Purchase Intention of Certified Coffee: Evidence from Thailand

Abstract

This study examines social identity and self-identity as the antecedents of the theory of planned behavior (TPB) model in predicting purchase intention of certified coffee, whereas perceived trustworthiness (PT) is evaluated whether it directly affects intention and/or indirectly through attitude. In addition, ethics and luxury are investigated as the salient beliefs affecting attitude formation in this regard. A face-to-face survey was conducted with 727 coffee consumers in Thailand. Confirmatory factor analysis is applied to assess the adequacy of the model, followed by structural equation modeling to evaluate the hypotheses proposed for the relationships between constructs in an extended TPB model. The results confirm that self-identity is the most influential antecedent on attitude when compared to social identity, and attitude, in turn, is the strongest determinant in predicting purchase intention. PT has a direct positive effect on purchase intention, meanwhile, ethical, luxury beliefs, and PT are confirmed to portrait the attitude formation. As such the marketing campaigns can address manipulating consumers' beliefs on both ethical and luxury aspects as well as PT, along with consumers' social identity and self-identity to fortify a positive attitude toward certified coffee. Then the actual purchase behavior can be foreseen based on empirical evidence.

1. Introduction

According to the World Bank, agriculture currently accounts for seventy percent of all freshwater consumption globally (The World Bank, 2020). In addition, recently, many studies in the field of production efficiency have addressed the impacts of agricultural crop production on the environment, e.g., environmental efficiency (Vu et al., 2020). Coffee production is not an exception as well, the world’s coffee production is now more damaging than ever to the environment than before (Jha et al., 2014). Coffee originally grew in tropical and subtropical regions with various verdure contributing to abundant diversity and shade that provided habitats for many indigenous species. In turn, these trees and animals aided to avoid topsoil erosion and fertilizer application. Since the 1970s, the demand for chemical fertilizer for growing coffee has increased significantly due to the emergence of coffee plantations by continuous deforestation (Moore, 2020), accompanied by the demand for “sun cultivation” for coffee, which destroyed the shady trees and animal habitats (Clay, 2013).

Apart from environmental damages, the coffee industry has experienced low prices of green coffee beans in the past twenty years (International Coffee Organization, 2020). On average, the typical third-world coffee farmers earn a meager ten percent of the final retail price; seventy percent of whom are smallholders suffered due to the low price and the undercutting by large retailers and has been left in the lurch to the volatile coffee market (Energy Makeovers, 2016; The Guardian, 2011). This has led to the concept of Fairtrade growing with a vision of more ethical coffee drinking by encouraging environmentally friendly production practices and ensuring fair returns to coffee farmers (Lentijo & Hostetler, 2011). Other coffee certification programs under the Voluntary Sustainability Standards (VSS) such as Organic, Utz Certified, and Rainforest Alliance have also been introduced to stakeholders in the coffee sector to promote the awareness of sustainable consumption and production (International Coffee Organization, 2020; International Trade Centre, 2011; Lentijo & Hostetler, 2011), i.e., to comply with the 12^th SDG proposed by the United Nations (United Nations, 2019). For customers who attempt to lead a greener lifestyle, they can seek eco-friendly products including certified coffee for practicing green consumerism (Han & Kim, 2010).

In Thailand, with the emerging of middle-class consumers, the demand for certified products has been increasing. More and more upscale supermarkets, specialized shops, as well as luxury coffee café are providing certified food and coffee under international sustainable labels to serve the need of consumers in urban areas (Zhang & Prasongsukarn, 2017). This study, therefore, attempts to investigate the insights affecting consumers’ behavior towards certified coffee by examining the case of Thailand. The contribution of this study can be useful for various stakeholders in the coffee industry (e.g., policymaker, roaster, marketer, coffee farmers, etc.) as a reference for strategies and policies in marketing to promote sustainable coffee production and consumption driven by consumers.

2. Literature Reviews and Hypotheses Development

2.1. Theory of Planned Behavior

Based upon the theory of planned behavior (TPB), human behavior is best predicted by intention, and intention, in turn, is a set of attitudes towards behavior, subjective norm, and perceived behavioral control (PBC) (Ajzen, 1991). TPB started as the Theory of Reasoned Action in 1980 to predict an individual’s intention to engage in a behavior at a specific time and place. Perceived behavioral control (PBC) refers to a person’s perception of the ease or difficulty of performing the behavior of interest. PBC varies across situations and actions, which results in a person having varying perceptions of behavioral control depending on the situation. This construct of the theory was added later and created the shift from the TRA to the TPB (Ajzen, 1991). The TPB had been successfully applied to a wide range of studies in agricultural production, social, ecological, environmental, and health-related issues (Chen & Tung, 2014; Senger et al., 2017). However, the TPB had received criticism for its limitation that focused on only the influence of three predictors (Ajzen, 1991).

The extended TPB models were then proposed and adapted into various fields to increase the explanatory power of human behavior (Paul et al., 2016; Wu et al., 2020). Nevertheless, a vast of literature investigated the determinants of green consumption and pro-environmental action by using the TPB frameworks (Maichum et al., 2016; Paul et al., 2016; Wu & Chen, 2014), yet to the authors’ knowledge, there is no research for studying the determinants of certified coffee purchase in the Thailand context. Therefore, this study challenges to incorporate identity theory into a TPB model by examining the purchase intention of Thai consumers.

2.2. Hypotheses Development

2.2.1. Belief

According to TPB, salient behavioral beliefs are assumed to form the attitude toward behavior (Ajzen, 1991; Fishbein & Ajzen, 1975). Previous literature regarding environmentally friendly products revealed that salient beliefs, i.e., ethics, environment, safety, and health were confirmed as successful determinants to predict attitude toward behavior, purchase intention, and further behavior (Lee & Yun, 2015; Maichum et al., 2016). In addition, a new perspective on sustainable purchase, which is seen as an indication of a luxury status, has emerged in Western society recently (Bain, 2017). Along with it, recent literature from China also supports that a salient luxury belief is a potentially determined attitude towards buying organic products (Bai et al., 2019). Accordingly, the following hypotheses are proposed:

H1: Ethical belief influences consumer attitude toward certified coffee.

H2: Luxury belief influences consumer attitude toward certified coffee.

2.2.2. Perceived Trustworthiness

Perceived trustworthiness (PT) refers to the amount of confidence that an individual has over another person (Strauss & Stüve, 2016). Scholars claimed that trust was significantly related to attitude toward purchase behavior (Nguyen et al., 2019b), indirectly related to attitude toward perceived risk (Nguyen, 2020), and directly influenced the purchase intention (Bai et al., 2019). Strauss and Stüve (2016) incorporated trust as the antecedent of the TPB model in predicting the purchase intention of organic hair care products, and the result revealed that trust had a positive relationship with TPB variables such as PBC. Accordingly, the following hypotheses are proposed:

H3: Perceived trustworthiness has an influence on attitude toward certified coffee.

H4: Perceived trustworthiness has an influence on consumer purchase intention.

2.2.3. Social Identity

Social identity is a person’s sense of who they are based on their group membership(s) (White et al., 2018). It’s associated with self-categorization such as work affiliation, nationality, race, gender, as well as salient identity characteristics (Salem & Salem, 2018; Stets & Burke, 2003). Individuals would likely use the norms and roles (i.e., goals, attitudes, and behaviors) sanctioned by group members; as a result, social identity could serve to monitor individuals’ status as group members and strengthen their self-concept (Tajfel & Turner, 1986). In Malaysia, social identity significantly predicted the purchase intention of luxury fashion goods through attitude, subjective norm, and perceived behavioral control (Salem & Salem, 2018). In addition, social factors which consisted of roles, norms, and self-concept were found to be the significant predictor of the bioethanol fuel purchase (Hoang, 2017). Likewise, social support and self-efficacy also predicted social entrepreneurial intention (Akhter et al., 2020). Accordingly, the following hypotheses are proposed:

H5: Social identity has an influence on attitude toward certified coffee.

H6: Social identity has an influence on the subjective norm.

H7: Social identity has an influence on perceived behavioral control.

2.2.4. Self-identity

Theoretically, the relationship between self-identity (SFI) and behavioral intention is determined based on identity theory, which considers the self as a social construct. Identity theory provides a clear rationale for incorporating self-identity as a predictor of intention, stating that in TPB, intention can be regarded as the most intimate determinant of behavior. Besides, self-identity can be formed through social interaction (Sparks, 2000). Some studies found that self identity had a significant effect on attitude and PBC but not on the subjective norm (Salem & Salem, 2018). However, it improved the predicting power of the TPB model. In the context of self-congruity theory, individuals could build their self-identity by choosing products that revealed their self image (Salem & Salem, 2018; Sirgy, 1986). In this sense, respondents who took into account the role of sustainable consumers as an important component of their self-identities, were more motivated to engage in the certified coffee purchase behavior than those who did not. Accordingly, the following hypotheses are proposed:

H8: Self-identity has an influence on attitude toward certified coffee.

H9: Self-identity has an influence on the subjective norm.

H10: Self-identity has an influence on perceived behavioral control.

2.2.5. Attitude

In the TPB model, attitude (ATT) is described as ‘‘the degree to which a person has a favorable or unfavorable evaluation or appraisal of the behavior in question’’ (Ajzen, 1991, p.188). Ajzen views attitude as a “summary evaluation of a psychological object captured in such attribute dimensions as good–bad, harmful–beneficial, pleasant–unpleasant, and likable–dislikable” (Ajzen, 2001, p.28). Previous studies applied the TPB model to explain the prevailing gap between attitude and behavior in the green or organic product segment (Armitage & Conner, 2001). Other research evidence also confirmed that attitude was the primary predictor of behavioral intention (Chen & Tung, 2014; O’Connor et al., 2017; Paul et al., 2016) and the purchase behavior. Thus, the following hypothesis is proposed:

H11: Attitude towards purchasing certified coffee influences consumer purchase intention.

2.2.6. Subjective Norm

Subjective norm (SN) refers to the social pressure to perform or not perform a particular behavior (Azjen & Fishbein, 1980). SN is believed to be a function of normative belief (NB) and motivation to comply (MC). NB is the probability of whether significant referents would approve or disapprove the behavior (e.g., family, relatives, friends, neighbors, or co-workers), and MC is related to individual’s desire to respond to his/her predominant referrals’ opinions with respect to behavior (Azjen & Fishbein, 1980). The strength of each NB is weighted by the corresponding MC, and the goods are summed to impose the subjective norm (Ajzen, 1991, 2001; Azjen & Fishbein, 1980). In previous literature, the subjective norm was proved as a significant predictor of behavioral intention (Arvola et al., 2008; Maichum et al., 2016), but weaker than attitude (Maichum et al., 2016). Therefore, it is postulated that:

H12: Subjective norm influences consumer purchase intention.

2.2.7. Perceived Behavioral Control

Perceived behavioral control (PBC) is a non-volitional factor, which makes the basic difference between TRA and TPB. PBC is defined as “the perceived ease or difficulty of performing the behavior, ” or in other words, it is the extent to which a person feels capable of performing a behavior (Ajzen, 1991). PBC includes such factors as the availability of time and money or having necessary skills and a person’s self-confidence in his/her ability to perform the action (Cheng et al., 2006). Scholars pointed out that the most crucial control factors influencing dietary intake were self-efficacy and convenience/availability (Olsen, 2004). In the domain of green products and organic products, previous literature revealed that PBC had been linked to the purchase intention (Bui et al., 2021; Chen & Tung, 2014; Moser, 2015; Paul et al., 2016). In accordance, the following hypothesis is proposed:

H13: Perceived behavioral control influences consumer purchase intention.

Based on the aforementioned literature review, the following is the research model (Figure 1).

Figure 1: Proposed Research Framework

3. Methodology

3.1. Sample Size and Data Collection

A sample size of 1, 000 is adopted due to the more accurate and above the minimum sample size of 500 based on the criteria of the Structural Equation Model (SEM) (Hair et al., 2019). The survey was conducted during May-July in 2020. The purposive sampling method was applied to select individuals who regularly drank coffee. A face-to-face interview was employed in the data collection. Two pre-tests of questionnaires were performed before the final survey. The response rate of questionnaires is 72.7%. The product in question of this study is certified roast coffee beans in a 250-gram package.

The descriptive statistics of respondents’ demographics are shown in Table 1. A majority of the respondents are females (68.1%). Most respondents are aged between 18–45 years old (81.6%), having full-time jobs (56.9%), bachelor’s degree (60.8%), single (57.1%), having 1–4 family members (71.9%), and a monthly household income of THB 20, 000– 39, 999 (39.9%).

Table 1: Sample Characteristics

3.2. Measures and Tools of Analysis

A total of nine constructs are applied, illustrating twenty eight items rated on a five-point Likert scale ranged from 1 (strongly disagreed) to 5 (strongly agreed) (Table 2).

Table 2: Descriptive Statistics of the Measurement Items

The mean scores of measurement items are higher than 2.50, ranging from 2.59 to 3.73. SPSS 22 is used for the descriptive statistic to analyze preliminary results and to find out the demographic characteristics of the respondents. With AMOS 22, confirmatory factor analysis (CFA) is measured for the adequacy of measurement to confirm the reliability, convergent and discriminant validity. Finally, SEM is conducted to test the hypothesized relationships between the extension of TPB constructs.

4. Results

4.1. Confirmatory Factor Analysis

The high composite reliability (CR) specifies the internal consistency which should be greater than 0.7. The average variance extracted (AVE) of 0.5 or higher is a good rule of thumb suggesting adequate convergence. As presented in Table 3, the CR and AVE values ranged from 0.85 to 0.95 and 0.66 to 0.91, respectively, beyond the recommended levels of 0.7 and 0.5. Consequently, the convergent validity of the CFA is confirmed by the CR and AVE results (Table 3).

Table 3: Reliability and Validity of the Measurement Model

Note: *p-value < 0.05; **p-value < 0.01; ***p-value < 0.001. Significant at the 0.05 level. a Values were not calculated because loading was set to 1.00 to fix construct variance.

Table 4 provides a discriminant validity of the measurement model. In this regard, the diagonal elements (bolded) are the AVE, whereas the off-diagonal elements are the correlation squared among constructs. To ensure the discriminant validity, the square root of each construct’s AVE should have a greater value than the correlations with other latent constructs (Hair et al., 2019). Most of the results provide better evidence of discriminant validity.

Table 4: Discriminant Validity of the Measurement Model

Note: The diagonal elements are the average variance extracted (AVE). Off-diagonal elements are the correlation squared among constructs.

The chi-square statistic is significant (χ²(310 df) = 1430.283, p < 0.001), which is acceptable in case of sample sizes greater than 250 (N > 250) (Hair et al., 2019). The goodness-of-fit index (GFI) is at an acceptable level (0.874) (Ting et al., 2019). Also, the other indices meet the recommended threshold values such as the comparative fit index (CFI) (0.950), the Tucker-Lewis index (TLI) (0.938), the normalized fit index (NFI) (0.937), and the relative fit index (RFI) (0.923), which are all above the common threshold (0.90) (Maichum et al., 2016; Nguyen et al., 2019a). The values of parsimonious normed fit index (PNFI) (0.768) and parsimonious comparative fit index (PCFI) (0.779) are higher than the recommended level (0.5) (Nguyen et al., 2019a). The root mean squared error of approximation (RMSEA) of 0.071 is less than the critical level (0.80) (Ting et al., 2019). Overall, the fit indicators’ results indicate the model fits well.

4.2. Testing of the Structural Equation Model

χ² is statistically significant (χ² = 1837.106; df = 323; p < 0.001) if the sample sizes are greater than 250 (N > 250) (Hair et al., 2019). Other indexes show that GFI is at an acceptable level (0.843) which is higher than the critical threshold of 0.8 (Ting et al., 2019). CFI, TLI, NFI, and RFI are 0.932, 0.920, 0.919, and 0.905, respectively, above the common threshold of 0.90, indicating a good fit(Maichum et al., 2016; Nguyen et al., 2019a). Likewise, the PNFI of 0.785 and the PCFI of 0.796 are all above the general criteria of 0.5, indicating a good fit as well (Nguyen et al., 2019a). RMSEA of 0.08 is within the acceptable range of fair fit (Ting et al., 2019). The diagnostics of these structures, when taken together, therefore, show a relatively good fit of the proposed theoretical models to the underlying data.

4.3. Hypotheses Testing

Table 5 shows the standardized parameter estimates (β) and directional significance levels for the hypothesized paths. All hypotheses are validated and confirmed as expected. The hypotheses H1, H2, H3 regarding the impact of ethical belief, luxury belief, and perceived trustworthiness (PT) on the purchase intention toward certified coffee are confirmed. The hypothesis H4 regarding the impact of PT on purchase intention is also confirmed. These results are in agreement with the previous research findings (Bai et al., 2019; Lee & Yun, 2015; Nguyen et al., 2019b;). The hypotheses H5, H6, H7, H8, H9, H10 are supported, i.e., social identity and self identity have a positive effect on attitude toward certified coffee, subjective norm, and PBC. These results are in line with the relevant literature ( Nguyen et al., 2019b; Salem & Salem, 2018; Willis et al., 2020; Zhang & Kim, 2013). The findings that attitude, subjective norm, and PBC have a significant positive effect on purchase intention completely confirm the hypotheses H11, H12, H13, and the results are in line with many previous research findings (Chen & Tung, 2014; Nguyen et al., 2019b; Ting et al., 2019).

Table 5: Path Analysis Results and Hypothesis Verification

Note: *p-value < 0.05; **p-value < 0.01; ***p-value < 0.001. Significant at the 0.05 level.

5. Discussion and Conclusion

This study aims to investigate the variables based on the extended TPB model. This empirical study has contributed to the existing literature by providing a preliminary analysis on the factors expanding the TPB framework that are likely to influence consumer purchase intention of certified coffee in the Thailand context.

The determinants influencing TPB, i.e., self-identity and social identity, can serve as the strategic reference for marketing policy to increase sales. Self-identity has the greatest impact on attitude; using the strong identities of consumers (i.e., ethical consumers, socially responsible consumers, etc.) as a key role in coffee advertising, promotes a positive attitude toward certified coffee and ultimately leads to purchase behavior. On the contrary, social identity mostly affects subjective norms; hence, marketers can use the influence of social groups that consumers belong to enhance positive subjective norms, and finally stimulate consumer purchase intention. Perceived trustworthiness (PT) directly influences purchase intention. PT is concerned with the certification system and quality label of certified coffee. Thai consumers trust certified coffee in general, including sustainable processing enterprises and certification authorities, who are involved in the production process. Sustainable labels e.g., Fairtrade, Utz certified, Rainforest alliance, etc. can serve as a hallmark for consumers at the point of purchase. However, attitude has the highest positive effect on consumers’ purchase intention in this study. Salient beliefs, e.g., ethical and luxury beliefs as well as perceived trustworthiness are suggested as a reference for developing consumer attitude toward purchasing certified coffee.

Finally, the government is recommended to design policies to support coffee producers and build the consumer perceptions on the trustworthiness of certified coffee consumption as a pro-environmental behavior. In this regard, the government is suggested to subsidize the cost of production and certification that could motivate conventional coffee farmers to certify their coffee beans by adopting the VSS program in their coffee farms and to enhance the higher score of environmental efficiency in the agricultural sector.