinese	99 (9.9)
Education:		Thai-Muslim	23 (2.3)
High school certificate or below	162 (16.2)	Others	2 (0.2)
Bachelor diploma	698 (69.8)	Domicile:
Master’s degree or higher	140 (14.0)	North	199 (19.9)
Average of individual income per month (THB):		Northeast	213 (21.3)
20,000 or less	526 (52.6)	Central	265 (26.5)
20,001 or greater	474 (47.4)	South	236 (23.6)
		Bangkok and Metropolitan	87 (8.7)

Note: The number in the bracket is the ratio of respondents with those particular characteristics to the total number of samples of 1000 persons.

, reports that more than half of them are male and about 90% of total respondents are considered to be in generations Y and X. Those two generations are a group born from the early 1980s to the late 1990s, accounting for a large proportion of the population with easy access to information, making communication with this age group important. The understanding of their behaviors and ways of thinking has caught the attention of not only the academic circle, but also the government and businesses, since they would draw useful information for strategizing communication channels. Almost half of them work for government-related agencies, while around 30% are company employees and business owners. A rate of 69.8% of total respondents have already graduated or still study for a bachelor/vocational certificate. The proportion of respondents with an average monthly income higher and lower than THB 20,000 is quite similar. Most respondents identify themselves as Thai-Thai, not belonging to any minority group. Most respondents’ domicile is in the Central region, accounting for 26.5%, while only 8.7% are from the Bangkok and Metropolitan area.

3.2. Preference for Locally Sourced Products: Cassava Chips

There are two separate sets of questions uses to retrieve respondents’ preferences for locally sourced cassava chips. Firstly, the question starts with no initial price to ask whether respondents are ready to pay an extra for chips made from local cassava. The results in

Table 4. Preference on willingness to pay an extra price for chips made from Thai cassava.

Preference	Number	Percentage
No minimum starting marginal price (THB 29 = general price of 75 g pack of potato chips)
Willing to pay	908	90.8
Not willing to pay	92	9.2
Reasons for not willing to pay an extra price for crispy chips made from Thai cassava
Do not believe it tastes like potato chips	27	29.3
Do not believe that cassava can be processed into chips	15	16.3
Do not respond to the incremental price of cassava chips. Government should take responsibility	50	54.3

unsurprisingly indicate that 908 respondents (90.8%) are willing to pay more for the product with local content in the case of chips made from Thai cassava on top of a general price of THB 29 for a 75 g packet of potato chips. On the other hand, only 92 individuals, or 9.2% of all respondents, are unwilling to pay more. Moreover, considering the reasons for unwillingness to pay more among those 92 surveyed individuals (Table 4), 50 of them feel reluctant to pay extra since the government should be solely responsible for the incremental price of cassava chips to promote products made from local raw materials for the benefit of the country’s economy. If the individuals argue that the government should be responsible for the excess price, in this case, they would be regarded as having hidden true preference. This is because they really are aware of and understand the extensive benefits, but do not want to pay more to get something better. Therefore, those samples are said to be a group of protest preferences that has to be excluded from the sample for further analysis to avoid biases. It must also be noted that another 42 individuals have shown their true preference of not willing to pay an extra price. Thus, a total of 950 samples is taken into account.

Secondly, the concept of a single-bounded closed questionnaire with an initial price is employed. According to

Table 5. Preference for chips made from Thai cassava classified by starting prices.

Willing to Pay an Extra Price	Number	Percentage
Additional willingness to pay with 4 starting prices:
Willing	773	77.3
Not willing	177	17.7
Protest preference	50	5.0
Classification of four cases of 950 samples showing true preference:
(1) Starting price 2 THB: willing	204	84.3
not willing	38	15.7
(2) Starting price 5 THB: willing	197	84.5
not willing	36	15.5
(3) Starting price 8 THB: willing	198	80.2
not willing	49	19.8
(4) Starting price 10 THB: willing	181	77.0
not willing	54	23.0

, the elicitation of price setting effect is thus designed by placing the four different scenarios in which the incremental prices or starting prices are offered, namely THB 2, 5, 8, and 10 from THB 29, the general price of 75 g pack of potato chips. Based on 950 samples, 773 respondents present their willingness to pay more, while 177 respondents are unwilling. There are 42 more respondents who switch from willing to not willing compared to the case of no initial price. Additionally, this corresponds to when the starting price is set to be higher sequentially. The number of samples willing to pay for the increase is even lower, especially when set at the first increment of THB 10. The finding reveals the price influencing purchasing decisions in accordance with the law of demand when categorizing those willing to pay based on the four starting prices.

3.3. Factors Influencing Consumers’ Behavior

Before working on the empirical intuition, the selection of proper specifications between logit and probit models must be performed. From the comparison of log- likelihood and pseudo-R² values between the two specifications, they show that logit and probit models had little difference. However, the probit models report somewhat better specifications. Moreover, their estimated results illustrate consistent outcomes. Therefore, to address the influencing factors, probit analysis is reasonably used to estimate the coefficients, and then, the marginal effect is calculated with respect to variables.

According to the three estimations shown in

Table 6. Estimation results of factors influencing consumers’ preference for cassava chips.

Variable	Model 1		Model 2		Model 3
	COFF	MFX	COFF	MFX	COFF	MFX
Constant	1.223 (0.33) a	-	1.270 (0.32) a	-	1.332 (0.32) a	-
PRS	−0.145 (0.07) b	−0.037 (0.02) b	−0.155 (0.08) b	−0.039 (0.02) b	−0.154 (0.07) b	−0.039 (0.02) b
GEN	0.153 (0.11)	0.038 (0.03)	0.92 (0.11)	0.023 (0.03)	0.159 (0.11)	0.040 (0.03)
LNUM	−0.068 (0.12)	−0.017 (0.03)	−0.035 (0.12)	−0.009 (0.03)	−0.058 (0.12)	−0.015 (0.03)
AGEZY	0.247 (0.11) b	0.063 (0.03) b	0.179 (0.11) c	0.045 (0.03) c	0.274 (0.11) a	0.069 (0.03) a
EDUC	−0.473 (0.16) a	−0.102 (0.03) a	−0.503 (0.16) a	−0.106 (0.03) a	−0.457 (0.16) a	−0.098 (0.03) a
OCCG	0.073 (0.11)	0.018 (0.03)	-	-	-	-
OCCPB	-	-	0.253 (0.12) b	0.061 (0.03) b	-	-
OCCS	-	-	-	-	−0.407 (0.18) b	−0.119 (0.06) b
INCL	0.046 (0.10)	0.012 (0.03)	−0.039 (0.11)	−0.010 (0.03)	−0.013 (0.11)	−0.003 (0.03)
RACET	0.229 (0.15)	0.063 (0.04)	0.244 (0.15)	0.067 (0.04)	0.227 (0.15)	0.062 (0.04)
HOMNE	−0.253 (0.12) b	−0.069 (0.03) b	−0.268 (0.12) b	−0.072 (0.03) b	−0.248 (0.12) b	−0.067 (0.03) b
FREQ	0.399 (0.10) a	0.101 (0.03) a	0.393 (0.10) a	0.098 (0.03) a	0.402 (0.10) a	0.101 (0.03) a
LNBSIGN	−0.097 (0.11)	−0.025 (0.03)	−0.102 (0.10)	−0.026 (0.03)	−0.098 (0.11)	−0.025 (0.03)
LNBBUY	−0.178 (0.12)	−0.045 (0.03)	−0.191 (0.12) c	−0.048 (0.03) c	−0.209 (0.12) c	−0.053 (0.03) c
LNBNUR	−0.013 (0.11)	−0.003 (0.03)	−0.018 (0.12)	−0.005 (0.03)	−0.034 (0.12)	−0.009 (0.03)
LNBSAF	−0.182 (0.11)c	−0.046 (0.03)c	−0.181 (0.11) c	−0.046 (0.03) c	−0.170 (0.11)	−0.043 (0.03)
LNBFAC	0.066 (0.11)	0.017 (0.03)	0.064 (0.11)	0.016 (0.03)	0.073 (0.11)	0.019 (0.03)
LNBCONT	0.164 (0.11)	0.042 (0.03)	0.176 (0.11) c	0.044 (0.03) c	0.172 (0.11)	0.043 (0.03)
UND	0.411 (0.17) a	0.104 (0.04) a	0.449 (0.17) a	0.113 (0.04) a	0.403 (0.17) b	0.101 (0.04) b
LNNPRI	−0.051 (0.11)	−0.013 (0.03)	−0.045 (0.10)	−0.011 (0.03)	−0.058 (0.11)	−0.015 (0.03)
ISHOP	−0.167 (0.12)	−0.044 (0.03)	−0.162 (0.12)	−0.043 (03)	−0.153 (0.12)	−0.040 (0.03)
# of obs.	950		950		950
R2	0.061		0.066		0.066
Log-likelihood	−428.770		−426.728		−426.588

Note: # of obs. and R² are denoted the numbers of observation and pseudo R², respectively. ^a, ^b, ^c represent the significant level of 0.01, 0.05, and 0.10. Number in the bracket is regression standard error. COFF stands for coefficient while MFX is the marginal effect.

, there are 19 relevant variables included which cover socio-economic characteristics, attitude, buying behavior, price setting, perception, and understanding. The estimations also intend to clarify the explanatory power of different occupations on the preference for products with local content. Occupations are thus distinguished into three groups with different regressing models. The results of the three estimated models show the strong explanatory power of five independent variables, namely, PRS, EDUC, HOMNE, FREQ, and UND. They are statistically significant in all three models.

The negative statistical significance at the 5% conventional level of PRS indicates that the higher the initial bid price offered to respondents, the lower they are willing to pay more for chips made from Thai fresh cassava. This, in other words, reveals the concept of market price setting that highlights the basic price mechanism under the law of demand. EDUC is statistically significant at 1% in explaining people’s intention to pay a premium price for chips made from fresh local cassava. The negative relationship shows that people with higher education tend to have less willingness to pay extra for the products in this case.

The statistical significance at the 5% conventional level of HOMNE shows its explanatory power for consumers’ willingness to pay more for chips made from fresh Thai cassava. Their negative relationship means respondents whose hometowns are from the Northeastern area are less willing to pay higher for chips made from local raw materials. FREQ is statistically significant at 1% in explaining consumers’ willingness to pay for cassava chips. Their positive relationship indicates that the more frequently consumers buy snacks, the more likely they are willing to pay more for chips made from fresh Thai cassava.

The statistical significance of UND indicates that consumers with knowledge and accurate understanding of local products, the processing of fresh cassava, and standards and safety for snacks are more likely to pay a higher price for chips made from Thai fresh cassava. As for occupation, the statistical significance of OCCS with a negative sign suggests that students seem to be less willing to pay a higher price for chips made from fresh Thai cassava. This may be because, generally, students have no income of their own, and thus cannot afford to pay an increased price.

Another interesting finding is the statistical significance of variables representing consumers’ generation, namely AGEZY. It can be seen that consumers’ generation can affect the purchase and consumption of chips made from fresh Thai cassava. The positive sign carried by the coefficient of AGEZY indicates that Generation Z and Y consumers tend to pay more for these cassava chips. Differences in consumers’ generations can clearly influence buying behaviors. Another interesting finding drawn from those three estimations is about consumers’ behaviors represented by two variables, LNBSAF and LNBBUY, which reflect the importance of the standard, food safety, food attributes, and food quality issues. Even though those two variables show only mild explanatory power with a 0.1 significant level, their negative signs inform insightful information. Thai people, in this case, appear to view products made from domestic raw materials or cassava chips as seemingly inferior products.

4. Discussion

As shown in this study, consumers’ behavior toward chips made from local cassava can be determined by several factors. The price offered to respondents (PRS) can affect their buying decision in a diverse direction. This is consistent with the law of demand, stating the adverse relationship between product price and quantity demanded. This is also consistent with the study on consumers’ behavior by Kucher et al. and Ali and Ali [43,44], which find the strong influence of price setting on people’s purchase decisions. The next factor that appeared to have a substantial impact on persons’ consumption behavior toward cassava chips is their educational level (EDUC). The opposing relationship between this factor and the opportunity of spending more on chips made from local cassava, found here, is different from the hypothesis set earlier. This may be because educated individuals, with their ability to access information from various sources, are more confident in the quality of imported raw materials. At the same time, they may consider local inputs as inferior to foreign counterparts. Taking this into account, they may not be eager to pay more for chips made from fresh local cassava. The opposing relationship between education level and people’s willingness to pay is supported by Jekanowski etal. and Giraud et al. [45,46], which found a negative linkage between education and willingness to pay for local agricultural products.

Another essential factor shown to have the ability to explain consumers’ readiness to pay extra for chips made from local raw materials is the individuals’ hometowns of the Northeastern region (HOMNE). The negative relationship, as presented earlier, reveals that although fresh cassava is most planted in this region, persons from the Northeastern part of Thailand are less likely to pay an extra amount for chips made from their own produce. It is possible that this area has always been hard hit by poverty; thus, its people are inclined to spend more on basic food than snacks.

Respondents’ frequency of buying snacks (FREQ) can also influence their behavior toward purchasing locally made cassava chips. The frequency of buying snacks can reflect buyers’ satisfaction with the products. More frequent buying can indicate a higher level of consumers’ satisfaction with snacks. Therefore, they have a tendency to pay a premium price in this case. It can be because they may not only enjoy the products, they like but also benefit from the freshness of local inputs as well as support Thai cassava farmers. Consumers’ frequent buying behavior implies their habit, which could indicate the purchase intention. This is in line with the work of Maksan and Jelic [47], who identify the intention to buy domestic potato chips as the reason behind consumers’ actual purchasing.

The strong explanatory power of respondents’ knowledge and correct understanding of local products and related issues (UND) shows its apparent effect on the likeliness to pay more for cassava chips. This can be explained by correctly understanding related information and situations. This allows consumers to recognize the advantages of consuming chips made from fresh Thai cassava possibly received by themselves and society. This is consistent with the work of Jekanowski et al. [45], who found that the perception of product quality increases consumers’ purchases.

The significance of persons’ generation (AGEZY) as the determinant of their chance of spending more money in buying chips made from local cassava, as explained above, can be beneficial for entrepreneurs. The results show that the most robust marketing opportunities lie in the younger generation. This result is in line with the work of Chancharoenchai and Saraithong [39], which finds that consumers under 20 years are more willing to pay a premium price for chips containing no carcinogens than those over 20 years. On the other hand, it contradicts the work of Winterstein and Habisch [48], which find that younger consumers are more sensitive to price and less willing to pay. This difference may be because Winterstein and Habisch [48] study a case of organic products whose features differ from snacks that are more popular among younger consumers.

On consumers’ attitudes and behavior toward such aspects as nutrition, food safety, and food quality, Zepeda and Li [49] provide comparable results. The statistical insignificance of LNBNUR is consistent with their results. In contrast, the mild explanatory power of LNBBUY is somewhat different from their finding, which shows that consumers’ attitudes toward food can significantly increase the chance of shoppers’ buying local food. Schneider and Francis, Dukeshire et al., and Memery et al. [36,50,51] show the statistical significance of individuals’ attitudes toward an environmental issue in encouraging them to buy products with local content. Unlike these works, this study’s result, the insignificance of LNNPRI, somewhat highlights an absence influence for an environmental issue on consumers’ preference for cassava chips. The diversion of these results may be because the variable LNNPRI, described in Table 2, is a mean score of a five-level Likert scale of many issues, the environment included. Thus, to assess the precise impact of an environmental issue on the consumption behavior of local cassava chips, further study should be performed with variables solely representing this matter in the estimated model.

From the findings stated earlier, through different aspects, strategies can be drawn to enable chips made from fresh local cassava to become more marketable. The strong influence of consumers’ generation on their consumption behavior suggests that younger customers could provide solid marketing opportunities for entrepreneurs. Thus, this group of buyers should be the focus of marketing and sale strategies. To effectively address this client group, the promotion should be carried out more through social media than traditional channels. Additionally, regular snack customers appear to be a high-potential client for chips made from local cassava as well. Thus, it could be more efficient for involved businesses to directly campaign with this group first.

The use of local cassava instead of imported potato in the production of chips naturally leads to a change in the nutritional qualities of products. Apart from the difference in the level of Acrylamide formation between cassava and potato chips, as mentioned earlier, both tubers are endowed with different nutrients. Although, in this study, the variable representing the nutritional issue is not statistically significant, providing and communicating this information to the public should allow consumers at large to have a precise understanding of their consumption choices. In turn, the variable for understanding about cassava chips-related issues is statistically significant in influencing consumers’ preference for cassava chips.

Regarding the study’s findings, having knowledge and understanding of local products, food safety, nutritional issues, as well as other related issues is even more important. This is because, thus far, in Thailand, there is no law regulating the maximum levels of acrylamide allowed in food products. In the EU and US, there are only official recommendations for the producers to mitigate the presence of acrylamide in foods [13,52]. The lack of mandatory regulation regarding acrylamide in foods forces consumers to be vigilant. For this reason, the availability of necessary information and their correct understanding of the subject is critical.

Hence, the government should facilitate and make sure that communication with the public on these issues can take place without delay. More importantly, fake news and inaccurate information regarding these matters should be swiftly dealt with before creating any misunderstanding among the public. As it seems like chips made from local cassava are not well received by educated individuals, their confidence needs to be elevated in fresh local cassava and Thai product standards in general through the provision of reliable and scientific, proof-based evidence. Making the quality of local raw materials more credible for educated individuals should entice these people to be more trustful of locally sourced cassava chips.

The results of this study are specific to the case of cassava chip production in Thailand. They can obviously provide direct insights for stakeholders in the country. Nonetheless, the findings’ implications can also be beneficial to other cases or other countries with comparable situations. Cassava is a staple crop in many countries sharing similar climates and geographical characteristics. As for the countries in the Southeast Asia region, as stated before, Indonesia and Malaysia are at the front row in promoting cassava chips to drive the local economy. Furthermore, African countries, to a large extent, produce a vast quantity of fresh cassava; thus, cassava-related products, including cassava chips, also play quite a major role in their economic development [53,54,55].

This study’s implications can be of value to these countries. For instance, the generation effect is likely to affect consumers’ buying decisions in the same manner in most countries. However, some specific features of each country/region need to be considered. For instance, Indonesia and Malaysia are predominantly Muslim countries; as a result, cassava chips are to be manufactured according to the Halal requirements. As for African countries, due to their high level of income inequality and the uneven access to various infrastructures, including education, such factors as income, education, may yield different results.

This study’s limitation is that it largely relies on data from respondents’ self-measurement of various issues, as explained above. Respondents may take part in the survey and provide information without any actual experience or knowledge about cassava chips. They may express their preference regarding these products based on prejudice or discrimination against local inputs. As a result, their responses may need to be more accurate. To improve this shortcoming, it is recommended that future research may undertake an experimental technique to measure preference. This should allow survey participants to actually experience the taste of or pay for cassava chips, for instance, thus revealing their true preference for the products.

5. Conclusions

As a consequence of the high opportunity and potential of cassava in the chip market in Thailand, cassava chips have thus been the subject of interest, especially of how to realize the market boundary. This study, then, attempts to fill a gap in the cassava value chain in Thailand. On the one hand, the country can produce plenty of fresh cassava. At the same time, it spends a large amount of foreign currency on importing potatoes as an input in the production of chips, which can also be substituted by local cassava. The missing link is the marketing difficulties of chips made from local inputs. The insightful information about consumers’ behavior is therefore necessary and helpful to not only the government, but also to entrepreneurs in formulating proper policies. Besides, successful commercialization can bring forward and backward benefits and indirect impacts such as local business, farmers’ welfare, and saving foreign currency. As consumers accept and buy more chips made from local cassava, farmers can enjoy more opportunities to sell their fresh produce and earn a higher income, consequently, attain a better standard of living. Using local cassava in the production of chips can add more value to the country’s cassava value chain, improving the quality of life of people who work along the chain, mostly living in the underdeveloped area of the Northeastern region. Moreover, looking at the broader picture, locally sourced cassava chips can lead to more sustainable development as they can save the country’s foreign exchange. Additionally, with the decreased flow of imports, international transportation declines, thus enhancing environmental conditions.

Regarding the probit analysis and marginal effect, the empirical evidence reveals several factors that could affect consumers’ behaviors, including their willingness to pay a higher price for chips made from local cassava. These factors include, for example, educational level, frequency of buying snacks, knowledge and correct understanding about local products, and a persons’ generation. More precisely, generation, which is classified as being in generation X and Y, has a significant effect on the preference to pay an extra price for local cassava to produce chips. Unsurprisingly, more understanding of the economic and welfare impact of using local materials would increase the chance of people paying an additional price. Students and people with higher education turn out to be less willing to pay more for chips produced from local cassava. On the contrary, more concerns about quality, safety, and standards show less probability of paying a premium price. The result shows the rather disappointing outcome of national pride for chip sales, which discloses a somewhat insignificant influence. Knowing these factors should allow entrepreneurs to adjust their operations and attract more customers, government agencies to prepare and provide necessary assistance and information to the public, and consumers to understand and be willing to buy chips made from local cassava.