Introduction
The prevalence of dining out has grown into a global phenomenon in modern lifestyles. The food service sector now generates multi-trillion-dollar revenues worldwide, reflecting the ubiquity of eating outside the home (Statista, 2024). For example, the global food service industry was valued at around USD 3.7 trillion in 2024 (Statista, 2024). The United States is one of the largest contributors to this market, with its restaurant industry accounting for nearly USD 0.9 trillion in annual sales (National Restaurant Association, 2024). Correspondingly, a substantial share of Americans dine out regularly; a 2022 survey indicated that many U.S. adults eat at restaurants at least once per week (Morning Consult, 2022). Japan likewise exhibits a robust dining-out culture, evidenced by over 1 billion restaurant dinners being served each year in the nation’s metropolitan regions (Ministry of Agriculture, Forestry and Fisheries Japan, 2023). In Taiwan, the practice of eating out is even more ingrained—nearly 90% of Taiwanese adults engage in dining out, with a considerable portion doing so almost daily (Statista, 2023a). These examples illustrate that eating out is not merely an occasional indulgence but rather an entrenched routine across different societies, raising the importance of the food service industry in daily life and the global economy.
Despite the diverse options provided by the hospitality industry for dining out, numerous studies have indicated that its commercial activities contribute to global warming, excessive energy consumption, and food wastage, among other unsustainable outcomes (Dani, Juyal and Rawal, Reference Dani, Juyal and Rawal2021; Langgat, Reference Langgat2020; Tehrani, Fulton and Schmutz, Reference Tehrani, Fulton and Schmutz2020). Studies have noted that restaurant operations are among the most resource-intensive commercial activities—the sector ranks as one of the heaviest consumers of energy and water, and it produces large volumes of waste by-products, including food waste and single-use disposables (Dani et al., Reference Dani, Juyal and Rawal2021). In societies with prevalent dining-out habits, these impacts are especially pronounced. East Asian countries, for instance, face heightened pressure to address the emissions and waste generated by their extensive food service industries. The case of Taiwan exemplifies the need for intervention: without changes in practice, the convenience of frequent dining out can exacerbate problems like food waste and carbon emissions. Since the hospitality industry is identified as one of the highest energy-consuming retail sectors (Kasim, Reference Kasim2009), transitioning the hospitality industry toward a greener, more sustainable direction is crucial.
The concept of green restaurants could be traced back to Lorenzini’s (Reference Lorenzini1994), who defined it as operating in an environmentally friendly manner. In societies where dining out is prevalent, such as in East Asia, green restaurants also bear a significant responsibility to influence and change people’s unsustainable eating habits. Furthermore, the promotion of green restaurants critically depends on consumer demand. Thus, exploring how to change people’s unsustainable eating habits while simultaneously increasing their demand for green restaurants is a topic worthy of ongoing research.
To further focus, this study will narrow down sustainable eating habits to those that contribute to reducing greenhouse gas (GHG) emissions. Using Taiwan as a case study, this study aims to understand how to enhance people’s willingness to visit green restaurants and adopt low-carbon foods in their diet, addressing the gaps in the academic research.
Literature review
The development of green restaurants in Taiwan
The Green Restaurant Association (GRA) was established in 1990 with the mission of fostering a sustainable restaurant industry in the United States. In 2009, it developed a certification system that evaluates restaurants’ green practices in energy, water resources, waste management, food, chemicals, disposable items, and building aspects to assess their overall environmental sustainability (GRA, n.d.). According to Jo, Choi and Taylor (Reference Jo, Choi and Taylor2020), GRA primarily uses a scoring system called green points to assess the green certification level of a green restaurant, with the highest score termed SustainaBuildTM.
Compared to grassroots initiatives in the United States promoting green restaurants, the development of green restaurants in Taiwan is primarily led or intervened by government agencies. The Ministry of Environment first introduced the eco-friendly restaurant certification system in 2012, which adopted a self-application approach for certification. Restaurants obtaining the Eco-Friendly Restaurant certification need to meet three main criteria: refraining from providing disposable items, prioritizing the use of domestic or local ingredients, and offering portion adjustment services (Ministry of Environment, 2024a). Additionally, the AMOT Traceability Restaurant, established in 2013, is jointly promoted by the Ministry of Agriculture and the Agriculture Multi-Discipline Association of Taiwan (AMOT). Its purpose is to create a trusted food traceability record for consumers (AMOT, 2019). Depending on the proportion of traceable ingredients used by the restaurant, it can be rated from one to three stars (AMOT, 2019). Finally, the Green Dining Guide (GDG) was initially established by a grassroots organization in 2018 and later received financial support and guidance from the Ministry of Agriculture. In 2023, it collaborated with the Sustainable Restaurant Association in the United Kingdom to promote a leaf-rating system. Restaurants are evaluated based on the proportion of organic-friendly ingredients used and the depth and breadth of sustainability standards they adhere to, resulting in ratings ranging from one to three leaves (Green Dining Guide, 2022). Currently, the three major green restaurant labels in Taiwan are shown in Fig. 1.
The label of three prominent green restaurants in Taiwan.
Source: Ministry of Environment (2024a).
According to Taiwan Trend Research (2023), there were approximately 163,643 restaurants across Taiwan in 2022. However, as of 2023, the number of restaurants in Taiwan with the Eco-Friendly Restaurant label, the AMOT Restaurant label, and the GDG label was only 3,760, 18, and 222, respectively (Liu, Reference Liu2022; Ministry of Environment, 2024b). In other words, after a decade of development, green restaurants in Taiwan still remain in their early stages, not yet widely integrated into people’s daily lives, accounting for less than 3% of all restaurants.
Dietary change and GHG mitigation actions
Although green restaurants often focus on broader sustainability aspects, they also bear the responsibility of enabling consumers to reduce GHG emissions through their dietary choices, especially given the thriving eating-out culture in Taiwan: 41% of individuals dine out at least daily, and 43% dine out several times a week (Statista, 2023b). According to Santos et al. (Reference Santos, di Sitizano, Pedersen and Borgomeo2022), the consumer demand value chain can reduce GHG emissions through reduced post-harvest losses, dietary changes, and reduced food waste (by consumers or retailers). Specifically, dietary changes have the potential to significantly mitigate large GHG emissions (more than 3 GtCO2eq yr-1) with high confidence.
According to Poore and Nemecek (Reference Poore and Nemecek2018), there can be significant differences in GHG emissions produced by the production of different foods. Therefore, consumers should be encouraged to consume foods with lower carbon emissions in their production process. Additionally, Scarborough et al. (Reference Scarborough, Appleby, Mizdrak, Briggs, Travis, Bradbury and Key2014) found that dietary GHG emissions in self-selected meat-eaters are approximately twice as high as those in vegans. This implies that to reduce GHG emissions, green restaurants should offer vegetarian options. Finally, several studies have shown that in most cases, purchasing seasonally and locally produced goods instead of imported ones can reduce GHG emissions by avoiding international transportation miles (Avetisyan, Hertel and Sampson, Reference Avetisyan, Hertel and Sampson2014; Striebig, Smitts and Morton, Reference Striebig, Smitts and Morton2019).
Consumer willingness to pay a premium for sustainable dining and the eco-label recognition gap
Recent studies across different markets indicate that many consumers are indeed willing to pay a premium for sustainable dining experiences. In developing Asian markets, for instance, visible green initiatives by restaurants can translate into higher customer spending. An empirical study of green restaurants in Ho Chi Minh City, Vietnam, found that implementing green practices (e.g., recycling programs, energy efficiency measures, and sustainable menu options) significantly enhanced customers’ emotional attachment and satisfaction—ultimately boosting their willingness to pay about 5% more for the restaurant’s offerings (Mai, Nhan and Nguyen, Reference Mai, Nhan and Nguyen2023). Similarly, a large 2023 survey of 1,377 consumers in five East Chinese cities quantified the premium people would spend on a ‘green’ lifestyle. For eco-friendly food products specifically, respondents were willing to pay an extra ¥81.8 per month on average (approximately US$11–12) compared to conventional options (Geng et al., Reference Geng, Yang, Zhang and Yang2023). Industry insights in Taiwan echo these trends: Taiwanese consumers, especially younger generations, are increasingly environmentally conscious and appear willing to spend extra at restaurants that adopt eco-friendly practices. Chen et al. (Reference Chen, Liao, Cheng, Shyr and Huang2023) note that consumers in Taiwan care deeply about sustainable actions, such as responsible ingredient sourcing, eco-friendly packaging, and waste reduction in dining, and are inclined to reward restaurants engaging in such practices with their patronage and wallets. On a broader scale, research on millennials’ dining preferences also supports this notion. Nicolau et al. (Reference Nicolau, Guix, Hernandez-Maskivker and Molenkamp2020) examined millennials’ willingness to pay a premium for green restaurants and found not only that many are open to paying more, but also identified key predictors for a higher willingness-to-pay. Using a Heckit model to analyze both the decision to pay any premium and the decision of how much to pay, they found that a strong pro-environmental attitude, high health consciousness, higher income, a clear preference for green restaurants, and a willingness to spend extra time/effort (e.g., traveling farther to dine sustainably) all significantly increase the likelihood that a consumer will pay a premium for sustainable dining. These studies collectively highlight that there is a substantive segment of consumers ready to support sustainable dining with extra spending, which underscores the market opportunity for green restaurant initiatives.
However, an important caveat has emerged in recent literature. Consumer willingness alone may not translate into action if there is a gap in trust or understanding of environmental labels and certifications. In other words, even as environmental awareness rises, a lack of confidence or familiarity with ‘green’ labels can hinder consumers from acting on their pro-sustainability intentions (Gorton et al., Reference Gorton, Tocco, Yeh and Hartmann2021; Natsir et al., Reference Natsir, Takai, Seo, Seo and Kim2025). Several studies have explored why consumers trust or distrust eco-labels. For example, Wang, Tao and Chu (Reference Wang, Tao and Chu2020) surveyed 844 shoppers in China and found that who issues a given eco-label heavily influences consumer trust. International certifications (such as globally recognized organic or sustainability standards) garnered the highest trust among Chinese consumers, whereas lesser-known or local labels were met with more skepticism. This suggests that credible backing and the reputation of the certifying body are crucial for a label’s effectiveness. Building on this, Yang, Xue and Qiao (Reference Yang, Xue and Qiao2024) investigated factors that improve trust in eco-labels through a survey of 1,072 consumers in Inner Mongolia, China. They identified three key ingredients for strengthening label trust: the consumer’s ability to acquire and understand product information, high institutional trust (i.e., confidence in regulators and certification agencies), and prior knowledge of eco-labels. Essentially, a well-informed consumer who trusts institutions and recognizes a label is far more likely to trust that label and act on it. Conversely, when consumers lack knowledge about a label or doubt the institutions behind it, the label’s impact on their behavior diminishes.
Research in Japan further illustrates how multiple labels and varying credibility can complicate consumer trust. Natsir et al. (Reference Natsir, Takai, Seo, Seo and Kim2025) examined Japanese consumers’ responses to products carrying two eco-labels: the national Organic JAS label (a government-backed organic standard in Japan) and the international RSPO label (Roundtable on Sustainable Palm Oil certification). They found that the awareness of the trusted, official Organic JAS label significantly increased consumers’ willingness to pay more for an eco-labeled organic cosmetic product. However, when a second logo (RSPO)—which was less familiar and less trusted by these consumers—was also present, it diluted the positive effect of the JAS label. In essence, the inclusion of a lesser-known or less credible eco-label introduced skepticism and confusion, undermining the influence of the more trusted label. This finding underscores a delicate point: more eco-labels are not always better. If a label is not widely recognized or believed, it may do more harm than good by planting doubts about the product’s overall credibility. Companies and policymakers, therefore, should be mindful to prioritize well-known, credible certifications and avoid ‘overloading’ consumers with too many logos—especially if some credentials are not strongly trusted.
In Western contexts, a similar gap in eco-label recognition and trust has been documented, which may help explain the slower growth in demand for sustainable dining despite rising environmental awareness. Gorton et al. (Reference Gorton, Tocco, Yeh and Hartmann2021) conducted a multi-country survey in Europe and pointed out that only about 37.6% of European consumers could correctly recognize the EU’s official organic eco-label (the ‘Euro-leaf’). This low recognition rate, combined with general skepticism, greatly dampens the influence that such labels have on purchase behavior. In other words, if nearly two-thirds of consumers do not even recognize a label meant to signal sustainable practices, its power to sway consumer choices is inherently limited. Gorton et al. argue that effective eco-labels require not just rigorous standards but also public understanding. They and other researchers have called for better public education around eco-labels and greater standardization of labeling practices to reduce consumer confusion. A recent study in Croatia provides evidence for how improving label visibility and clarity can build trust. Kovač, Dunković and Kovač (Reference Kovač, Dunković and Kovač2025) surveyed consumers in Croatia—a market increasingly shaped by green marketing but also marred by greenwashing—and found that a few straightforward steps could significantly reduce consumer skepticism. Standardizing the visual presentation of eco-labels on packaging and ensuring strong institutional backing and transparency were identified as pivotal measures. When labels were clearly displayed in a consistent manner and backed by trusted institutions (with transparent verification processes), consumers reported higher trust and were more inclined to make environmentally responsible purchase decisions. These findings suggest that bridging the ‘trust gap’ in eco-labels is feasible through policy and design: for example, governments and industry groups might collaborate on uniform label designs or public awareness campaigns that help consumers more easily identify and understand credible sustainability labels. In short, consumers need to know what a label truly stands for and trust that it is enforced; otherwise, even the most well-intentioned eco-label may have little effect on actual consumer behavior.
By incorporating these recent insights, our study is positioned at the intersection of consumer willingness and consumer trust. On one hand, the growing willingness to pay a premium for sustainable dining options (especially among younger and environmentally conscious consumers) provides a strong motivation for restaurants and food businesses to go green. On the other hand, the persistent gap between environmental awareness and actual purchasing, partly due to limited recognition and trust of eco-certifications, highlights a crucial challenge that must be addressed.
Materials and methods
The AID analysis framework and the survey
Between November 4 and 13, 2023, a sample of 562 respondents from across Taiwan participated in a web-based survey designed to conduct the current research. As mentioned earlier, due to the relatively new concept and form of green restaurants in Taiwan, which have not become fully widespread, it is difficult to measure respondents’ actual consumption behavior. Therefore, the survey aimed to examine whether respondents, under the AID model, could enhance their desire to consume GHG-mitigating ingredients in green restaurants through attention to domestic green restaurant labels and various factors that spark their interest and attract them to patronize green restaurants.
The final questionnaire was divided into five sections, addressing specific issues as follows:
-
(i) The first section (three questions) examined respondents’ understanding of green restaurant labels in Taiwan, including eco-friendly restaurants, AMOT traceability restaurants, and the GDG. A 6-point Likert scale question was used in this section, with 1 corresponding to ‘never heard of’ and 6 to ‘full knowledge’.
-
(ii) The second section (eight questions) covered eight factors that may spark their interest and attract respondents to patronize green restaurants, including the quality and deliciousness of the food, the use of healthy ingredients, the reasonableness of prices, the ambiance of the restaurant, the quality of the restaurant’s service, the convenience of the location, and the restaurant’s support for sustainability goals. A 6-point Likert scale question was used to measure the degree of agreement, with 1 indicating ‘completely disagree’ and 6 indicating ‘completely agree’.
-
(iii) The third section (three questions) includes respondents’ views on green restaurants adopting ingredients that help mitigate GHG emissions. Based on the aforementioned review, key points of dietary change conducive to reducing GHG emissions include using local seasonal ingredients, using ingredients with lower carbon emissions, and offering vegetarian options. In addition, the previously used 6-point Likert scale question was employed (1 = ‘completely disagree’ and 6 = ‘completely agree’).
-
(iv) The fourth section (six questions) analyzed respondents’ expected consumption behavior at green restaurants. Questions 1–3 also used the same Likert scale as before (1 = ‘completely disagree’ and 6 = ‘completely agree’). Meanwhile, Questions 4–6 assessed the frequency of respondents’ willingness to patronize green restaurants (1 = ‘not willing’ and 6 = ‘every week’), the extent of additional money they are willing to spend (1 = ‘not willing’ and 6 = ‘15%’), and the amount they are willing to spend per meal at green restaurants (1 = ‘under 200 NT Dollars’ and 6 = ‘over 1,001 NT Dollars’), respectively.
-
(v) The fifth section (five questions) used demographic questions to cover the sociodemographic aspects of the respondents; the formulated questions evaluated the participants’ sex, age, educational degree, disposable income per month, and area of residence. The 25 indicators across the five sections of the questionnaire are reported in Table 1.
Table 1.The 25 indicators in the questionnaire with the different measurements
Source: Designed and edited by the research group.
Data credibility, validity test, and rotated component
Based on the 562 samples of data collected from the questionnaire, we adopted SPSS 25.0 for the exploratory factor analysis of the 20 observed variables (Sections 1–4), and the results show that the KMO value is 0.867, indicating good data validity (Table 2). Then, we conduct the credibility analysis of the internal consistency of data according to the divided sections. The results show that the α value of the first section—the respondent ‘label understanding’—is 0.833, followed by the respondent’s ‘the factors of dining at a GR’ (0.885), ‘dietary change at GRs’ (0.736), and the respondent ‘expected consumption behavior at GRs’ (0.702), all of which indicate a good internal consistency of data (Table 3).
The KMO test and the Bartlett’s test of sphericity of 20 variables
Source: Derived from the sample data by the authors using SPSS 25.0.
Reliability testing of the four sections
Source: Derived from the sample data by the authors using SPSS 25.0.
Additionally, when we further analyze the rotated component in the orthogonal rotation matrix of the questionnaire, we found that the factors related to dining at green restaurants could be divided into two factors. The first factor consists of aspects related to restaurant service and experience, encompassing Q4, Q5, Q7, Q8, Q9, and Q10, whereas the second factor consists of Q6 and Q11 related to health and sustainability (Table 4). Additionally, Table 4 demonstrates the differences in the rotated component of expected consumption behavior. Q15–Q18 can be interpreted as the willingness to consume, whereas Q19 and Q20 can be interpreted as the actual willingness-to-pay amounts.
Rotated component of the matrix of orthogonal rotation of 20 variables
Source: Derived from the sample data by the authors using SPSS 25.0.
To extract the underlying constructs from the questionnaire items, we performed an exploratory factor analysis with orthogonal (Varimax) rotation using SPSS 25.0. This method allowed us to identify distinct factors with minimal overlap, improving interpretability. Based on the rotated component matrix (see Table 4), we grouped Q15–Q18 as one factor representing general willingness to consume (ECB1), and Q19 and Q20 as a second factor representing actual willingness to pay (ECB2). This clear separation supports the validity of using these two dimensions in the subsequent analysis.
The AID model and the analysis framework
In order to stimulate consumer engagement in green restaurants, these establishments introduce new concepts that encourage the use of ingredients that help mitigate GHG emissions, thereby aiming to change consumers’ dietary habits. The AID model can provide an appropriate theoretical basis for this effort. First proposed by E. St. Elmo Lewis (Reference Lewis1899), the AID model has been widely applied in commercial sales strategies for over a century (with various adjustments over time), particularly in personal selling research (Pramita and Manafe, Reference Pramita and Manafe2022; Seiler and Klaas, Reference Seiler and Klaas2016). According to Montazaribaforoushi, Keshavarzsaleh and Zoëga (Reference Montazaribaforoushi, Keshavarzsaleh and Zoëga2017), the AID model comprises four stages: first, the cognitive stage where the user’s Attention is captured; second, the affective stage where the consumer develops Interest in and understanding of the offering, leading to a Desire to purchase; and finally, the behavioral stage where Action occurs. Therefore, in the context of our study, we apply the AID framework to examine how consumers’ desire to consume GHG-mitigating ingredients in green restaurants can be significantly enhanced.
Within our AID-based analysis framework, we operationalize the constructs of Attention, Interest, and Desire for measurement. To quantify Attention, we assess respondents’ familiarity with green restaurant labels. Specifically, we average each respondent’s understanding scores for the three prominent green restaurant labels currently used in Taiwan (Fig. 1) to create a composite Attention variable. An average score above 4 (corresponding to at least ‘basic understanding’) is coded as 1, indicating a meaningful level of awareness of these labels.
To capture the Interest stage while simplifying its complexity, we categorize the factors sparking respondents’ interest into two groups. The first interest dimension, I1, encompasses service- and experience-related factors in the restaurant (including survey items Q4, Q5, and Q7–Q10). The second interest dimension, I2, covers health- and sustainability-related factors (including items Q3 and Q11). For each respondent, we calculate the average score for the items in each category. If the average score in a category exceeds 5 (equivalent to at least ‘agree’ on the interest-related statements), we interpret that as indicating a strong interest in that dimension, and we code it as 1 for analysis purposes.
Finally, for the Desire stage, we evaluate respondents’ willingness to consume GHG-mitigating ingredients at green restaurants. We conducted a factor analysis (using orthogonal rotation) on six survey questions in Section 4 related to expected consumption behavior. This analysis yielded two constructs: ECB1 (comprised of Q15–Q18), which measures general willingness to consume such ingredients, and ECB2 (comprised of Q19 and Q20), which measures willingness to pay for them. Notably, because ECB2 involves specific expenditure amounts, it serves as a more concrete indicator of intent. We then combined these constructs with responses to the three dietary-change questions from Section 3 to derive two indices: D1 and D2. D1 denotes the willingness to consume seasonal ingredients with lower carbon emissions, and D2 denotes the actual willingness to pay amount. A higher value of D1 indicates that a respondent not only believes green restaurants should provide GHG-mitigating ingredients but is also more willing to consume them, whereas a higher value of D2 reflects a stronger willingness to pay a premium for such offerings in green restaurants. Conversely, lower scores on these indices suggest weaker support or willingness. The models are specified as follows:
$$ D{1}_i=\alpha +{\gamma}_1{A}_i+{\gamma}_2\left({A}_i\times I{1}_i\right)+{\gamma}_3\left({A}_i\times I{2}_i\right)+\beta {X}_i+{\varepsilon}_i, $$
$$ D{2}_i=\alpha +{\gamma}_1{A}_i+{\gamma}_2\left({A}_i\times I{1}_i\right)+{\gamma}_3\left({A}_i\times I{2}_i\right)+\beta {X}_i+{\varepsilon}_i, $$
where the
$ D{1}_i $
and
$ D{2}_i $
are the dependent variables for respondent
$ i $
in two different expected consumer behaviors.
$ {\mathrm{A}}_{\mathrm{i}} $
is a dummy variable that equals 1 if the attention score is above 4 points.
$ \mathrm{I}{1}_{\mathrm{i}}\;\mathrm{and}\;\mathrm{I}{2}_{\mathrm{i}} $
are the dummy variables that equal 1 if the interest score is above 5 points.
$ {\mathrm{X}}_{\mathrm{i}} $
represents a vector of control variables related to respondent’s sociodemographic characteristics.
$ {\unicode{x025B}}_{\mathrm{i}} $
is the random error term.
$ {\gamma}_1 $
,
$ {\gamma}_2 $
,
$ {\gamma}_3 $
, and
$ {\beta}_1 $
are the parameters to be estimated. This study focuses on
$ {\gamma}_2 $
and
$ {\gamma}_3 $
as it captures the interaction effects of attention and interest on desire.
Results and discussion
Description of the sample
Firstly, the results reporting the sociodemographic aspects of the sample used in the present study are depicted in Table 5. It can be seen that there is an overrepresentation of women and older respondents, with 47.2% of the sample aged 50 and above, which may be because people tend to prioritize health as they age, thus having a greater willingness to respond to this questionnaire about green restaurants. A total of 60.1% of the respondents were women, and this higher proportion can be explained by the fact that our respondents were responsible for food purchases in their households, a responsibility typically assumed by women (Sampalean, Rama and Visentin, Reference Sampalean, Rama and Visentin2021). Therefore, women are more likely to be interested in this new concept of restaurant, green restaurants, and are more willing to respond to this questionnaire.
Sociodemographic indicators (N = 562)
More than 40% of the respondents possessed a master’s degree or higher. The prevalence of higher education in Taiwan may explain the high level of education observed in the sample (Tsai, Reference Tsai2015). Additionally, the eastern region of Taiwan showed overrepresentation (42.3%). This could be attributed to the online distribution of the questionnaire, potentially leading to an overrepresentation of respondents from specific geographic areas. Finally, 17.3% reported a monthly disposable income below 20,000 NT dollars, whereas 30.4% reported a disposable income exceeding 50,001 NT dollars.
The overrepresentativeness of our sample might have some influence on the final results. For example, research by Verbeke, Guerrero and Hersleth (Reference Verbeke, Guerrero and Hersleth2012) and Dekhili and Cohen (Reference Dekhili and Cohen2011) indicates that older individuals tend to exhibit greater awareness and utilization of EU quality certifications. As the older groups were overrepresented in our sample, we believe this could result in higher attention scores for green restaurant labels. Having a higher educated sample might have introduced some bias, as higher education levels are typically associated with greater knowledge and attention to green restaurant labels.
Sample statistics in the AID model
Table 6 presents the sample statistics of respondents’ attention, interest, and desire for green restaurants under the AID model. In the first section of the questionnaire, respondents were presented with three of Taiwan’s GR labels: eco-friendly restaurant, AMOT traceability restaurant, and GDG. They were then asked to rate their understanding of each label on a 6-point scale. As the eco-friendly restaurant label was the earliest initiative in 2012 in Taiwan’s green restaurant movement (Ministry of Environment, 2024a), it was found to be the most recognized, with an average score of 3.18. Furthermore, according to the statistical findings from the second section, the use of healthy ingredients was found to be the most compelling factor driving respondents’ interest in patronizing green restaurants, with an average score of 5.22 points. Previous studies have consistently shown that health consciousness is a key factor explaining individuals’ intentions to dine at green restaurants (Kim et al., Reference Kim, Lee, Kim and Kim2013; Lee et al., Reference Lee, Conklin, Cranage and Lee2014; Shin et al., Reference Shin, Im, Jung and Severt2019).
Min., max., means, and standard deviation of the variables in the different sections (N = 562)
The strong correlation between health and green restaurants is further evidenced by respondents’ emphasis on the importance of green restaurants using ingredients that help mitigate GHG emissions. In the third section of the questionnaire, respondents indicated that the use of local seasonal ingredients by green restaurants was the most important factor in dietary change (5.51 points). This preference is likely due to the perception that local seasonal foods/ingredients are often less processed, which has positive effects on both human health and GHG emissions (Fardet and Rock, Reference Fardet and Rock2020).
Additionally, it is noteworthy that in the fourth section of the questionnaire, respondents, on average, tended to agree with prioritizing green restaurants (4.63 points), recommending them (4.77 points), and paying higher prices for meals. Nevertheless, when asked about the actual amounts they were willing to pay, the results were less supportive: only 5%–7% of respondents indicated a willingness to pay more (3.91 points on average), and the specific amounts they were willing to pay for meals fell between 201 and 400 NT dollars, and between 401 and 600 NT dollars (2.78 points). This indicates that in Taiwan’s green restaurants, there also exists a need to balance green practices with restaurant-quality factors (e.g., price and quality of other attributes) (Park et al., Reference Park, Chae, Kwon and Kim2020).
Group comparison of different label attention levels of green restaurants
Table 7 presents the differences in green restaurant label attention across various levels: low (3–7 points), middle (8–12 points), and high (13 points and above, out of a total of 18 points), among different sample groups. Perhaps owing to the prevalence of the internet, such as Google reviews and Facebook, the results indicate that sociodemographic variables such as gender, education, and monthly disposable income exhibit no significant differences in the different levels of green restaurant label attention (Božić and Milošević, Reference Božić and Milošević2021; Božić and Zubanov, Reference Božić and Zubanov2018).
Group comparison of different green restaurant label attention levels (N = 562)
Note: ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively. ✮ indicates continuous variables, with values being the mean, and standard deviations shown in parentheses.
Conversely, younger respondents tend to have lower levels of label attention, whereas older respondents tend to exhibit higher levels of label attention. These findings are in line with the study by Sarmiento and Hanandeh (Reference Sarmiento and Hanandeh2018), which finds that older age groups (above 45 years old) are more supportive of the concept of green restaurants. Additionally, across all regions in Taiwan, there is a decreasing trend in the label attention level, indicating fewer respondents with higher label attention as geographic regions become more specific. These results may indirectly suggest that green restaurants are not yet widespread in Taiwan, leading respondents from all regions to have relatively low attention to green restaurant labels.
Furthermore, respondents with higher label attention levels tend to have more interest in health and sustainability (I2). Conversely, there are no differences in green restaurant label attention levels for aspects related to restaurant service and experience (I1), which spark customer interest. These findings are similar to those of Lo, King and Mackenzie (Reference Lo, King and Mackenzie2020), who noted that customers with higher levels of health and environmental consciousness tend to perceive green restaurant menus more favorably. Additionally, Kwok, Huang and Hu (Reference Kwok, Huang and Hu2016) and Jo, Joung and Taylor (Reference Jo, Joung and Taylor2023) suggest that restaurant sustainability practices positively influence customer evaluations of sustainability efforts, rather than food-focused and administration-focused attributes.
Finally, there are significant differences in the desire to consume at green restaurants that contribute to mitigating GHG emissions among respondents with different levels of label attention. Overall, respondents express the highest desire to consume local seasonal ingredients at green restaurants, followed by low-carbon ingredients and vegetarian options. Moreover, as label attention levels increase, respondents also exhibit a higher desire to consume these three kinds of ingredients at green restaurants.
The results of the AID model estimation
Table 8 presents the estimation results of the AID model, examining the impact of label attention and various interests on desired dietary changes in green restaurants. First, without any interaction effects of interest, respondents purely with higher attention to green restaurant labels (above 4 points) show a significant decrease of −3.417, −3.771, and −3.706 points in Desires 1-1 to 1-3. That is to say, even though respondents may be familiar with green restaurant labels through others or the internet, they will deliberately avoid visiting green restaurants when they lack interest. The results of this study lend support to some studies that indicate the weak connection between subjective (social) norms and intention, especially in areas where green restaurants are newly introduced, as individuals may lack environmental awareness (Armitage and Conner, Reference Armitage and Conner2001; Elhoshy, Reference Elhoshy2020; Visschers, Wickli and Siegrist, Reference Visschers, Wickli and Siegrist2016).
Estimation results of dietary change desired (D1) in green restaurants (N = 562)
Note: ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.
Additionally, it is noteworthy that when combined with attention, service and experience-focused interest (I1) and healthy and sustainability-focused interest (I2) can positively enhance people’s desire to consume GHG-mitigating ingredients in green restaurants. Moreover, as observed in a study by Liu, Cai and Zhu (Reference Liu, Cai and Zhu2015), the dining choice of Chinese ethical eaters in green restaurants is culturally influenced, particularly regarding the adoption of green eating practices, such as vegetarian diets. Therefore, when controlling for other variables, consumers influenced by attention and I1 show the highest willingness to consume vegetarian dishes (D1-3) in green restaurants (3.814). Conversely, consumers influenced by attention and I2 exhibit the highest willingness, particularly driven by environmental considerations, to consume low-carbon-emission (D1-2) ingredients in green restaurants (4.387).
However, in stark contrast to Table 8, when we further change the desire for consumption into the concrete concept of actual payment amounts, we find that the AID model fails to enhance people’s desire to consume GHG-mitigating ingredients in green restaurants. Furthermore, age and disposable income emerge as the primary factors influencing desire. In Table 9, older individuals, especially those over 60 years old, exhibit a stronger inclination toward consuming GHG-mitigating ingredients in green restaurants and are willing to pay higher amounts specifically, with Desire scores increasing by 4.894, 4.722, and 4.923 points for Desires 2-1 to 2-3, respectively. These research findings may be culturally influenced. Studies on dietary habits among the elderly show differences between Eastern and Western cultures. For instance, in the study by McKie et al. (Reference McKie, Maclnnes, Hendry, Donald and Peace2001), Scottish older adults were found to overlook balanced dietary habits; however, in the research by Danyuthasilpe et al. (Reference Danyuthasilpe, Amnatsatsue, Tanasugarn, Kerdmongkol and Steckler2009), Thai elderly individuals valued balanced diets due to shared beliefs, values, and customs. Moreover, respondents with higher disposable income are also more willing to spend more on GHG-mitigating ingredients in green restaurants, unaffected by the AID model. These findings align with the research conducted by Langgat (Reference Langgat2020) and Manaktola and Jauhari (Reference Manaktola and Jauhari2007).
Estimation results of dietary change desired (D2) in green restaurants
Note: ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.
That is to say, when not involving specific monetary concepts, the AID model effectively enhances people’s desire for GHG-mitigating ingredients in green restaurants. However, when considering the amount they are willing to pay, the only factors that still influence GHG-mitigating ingredients desire are age and disposable income influenced. This may suggest that while most Taiwanese are indeed willing to make dietary changes in green restaurants, the positioning of green restaurants in Taiwan is still unclear, leading consumers to perceive them as either general fast food restaurants or upscale casual ones. DiPietro and Gregory’s (Reference DiPietro and Gregory2013) study showed that fast food customers are more concerned about price and restaurant appearance, whereas upscale casual customers prioritize environmental record and whether the restaurant has recycling bins.
Conclusion and implications
This study explores how individuals’ desire to consume GHG-mitigating ingredients in green restaurants can be enhanced within the analysis framework of the AID model. Given that this novel type of restaurant, green restaurants, is not yet widespread in Taiwan, and there may be a lack of sustainable social norms and values, individuals with higher attention to the green restaurant label may deliberately oppose supporting green restaurants and their offerings of dietary change if they are not interested.
Additionally, although label attention combined with either service and experience-focused interest or healthy and sustainability-focused interest can significantly enhance consumers’ desire to consume GHG-mitigating ingredients in green restaurants, the AID model becomes insufficient when further investigating the actual price consumers are willing to pay for this desire. Instead, factors such as age and disposable income, which may be influenced by cultural and financial factors, come into play. These findings are inconsistent with previous studies revealing consumers’ willingness to pay a premium for green restaurants (Namkung and Jang, Reference Namkung and Jang2017; Nicolau et al., Reference Nicolau, Guix, Hernandez-Maskivker and Molenkamp2020).
Considering the prevalence of dining out culture and the current developmental bottleneck faced by green restaurants in Taiwan, this study suggests that government agencies should provide enhanced subsidies to restaurants with green restaurant labels. This would enable them to compete with regular restaurants in terms of pricing and expedite their integration into people’s daily lives, thereby changing dietary habits and promoting the adoption of GHG-mitigating ingredients.
The research acknowledges limitations, including issues of sample overrepresentation, such as an overabundance of respondents residing in Eastern Taiwan, a higher proportion of females compared to males, and a lack of measurement regarding specific actions of respondents dining at green restaurants. It suggests that future research could incorporate relevant indicators of actions to ensure that desire transforms into concrete actions, completing the application study of the AID model in promoting GHG-mitigating ingredients in green restaurants.
Use of artificial intelligence (AI) tools
Artificial intelligence tools were used solely to assist in formatting and organizing the reference list during manuscript preparation. No part of the manuscript’s conceptual development, data analysis, writing, or interpretation was generated by AI. All research design, content creation, and analysis were conducted entirely by the authors.
Data availability statement
The data supporting the findings of this study are available from the corresponding author upon reasonable request.
Funding statement
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Competing interests
The authors declare no competing interests.
Open access
