Green Purchase Behavior in Indonesia: Examining the Role of Knowledge, Trust and Marketing

Abstract

This study investigates the factors influencing green purchase behavior in emerging economies, focusing on Indonesian consumers’ preferences for organic food products. While sustainability awareness is growing globally, limited research has examined how environmental knowledge and trust interact with marketing efforts to shape green purchasing decisions in developing market contexts like Indonesia. The research model incorporates five constructs: environmental knowledge (awareness of ecological issues), green trust (confidence in environmental claims), green marketing mix (eco-oriented strategies for product, price, place, and promotion), green purchase intention (likelihood of buying eco-friendly products), and green purchase behavior (actual sustainable buying decisions). Data from 211 valid respondents were analyzed using structural equation modeling. The results indicate that environmental knowledge directly influences green trust and the green marketing mix but not green purchase intention or behavior. Instead, it affects behavior indirectly through trust and intention. Contrary to expectations, green trust does not influence the green marketing mix, suggesting it may operate independently of marketing strategies. Similarly, the green marketing mix does not significantly influence green purchase intention or behavior, suggesting that marketing strategies alone may be insufficient in driving sustainable consumer choices. These findings highlight the important role of environmental knowledge in fostering consumer trust and indirectly guiding green purchasing behavior in emerging markets. By promoting sustainable consumption through knowledge and trust, this study offers insights into consumer behavior as a pathway to advancing planetary health. This study advances the Theory of Planned Behavior by integrating green trust and the green marketing mix to explain how trust and economic factors shape green purchasing behavior. Practical implications suggest that businesses should adopt targeted green marketing strategies, such as educational campaigns, eco-labeling, or certifications, to enhance environmental awareness, build consumer trust, and encourage sustainable purchasing decisions. This study contributes to the literature by examining how environmental knowledge indirectly influences green purchase behavior through the mediation of trust and intention within the context of an emerging market.

1. Introduction

The global population increased by over 10% between 2000 and 2019 [1,2]. This growth led to higher demand for natural resources and accelerating environmental degradation, particularly in consumer-driven sectors [3]. In response, consumers and businesses have increasingly shifted toward sustainable consumption, with eco-friendly or “green” products emerging as a central strategy for mitigating the adverse impacts of consumption on climate and ecological systems, especially in the retail sector [4]. Retailers face increasing pressure from stakeholders and consumers to adopt sustainability-oriented practices, while consumers demand greater transparency and ethical responsibility from businesses [5,6]. Consequently, green consumption behavior has become a critical focus for researchers and practitioners seeking to align commercial practices with environmental goals [7].

Consumer behavior is progressively shifting toward sustainability, as evidenced by a growing interest in green product purchases [8]. Consumers increasingly recognize such products for their resource efficiency, waste reduction, and minimal ecological impact [9,10,11]. Accordingly, green consumption encompasses behaviors intentionally directed toward environmental preservation, such as selecting items with low carbon footprints or sustainably sourced materials [12,13].

Despite increasing awareness, numerous studies highlight a persistent attitude–behavior gap, a phenomenon where consumers express pro-environmental attitudes but fail to engage in consistent sustainable purchasing [14,15,16]. Researchers attribute this disconnect to multiple barriers, including social norms, cost, product availability, and lack of trust in environmental claims [17,18]. While these factors have been extensively examined in global contexts, scholars have not yet thoroughly explored how these factors manifest in emerging markets such as Indonesia, particularly concerning green food products. Addressing this research gap, this study is among the first to explore the combined influence of green trust and the green marketing mix within Jakarta’s rapidly urbanizing consumer context.

To provide a comprehensive understanding of green purchasing behavior in this setting, the present study integrates three theoretical frameworks: the Theory of Planned Behavior (TPB), green trust, and the green marketing mix. TPB provides a foundational model for understanding the link between intentions and behavior, positing that behavioral intention is shaped by individual attitudes, perceived behavioral control, and subjective norms [19,20,21]. Scholars define green trust as consumers’ confidence in the credibility of environmental claims, which is especially relevant in markets susceptible to greenwashing [22,23,24,25]. Additionally, the green marketing mix, comprising eco-adaptations of the traditional 4Ps (product, price, place, and promotion), serves as a key mechanism for aligning business practices with sustainability principles [26,27,28,29].

In the Indonesian context, climate change and unsustainable land-use practices exacerbate environmental degradation [30]. Indonesia has committed to reduce its greenhouse gas emissions under the Paris Agreement [31], yet rising consumption demands continue to drive deforestation, palm oil production, and agricultural expansion, which put pressure the environment [32,33,34]. Despite these challenges, Indonesia’s organic food market and green product adoption have grown gradually in recent years, signaling a shift toward sustainable consumption [35,36]. Rising incomes, urbanization, and higher education levels contribute to changing consumption habits, especially in urban centers like Jakarta, which suffer from severe air pollution and ecological stress [37,38].

Several recent studies have investigated green consumption in Indonesia. Suhartanto et al. [39] identified a rise in plant-based diets among younger consumers, while Suhartanto et al. [40] examined the roles of environmental concern, cultural values, and religiosity. Building on these findings, Suhartanto et al. [41] integrated cultural and religious perspectives, offering a more holistic view of green purchasing behavior across Southeast Asia. While these studies have enriched our understanding of green attitudes in Indonesia, they have largely overlooked the urban-specific dynamics shaping consumer behavior.

Jakarta, as a rapidly urbanizing metropolis, presents a distinctive and under-researched context. Although prior work has addressed general environmental attitudes [39,40], researchers have yet to fully identify the specific drivers of green purchasing behavior in Jakarta’s retail sector. To address this gap, the present study examines the influence of environmental knowledge, green trust, and perceptions of the green marketing mix on eco-conscious purchasing decisions. Using a quantitative approach, this research seeks to uncover the behavioral mechanisms underlying sustainable consumption in Jakarta, offering insights for retailers and policymakers aiming to advance green consumer practices in alignment with Indonesia’s Sustainable Development Goals (SDGs).

By investigating the psychological and contextual drivers of sustainable consumption, this study advances a systems-based perspective on planetary health, demonstrating how individual purchasing decisions in emerging urban markets influence both environmental sustainability and human well-being [42,43,44]. This approach aligns with global sustainability agendas that emphasize integrated frameworks linking ecological resilience with consumer behavior [42,45,46].

Thus, the present study examines how environmental knowledge, green trust, and the green marketing mix shape green purchase intention and actual behavior among urban consumers in Jakarta. Specifically, it investigates (1) the mediating roles of green trust and purchase intention in the relationship between environmental knowledge and green behavior, and (2) the moderating effect of contextual economic constraints, particularly price sensitivity, on sustainable consumption pathways.

The remainder of this article is structured as follows: Section 2 reviews the literature on green purchasing and trust-based theoretical frameworks. Section 3 outlines the methodology, including data collection procedures and the application of structural equation modeling (SEM). Section 4 presents the empirical results, followed by a discussion of theoretical and managerial implications in Section 5. Section 6 concludes this study.

2. Literature Review

2.1. Conceptual Framework

Researchers frequently use the terms “green,” “sustainable,” and “environmental” interchangeably to describe consumer behaviors that positively impact ecological outcomes [47]. They typically associate green products with sustainable resource use, eco-friendly manufacturing, waste reduction, recycling, and energy conservation [48].

This study’s conceptual framework identifies five key determinants of green purchasing behavior: environmental knowledge, green trust, the green marketing mix, green purchase intention, and actual purchasing behavior. Within this model, consumers perceive environmental knowledge and green marketing practices as antecedents that shape both attitudes and perceived behavioral control [49]. The framework extends the Theory of Planned Behavior (TPB) by incorporating green trust, which refers to consumer confidence in the credibility and integrity of environmental claims, which enhances attitudinal certainty and strengthens purchase intention [50,51].

Although prior studies often utilize the Stimulus–Organism–Response (S-O-R) framework, positioning the green marketing mix as an external stimulus that generates trust [52], this study offers an alternative perspective. It conceptualizes green trust as an internal cognitive–affective filter that mediates how consumers interpret the four Ps of the green marketing mix: product, price, promotion, and place [53]. This framework is particularly salient in emerging markets such as Indonesia, where economic pressures and consumer skepticism often temper pro-environmental attitudes [54,55].

To further contextualize the model, this study integrates regional insights from Southwest and Southeast Asia. In Gulf Cooperation Council (GCC) countries, such as Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the UAE, green marketing strategies rooted in Islamic ethical values have enhanced green trust and increased sustainable purchase intentions [56,57]. Similarly, research in Malaysia highlights how religiosity, normative beliefs, and environmental knowledge shape green behavior [58]. In Indonesia, particularly in urban areas like Jakarta, green consumption increasingly reflects a combination of cultural, environmental, and religious influences [46,59]. While sustainability awareness continues to grow, significant gaps remain between awareness and consistent green purchasing behavior among Jakarta’s consumers [60]. These regional parallels underscore the importance of religion-based environmental ethics, urban consumer dynamics, and trust in green claims as key drivers of green consumer behavior.

Based on this synthesis, the proposed model positions environmental knowledge, green trust, and the green marketing mix as key antecedents of green purchase intention. The green marketing mix, comprising product, price, place, and promotion, serves as a strategic lever that enables firms to translate sustainability values into concrete market offerings and communication strategies [61]. By embedding sustainability across all marketing elements, firms influence consumer perceptions, align attitudes, and build credibility in their environmental positioning [62]. This approach is particularly relevant in emerging markets, where price sensitivity, value-driven consumption, and economic pragmatism often moderate pro-environmental attitudes [54,55]. Accordingly, this study examines how the four components of the green marketing mix shape green purchase intentions and behavior beyond mere environmental awareness or trust.

Aligned with the Theory of Planned Behavior, this study positions green purchase intention as a mediating variable that links environmental knowledge, green trust, and marketing perceptions to actual consumer behavior [63]. By grounding the model in both theoretical and regional contexts, this study provides a comprehensive and culturally sensitive framework to understand sustainable consumption in Indonesia. Moreover, it fills a critical gap by examining how psychological and contextual factors interact to shape green behavior in Jakarta, a rapidly urbanizing city facing significant environmental challenges.

2.2. Theoretical Framework and Hypotheses

2.2.1. Environmental Knowledge (EK)

Environmental knowledge refers to an individual’s awareness and understanding of environmental challenges and sustainable practices, comprising both system-oriented (“what”) and action-oriented (“how”) knowledge [64]. Consumers who possess a strong understanding of environmental issues are better equipped to evaluate the credibility of corporate green initiatives, thereby enhancing trust in sustainability claims [22,65]. Moreover, environmental knowledge bolsters consumer perceptions of corporate authenticity and commitment to sustainability, which in turn reinforces trust [66,67].

Environmentally literate consumers expect sustainability to be embedded across all four elements of the marketing mix [68], and this awareness enables them to make more informed, values-driven purchasing decisions [69,70]. However, despite growing awareness, many consumers still lack the depth of understanding necessary to translate environmental concern into concrete green purchasing behavior [71,72]. Nonetheless, environmental knowledge remains a critical enabler of pro-environmental action [73].

In Indonesia, particularly in Jakarta, consumers with higher environmental knowledge show greater trust in brands that visibly demonstrate environmental responsibility [59]. This literacy enhances consumer responsiveness to product claims, price positioning, promotional strategies, and distribution practices [74]. Research also suggests that environmentally aware millennials exhibit significantly stronger intentions to purchase eco-friendly products [75]. However, persistent knowledge gaps, especially among younger demographics, continue to impede the adoption of sustainable consumption behaviors [60,76].

Given these insights, the following hypotheses are proposed:

• H1: Environmental knowledge positively influences green trust.
• H2: Environmental knowledge positively influences the green marketing mix.
• H3: Environmental knowledge positively influences green purchase intention.

2.2.2. Green Trust (GT)

Green trust refers to a consumer’s willingness to rely on a brand’s environmental claims, grounded in perceptions of the firm’s competence (i.e., its ability to deliver on sustainability) and benevolence (i.e., its sincerity in pursuing environmental goals) [22]. It functions as a cognitive–affective mechanism that shapes how consumers evaluate product quality, pricing, promotional messaging, and distribution strategies within the green marketing mix [52,53]. Higher levels of green trust enhance the credibility and effectiveness of sustainability-focused marketing efforts, thereby promoting stronger intentions to engage in green purchasing behavior [61].

Conversely, greenwashing—where firms exaggerate or falsify environmental claims—can erode this trust and foster consumer skepticism [77,78]. This concern is particularly salient in contexts where consumers may lack the expertise to distinguish genuine sustainability practices from superficial marketing efforts [25]. In Jakarta, green trust is growing among environmentally conscious youth; however, persistent skepticism highlights the need for transparent environmental communication and credible eco-labeling practices [74,79].

Green trust not only enhances consumer interpretation of the marketing mix but also serves as a mediator between environmental messaging and green purchase intention.

Accordingly, the following hypotheses are proposed:

• H4: Green trust positively influences the green marketing mix.
• H5: Green trust positively influences green purchase intention.

2.2.3. Green Marketing Mix (GMM)

Green marketing refers to the integration of sustainability principles across the marketing lifecycle, from product development to post-consumption disposal [26]. It comprises the traditional 4Ps:

• Product: Environmentally sustainable goods that support ecological well-being [27].
• Price: Reflects sustainability attributes, potentially justifying higher prices through long-term environmental and social benefits [80,81].
• Place: Emphasizes efficient and sustainable distribution that reduces ecological footprints [82].
• Promotion: Communicates the environmental value and benefits of the offering [28,29].

The green marketing mix shapes consumer perceptions of product quality and brand integrity, thereby influencing purchase intentions [83,84]. Consumers often associate green products with higher quality and ethical value, increasing their likelihood of purchase [85]. Alignment of marketing strategy with consumer eco-values further amplifies this effect [86].

Nevertheless, the effectiveness of the green marketing mix is not guaranteed. When sustainability claims are vague, opportunistic, or poorly communicated, consumers may become skeptical [87,88,89].

In Jakarta, targeted green promotions and eco-labeling have proven especially effective among younger, urban consumers [74,90]. Transparent communication and value-driven branding have been shown to convert intention into behavior [59,75]. However, in emerging markets, price sensitivity, limited access, and lack of trust can undermine even well-designed green strategies [91,92].

Based on these insights, we propose the following hypotheses:

• H6: The green marketing mix positively influences green purchase intention.
• H7: The green marketing mix positively influences green purchase behavior.

2.2.4. Green Purchase Intention (GPI)

Green purchase intention reflects a consumer’s deliberate plan to select environmentally responsible products. Grounded in the Theory of Planned Behavior [19,63], attitudes, social norms, and perceived behavioral control shape this intention. Additional factors, such as price sensitivity, perceived value, product attributes, and ethical alignment, further influence the formation of these intentions [93,94].

Strong intentions are generally reliable predictors of actual green purchasing behavior, particularly when attitudes align with consumer actions [95,96,97,98,99]. However, the well-documented “intention-behavior gap” suggests that consumers often fail to act on their sustainable intentions due to barriers such as high cost, limited accessibility, or ingrained consumption habits [100,101].

In Indonesia, this gap is especially prominent, as economic trade-offs often outweigh environmental motivations [102,103,104].

Accordingly, the following hypothesis is proposed:

• H8: Green purchase intention positively influences green purchase behavior.

2.2.5. Green Purchase Behavior (GPB)

Green purchasing behavior involves the intentional selection of products and services that support environmental sustainability [105]. Researchers commonly employ the TPB and other behavioral models to explain and predict these behaviors [20,21]. Studies show a positive correlation between environmental literacy and pro-environmental actions, particularly when consumers recognize sustainability efforts within the marketing mix [106,107,108,109].

However, environmental knowledge does not automatically translate into behavior. As Goh and Balaji [110] highlight, even knowledgeable consumers may fail to act sustainably in the absence of behavioral control or personal norms. Similarly, Zsóka et al. [111] demonstrated that environmental education does not consistently translate into action unless personal norms and contextual facilitators support it. This “knowledge–action gap” is particularly evident in developing economies, where affordability, convenience, and habitual purchasing patterns often outweigh eco-conscious preferences [103,112]. In Jakarta, although interest in sustainable consumption is rising, practical constraints such as affordability and access remain significant barriers [60,76].

In light of this, the final hypothesis is proposed:

• H9: Environmental knowledge positively influences green purchase behavior.

2.3. Research Model

This study’s conceptual framework examines the interplay between environmental knowledge, green trust, and the green marketing mix, evaluating their combined influence on green purchase intention and actual sustainable purchasing behavior. Figure 1 illustrates the proposed model, emphasizing the interrelationships among these constructs and their cumulative effect on sustainable consumer decision making.

3. Materials and Methods

3.1. Research Design

This study employed a quantitative, cross-sectional, and correlational research design grounded in the positivist paradigm to examine behavioral drivers of green food consumption in the Jabodetabek region of Indonesia (Jakarta, Bogor, Depok, Tangerang, and Bekasi). The research focused on consumer behavior toward eco-friendly food products in supermarkets and hypermarkets, specifically assessing the influence of environmental knowledge, green trust, and the green marketing mix on green purchase intention and green purchase behavior [113].

We encountered several methodological constraints due to pandemic-related restrictions and disparities in digital access, which are discussed in detail in Section 5.3 (Research Limitations). Therefore, we adopted a fully quantitative online survey to ensure both feasibility and methodological rigor. We chose a quantitative approach due to its capacity to statistically measure relationships among variables and extrapolate findings to a wider population [114]. This study employed a cross-sectional design, with the main survey data collected between October and November 2022. This design effectively captured behavioral trends during the post-COVID-19 recovery period [115,116].

The correlational nature of this study enabled the examination of associations between the independent variables (environmental knowledge, green trust, and green marketing mix) and the dependent variables (purchase intention and behavior) without manipulation [117]. This methodological approach aligns with the positivist paradigm, which assumes an objective reality measurable through systematic observation and scientific reasoning [118].

3.2. Data Collection

This study employed both primary and secondary data sources to ensure a comprehensive understanding of consumer behavior toward green food products in the Jabodetabek region.

The overall data collection process for this study occurred between March and November 2022. During the preliminary phase (March–October), the research team developed the questionnaire, conducted a pilot test, and finalized the distribution strategy. We administered the primary survey between October and November 2022 using an online questionnaire distributed via Google Forms. The target population consisted of consumers aged 18 years and above who had previously purchased green food products from supermarkets and hypermarkets within Jabodetabek. These individuals were considered more likely to be responsive to ethical and environmental concerns and influence household purchasing decisions [119,120]. Furthermore, Jabodetabek was selected as the study location due to its demographic diversity, high green consumption growth [121], and its role as a microcosm of Indonesia [122]. Consumers in this area also tend to exhibit higher levels of environmental awareness and education [123].

Although stratified random sampling was initially considered, pandemic-related limitations necessitated the use of convenience sampling [124]. The researchers disseminated the questionnaire through their personal networks using widely used digital platforms such as WhatsApp, due to government-imposed restrictions and health concerns that limited in-person data collection and institutional access. While this method restricted generalizability, they made efforts to ensure representation across all five cities in Jabodetabek to enhance this study’s contextual relevance.

We collected 211 valid responses from an initial pool of 262 participants, yielding a valid response rate of 80.5%. This exceeds the minimum sample size requirements for structural equation modeling (SEM), which typically requires 150–200 observations [125]. Additionally, following Haryono and Wardoyo’s guideline suggesting 5–10 responses per estimated parameter, and given that this study includes five constructs and eight parameters, the minimum required sample size would be 40 (5 × 8). Thus, the final sample size exceeds the thresholds necessary for robust SEM using Maximum Likelihood Estimation (MLE), which performs reliably with 150–400 observations [126,127].

The researchers also reviewed secondary data to support and contextualize the primary findings. These included the academic literature, market trend reports, and prior research on green consumer behavior and sustainable purchasing trends. Insights from these sources informed this study’s theoretical foundation, strengthened its analytical framework, and added contextual depth to the empirical analysis [128].

3.3. Instrumentation

The final questionnaire consisted of 33 items divided into two sections. The first section collected demographic data and included filter questions on prior green food purchases and current residence in Jabodetabek. Additional demographic variables included gender, age, education, total monthly household expenditure, and place of residence (Appendix A,

Table A1. Demographic and filter questions.

Variable	Measured Variable	Scale	Rationale
Soppping Experience: Have you ever shopped organic food product at a supermarket or hypermarket?	Yes No	0 1	Filters respondents to ensure they have lifetime experience purchasing organic food at a supermarket or hypermarket. If respondents don’t have the shopping experience, they cannot access the next questions.
Area/Region: In which area/region of Jabodetabek do you live in Indonesia?	Jakarta	0	Filters respondents to ensure they are from the defined study region. If respondents live outside the Jabodetabek area in Indonesia, they cannot access the next questions. This question helps the researcher track the unit of analysis.
	Bogor	1
	Depok	2
	Tangerang	3
	Bekasi	4
	Others	5
Gender	Male	0	Explores differences in green purchasing behavior by gender.
	Female	1
	Others	2
Age	18–24	0	Examines the relationship between age and the research model variables.
	25–35	1
	36–45	2
	46–55	3
	55 and more	4
Total household Expenditure Rate per Month	≥Rp7,500,000	0	Assesses whether household expenditure influences green product purchasing behavior.
	Rp5,000,001–Rp7,500,000	1
	Rp3,000,000–Rp5,000,000	2
	Rp2,000,001–Rp3,000,000	3
	Rp1,500,001–Rp2,000,000	4
	Rp1,000,001–Rp1,500,000	5
	≤Rp1,000,000	6
Education Level	Postgraduated (Master/Doctoral)	0	Evaluates the influence of educational background on green purchasing behavior.
	Graduated (Bachelor)	1
	Secondary (High School)	2
	Vocational and less	3
Place of Living	Village (Rural Area)	0	Investigates whether place of residence affects green behavior as a daily lifestyle factor.
	Town (Suburban Area)	1
	City (Urban Area)	2

Note: The design and adaptation of demographic and filter questions were informed by previous studies, including Abdel Wahab et al. [211], Dąbrowski et al. [212], Feng et al. [213]), İnan & Konyalı [214], Simanjuntak et al. [215], Wang et al. [216], Witek and Kuźniar [217], and Zeynalova & Namazova [218]. Socio-economic classifications follow standards from Badan Pusat Statistik (BPS) [155,156].

).

The second section focused on behavioral constructs related to green consumption. These constructs were selected based on prior validated research and grounded in this study’s theoretical framework to ensure conceptual alignment [129]. Specifically, five behavioral constructs were measured: environmental knowledge (3 items), green trust (5 items), green marketing mix (12 items, with 4 items each for product, price, place, and promotion), green purchase intention (5 items), and green purchase behavior (4 items). Details are shown in Appendix A,

Table A2. Measurement items.

Latent Variable	Code	Measurement	No. of Item
Environmental Knowledge (ENK)	ENK1–ENK3	ENK1: I can tell if the appliances I bought are good for the environment ENK2: I know more about recycling than other ordinary people ENK3: I thoroughly know about environmental issues	3
Green Trust (GTR)	GTR1–GTR5	GTR1: I feel that organic food product environmental commitments are generally reliable GTR2: I feel that organic food product environmental performance is generally dependable GTR3: I feel that organic food product environmental argument is generally trustworthy GTR4: Organic food product environmental concern meets my expectations GTR5: Organic food product keeps promises and commitments for environmental protection	5
Green Marketing Mix (GMM)	GPO1–GPO4	Green Product (GPO) GPO1: Green product companies focus on manufacturing products that have the lowest rate of negative human reflection. GPO2: Green product companies contribute to producing green products with less pollution. GPO3: There is an effective control on green products that green product companies produce. GPO4: Green product companies make products free of strong toxic materials.	16
	GPR1–GPR4	Green Price (GPR) GPR1: Green product companies raise the prices of their products, which negatively affects the environment that happens due to misusage. GPR2: Increased prices of green products sometimes stop me from purchasing them. GPR3: The price difference between green products and conventional products is large. GPR4: Green products have a price that is proportional to their quality
	GPL1–GPL4	Green Place (GPL) GPL1: Environmentally friendly products are sold at reputable agents GPL2: Green product companies make delivery easy. GPL3: Green products companies aim to work with environmentally friendly agents. GPL4: The stores of green product companies are clean.
	GPM1–GPM4	Green Promotion (GPM) GPM1: Green product companies devote a special day to the environment. GPM2: Green product companies favor hosting environmental activities, festivities, seminars, and conferences. GPM3: Employees of green products companies advise customers on how to use their products not to harm the environment GPM4: Green product companies contribute to supporting environmental centers.
Green Purchase Intention (GPI)	GPI1–GPI5	GPI1: I will consider buying green products as they are less polluting in the near future/in coming times GPI2: I will consider switching to green product brands for ecological reasons GPI3: I aim/plan to pay out/to spend more on green product rather than conventional product GPI4: I expect to purchase product in the future because of its positive environmental contribution GPI5: I certainly want to buy green products in coming future	5
Green Purchase Behavior (GPB)	GPB1–GPB4	GPB1: I try to buy green product. GPB2: I have switched to buy green products because of the environmental benefits. GPB3: When I choose between the same type of products, I purchase the ones that are less harmful to the environment. GPB4: I buy green products even if they are more expensive than nongreen ones	4
Total Item			33

Note: Likert scale: (1) Strongly Disagree; (2) Disagree; (3) Neutral; (4) Agree; (5) Strongly Agree. The design and adaptation of measurement items were informed by previous studies, including Mostafa [131], Maichum et al. [132], Kamalanon et al. [133], Guerreiro and Pachero [134], Augusto et al. [135], Witek and Kuźniar [136], Munamba and Nuangjamnong [137], Liu and Kim [138], Cai et al. [139], Chen and Deng [140], and Correia et al. [141].

. All items were adapted from previously validated instruments and measured on a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree), consistent with the standard practices in green consumer behavior research [130].

Three items adapted from Mostafa [131], Maichum et al. [132], and Kamalanon et al. [133], assessed environmental knowledge (ENK), capturing consumers’ awareness of eco-friendly products, recycling practices, and environmental issues. Green trust (GTR), based on Guerreiro and Pacheco [134], Román-Augusto et al. [135], and Witek and Kuźniar [136], included five items evaluating trust in organic products’ reliability, dependability, environmental concern, and fulfillment of environmental commitments. The green marketing mix (GMM), based on Munamba and Nuangjamnong [137] and supported by Liu and Kim [138] and Cai et al. [139], included four areas, product, price, place, and promotion, each measured with four questions about how consumers view the quality of green products, fair pricing, store availability, and marketing efforts that support environmental issues. Five items adapted from Maichum et al. [132], Kamalanon et al. [133], Witek and Kuźniar [136], and Chen and Deng [140], measured green purchase intention (GPI), which examined consumers’ intention to buy green products, including their willingness to switch brands, pay a premium, and prioritize environmentally friendly alternatives. Finally, four items based on Kamalanon et al. [133] and Correia et al. [141] assessed green purchase behavior (GPB), capturing actual purchasing actions such as consistent selection of green products, brand switching, and choosing less harmful products despite higher costs. By structuring the questionnaire with clearly defined constructs and adapting established scales, this research ensured the validity and reliability of the instrument. This structured measurement approach provided a clear and reliable basis for analyzing the psychological and behavioral factors driving green product purchasing.

To ensure the clarity, reliability, and initial construct validity of the measurement items, a pilot study was conducted with a convenience sample of 30 participants prior to full-scale data collection. This sample size aligns with established recommendations that 20–30 respondents are sufficient for detecting potential weaknesses in the instrument and for preliminary validation purposes [142,143]. The sample included university students, early career professionals, and green consumers aged 18 years and above who were residing in the Jabodetabek area at the time of this study.

We distributed the questionnaire in both Bahasa Indonesia and English. To ensure both linguistic and conceptual equivalence between the original English and the translated Indonesian instruments, we employed Brislin’s [144] back-translation method. A bilingual marketing expert conducted the initial translation from English to Indonesian. Subsequently, a second bilingual translator—unaware of the original version—independently translated the Indonesian version back into English. An expert panel, comprising two specialists in green marketing and linguistics, reviewed the original and back-translated items to assess semantic and conceptual consistency. The panel reached a consensus on the equivalence of the items, noting only minor lexical variations that did not affect meaning. All discrepancies were resolved through collaborative discussion and iterative refinement, thereby ensuring strong translation validity and cultural appropriateness of the measurement instrument. Table S1 in Supplementary Materials presents further details of the procedure.

We administered the pilot survey online using a structured Google Form. Based on participants’ feedback, we made minor revisions to wording and layout, which enhanced both content validity and linguistic precision [145]. Preliminary reliability testing showed that all constructs exceeded the Cronbach’s alpha threshold of 0.70, while factor loadings were above 0.60, confirming internal consistency and convergent validity [117,125].

3.4. Data Analysis

The research team used a combination of SPSS version 25 and LISREL version 8.8 to conduct the data analysis, assess the psychometric properties of the measurement model, and evaluate the structural relationships among constructs. The analytical procedure followed a sequential approach comprising data screening, reliability and validity testing, and structural equation modeling (SEM).

First, preliminary data screening was performed in SPSS to identify incomplete responses, outliers, and inconsistencies, ensuring clean and analyzable data [146]. We also generated descriptive statistics to profile respondents based on demographic variables and purchasing behavior [147].

Next, we assessed the internal consistency reliability using Cronbach’s alpha. All constructs achieved alpha values equal to or exceeding 0.70, indicating acceptable to high reliability [120]. Construct validity was evaluated through item-total correlations and Pearson’s correlation coefficients, with values of r ≥ 0.40 interpreted as evidence of convergent validity and strong item–construct relationships [117].

We conducted Confirmatory Factor Analysis (CFA) in LISREL to validate the measurement model. Items with standardized factor loadings (SLF) of ≥ 0.50 and t-values ≥ 1.96 were retained to ensure construct validity [125]. Reliability was further confirmed using Construct Reliability (CR) and Variance Extracted (VE), with thresholds of CR ≥ 0.70 and VE ≥ 0.50 considered acceptable [148].

Finally, structural equation modeling (SEM) was conducted to test the hypothesized relationships. Model fit was assessed using absolute and incremental indices, including Chi-square (χ²), Root Mean Square Error of Approximation (RMSEA), Comparative Fit Index (CFI), and Tucker–Lewis Index (TLI). Thresholds for good model fit followed standard guidelines: RMSEA < 0.08, CFI ≥ 0.90, and TLI ≥ 0.90 [125,149].

3.5. Ethical Consideration

This study did not require formal ethical approval from an institutional review board, as it posed minimal risk to participants and was conducted independently during a period of restricted institutional access due to COVID-19 [150]. Nonetheless, it adhered strictly to academic ethical standards. Participants received a brief explanation of the study’s purpose, and informed consent was implied through the voluntary completion of an anonymous online questionnaire [151]. No personally identifiable data were collected, ensuring both confidentiality and anonymity [152].

Given the post-pandemic context, an entirely online data collection approach was employed to enhance participant safety, accessibility, and convenience. No financial or material incentives were provided, in line with ethical best practices aimed at avoiding coercion and bias [153]. The use of familiar digital platforms allowed participants to complete the survey in a secure and private setting, upholding principles of autonomy and data protection [150].

4. Results

4.1. Samples and Data Collection Procedures

We conducted this empirical study in the Greater Jakarta (Jabodetabek) area between October and November 2022. The survey was administered via Google Forms and distributed through WhatsApp within Indonesian community networks. To ensure respondent relevance, we included a filter question: “Have you ever purchased organic food products in a supermarket or hypermarket?”. While the filter question does not specify a timeframe (e.g., within the last three or six months), it effectively identified individuals with prior experience purchasing organic food, thus establishing baseline familiarity necessary for subsequent attitude and perception assessments. We also verified geographic eligibility by asking respondents to indicate their place of residence within Jabodetabek. Out of 350 individuals invited to participate, 262 completed the survey, yielding a response rate of 74.86%.

4.2. Data Screening and Data Analysis

We conducted data analysis using SPSS after screening and preparing the dataset. We excluded respondents who answered “No” to the organic food filter question or reported living outside the Jabodetabek area, as outlined in Section 4.1. We conducted outlier analysis using Mahalanobis Distance (p < 0.001), in line with prior methodological standards [154]. We found no significant outliers. After validation, the final dataset comprised 211 responses, which served as the basis for all subsequent analyses.

4.3. Demographic Profile

Table 1. Demographic characteristics of survey participants.

Variable Characteristics	Classification Standards	Frequency	Percentage
Gender	Male	145	68.70%
	Female	65	30.92%
	Others	1	0.38%
Age	18–25	10	4.96%
	26–35	67	31.68%
	36–45	87	41.22%
	46–55	35	16.41%
	55 and more	12	5.73%
Total Household Expenditure Rate per Month	≥Rp7,500,000	58	27.50%
	Rp5,000,001–Rp7,500,000	56	26.70%
	Rp3,000,000–Rp5,000,000	55	26.00%
	Rp2,000,001–Rp3,000,000	23	10.70%
	Rp1,500,001–Rp2,000,000	10	4.60%
	Rp1,000,001–Rp1,500,000	7	3.40%
	≤Rp1,000,000	2	1.10%
Education Level	Postgraduated (Master/Doctoral)	15	7.30%
	Undergraduated (Bachelor)	154	73.30%
	Secondary (High School)	35	16.40%
	Vocational and less	7	3.10%
Place of Living	Village (Rural Area)	7	3.40%
	Town (Suburban Area)	39	18.40%
	City (Urban Area)	164	78.20%

presents the demographic characteristics of the respondents relevant to their organic food purchasing behavior. A large majority (93.9%) reported purchasing organic food in supermarkets or hypermarkets. The most common shopping frequencies were bi-weekly (38.9%), monthly (30.2%), and weekly (20.2%). Males comprised 68.7% of respondents, while females accounted for 30.9%. The largest age group was 36–45 years (41.2%), followed by 26–35 years (31.7%), 46–55 years (16.4%), and those aged 55 and above (5.7%). A substantial proportion (73.3%) held at least a bachelor’s degree.

In terms of monthly household expenditure, 27.5% of respondents reported spending more than IDR 7,500,000, while 26.7% spent between IDR 3,000,000 and IDR 5,000,000. These figures, based on classifications from the Indonesian Central Statistics Agency (Badan Pusat Statistik) [155,156], reflect total household expenditures rather than food-specific spending. To help international readers interpret these values, we included a comparison with the regional minimum wage. In the Jabodetabek area, the average monthly minimum wage rose from IDR 4,517,064 (≈ USD 300) in 2022 to IDR 5,283,982 (≈ USD 350) in 2025, a 16.98% increase. Although the minimum wage does not reflect average income, it serves as a useful benchmark given the lack of consistent national salary data. Since urban households often have multiple earners, these expenditure levels align with middle-income consumption patterns typical of respondents who shop at supermarkets and complete online surveys [157]. Most respondents lived in urban areas (78.2%), while 18.3% lived in suburban areas and 3.4% in rural regions. See Table 1 for detailed demographic characteristics.

4.4. Normality, Collinearity, Homogeneity, and Reliability

• Normality: The Kolmogorov–Smirnov and Shapiro–Wilk tests showed non-normality (p < 0.05), confirming the need for non-parametric or robust analytical methods.
• Collinearity: Spearman’s correlation and Variance Inflation Factor (VIF) results showed no multicollinearity; all values were below the 0.85 threshold.
• Homogeneity: Levene’s test indicated unequal variances in expenditure level (GPI, p = 0.025) and residence (GMM, p = 0.014), suggesting socio-economic influence on these variables.
• Reliability: Cronbach’s alpha values exceeded 0.70 for all constructs (ENK, GTR, GMM, GPI, GPB), confirming acceptable internal consistency [158].

4.5. Descriptive Statistical Analysis

Table 2. Descriptive statistics and reliability coefficients of key constructs (N = 211).

Construct	Mean	SD	Min	Max	Cronbach’s α
Environmental Knowledge (ENK)	3.96	0.69	2.00	5.00	0.874
Green Trust (GTR)	3.99	0.61	1.80	5.00	0.879
Green Marketing Mix (GMM)	4.04	0.55	2.58	5.00	0.939
Green Product	4.03	0.62	2.00	5.00	–
Green Price	4.11	0.59	2.75	5.00	–
Green Place	3.99	0.60	2.50	5.00	–
Green Promotion	4.03	0.61	2.00	5.00	–
Green Purchase Intention (GPI)	4.01	0.60	1.80	5.00	0.869
Green Purchase Behavior (GPB)	3.97	0.68	1.50	5.00	0.901

Note: Composite mean and Standard Deviation (SD) values are presented for each construct. Cronbach’s α = internal consistency reliability.

presents the descriptive statistics for the five main constructs in the research model. We conducted the analysis using SPSS based on responses from 211 participants.

The green marketing mix (M = 4.04, SD = 0.55) recorded the highest overall mean among all constructs, suggesting that consumers placed the greatest emphasis on marketing-related elements such as green pricing and product attributes when making environmentally conscious decisions. The green price dimension, in particular, had the highest mean (M = 4.11), indicating that the affordability of eco-friendly products significantly influences consumer behavior.

Green purchase intention (M = 4.01, SD = 0.60) and green trust (M = 3.99, SD = 0.61) followed closely, highlighting that consumers are willing to engage in sustainable purchasing when they trust green product claims and perceive long-term environmental benefits. Environmental knowledge (M = 3.96, SD = 0.69) also scored relatively high, indicating that awareness and understanding of environmental issues remain important drivers of behavior. Similarly, green purchase behavior (M = 3.97, SD = 0.68) suggests moderate to high engagement in actual green consumption practices.

4.6. Structural Equation Modeling (SEM) Analysis

We employed LISREL version 8.80 to conduct SEM analysis examining the interrelations among five latent variables: ENK, GTR, GMM, GPI, and GPB.

4.6.1. Measurement Model Analysis

Table 3. Summary of validity and reliability of latent variables.

Latent Variable	Dimension	SLF *	Error	Construct Reliability	Variable Extracted	Conclusion
				>0.70	>0.5
ENK	ENK			0.75	0.50	Reliable
	ENK1	0.76	0.42			Valid
	ENK2	0.72	0.48			Valid
	ENK3	0.64	0.59			Valid
GTR	GTR			0.83	0.50	Reliable
	GT1	0.81	0.34			Valid
	GT2	0.74	0.45			Valid
	GT3	0.50	0.78			Valid
	GT4	0.71	0.51			Valid
	GT5	0.74	0.46			Valid
GMM	GMM			1.00	0.99	Reliable
	GMM.GPO	0.92	0.14			Valid
	GMM.GPR	1.00	−0.0059			Valid
	GMM.GPL	1.03	−0.66			Valid
	GMM.GPM	0.94	0.11			Valid
GPI	GPI			0.80	0.45	Reliable
	GPI1	0.65	0.58			Valid
	GPI2	0.70	0.51			Valid
	GPI3	0.65	0.58			Valid
	GPI4	0.62	0.62			Valid
	GPI5	0.72	0.49			Valid
GPB	GPB			0.85	0.59	Reliable
	GPB1	0.77	0.41			Valid
	GPB2	0.73	0.47			Valid
	GPB3	0.68	0.54			Valid
	GPB4	0.87	0.24			Valid

Note: SLF: Standardized Factor Loadings, showing item–construct relationships. * Values obtained from confirmatory factor analysis (CFA).

summarizes the validity and reliability of all observed variables in the measurement model, showing that all dimensions have good validity and latent variables have strong reliability.

The measurement model demonstrated a good overall fit based on multiple indices. Most observed variables met the minimum validity and reliability thresholds: standardized factor loadings (SLF) ≥ 0.50, t-values ≥ 1.96 [125], Construct Reliability (CR) ≥ 0.70, and Variance Extracted (VE) ≥ 0.50 [148]. Although some indicators, such as GT3 and GPB4, had marginal loadings or error variances, they were retained due to their overall contribution and acceptable statistical thresholds.

Notably, the green marketing mix (GMM) construct showed exceptionally strong reliability (CR = 1.00; VE = 0.99). However, the indicator GMM.GPL presented a Heywood case (loading >1, negative error variance), which suggests a statistical anomaly possibly caused by high multicollinearity or sample size limitations. While the model fit remained robust, this issue indicates the need for further assessment in future studies.

4.6.2. Overall Measurement Model

The overall measurement model in this research uses latent variable scores (LVSs) for each indicator, with standardized load factors (SLF) below 1.0 and acceptable variance errors [159].

Table 4. Overall measurement model goodness of fit indices.

GOFI	Criteria	Value	Result	GOFI	Criteria	Value	Result
NCS	≤2	1.05401	Good Fit	NFI	≥0.90	0.99	Good Fit
p-value	>0.05	0.30	Good Fit	NNFI	≥0.90	1.00	Perfect Fit
RMSEA	≤0.08	0.016	Good Fit	CFI	≥0.90	1.00	Perfect Fit
RMR	≤0.05	0.040	Good Fit	IFI	≥0.90	1.00	Perfect Fit
GFI	≥0.90	0.90	Good Fit	RFI	≥0.90	0.98	Good fit

shows the overall model achieved excellent fit based on multiple indices: NNFI, CFI, and IFI values were all at or above 1.00; RMSEA was low (0.016); and NCS was well below the cutoff (1.05). The only marginal indicator was GFI (0.90), which still met acceptable fit criteria.

All latent variables demonstrate good validity and reliability (

Table 5. Summary of overall measurement model validity and reliability.

Latent Variable	Dimension	SLF *	Error	Construct Reliability	Variable Extracted	Conclusion
				>0.70	>0.5
ENK	ENK			0.76	0.51	Reliable
	ENK1	0.70	0.51			Valid
	ENK2	0.75	0.44			Valid
	ENK3	0.70	0.51			Valid
GTR	GTR			0.82	0.49	Reliable
	GT1	0.77	0.41			Valid
	GT2	0.75	0.44			Valid
	GT3	0.53	0.72			Valid
	GT4	0.72	0.48			Valid
	GT5	0.69	0.52			Valid
GMM	GMM			0.90	0.74	Reliable
	GMM.GPO	0.80	0.35			Valid
	GMM.GPR	0.83	0.32			Valid
	GMM.GPL	0.90	0.19			Valid
	GMM.GPM	0.85	0.27			Valid
GPI	GPI			0.80	0.45	Reliable
	GPI1	0.62	0.61			Valid
	GPI2	0.66	0.56			Valid
	GPI3	0.76	0.42			Valid
	GPI4	0.64	0.59			Valid
	GPI5	0.64	0.58			Valid
GPB	GPB			0.84	0.56	Reliable
	GPB1	0.74	0.45			Valid
	GPB2	0.74	0.45			Valid
	GPB3	0.73	0.47			Valid
	GPB4	0.79	0.37			Valid

Note: SLF: Standardized Factor Loadings, showing item–construct relationships. * Values obtained from confirmatory factor analysis (CFA).

). While the GPI variable has a VE of 0.45, slightly below the recommended threshold of 0.50 [125], its CR of 0.80 exceeds the minimum requirement of 0.70, indicating acceptable reliability. The highest SLF was found for GMM.GPL (0.90), reinforcing the robustness of the GMM construct. Overall, the model demonstrated a sound measurement structure suitable for SEM analysis.

Table 5 also shows the measurement model path, summarizing all survey variables (ENK, GTR, GMM, GPI, GPB). ENK2 (SLF = 0.75) contributes most to ENK. GT1 (SLF = 0.77) has the greatest effect on GTR, followed by GT2 (SLF = 0.75) and others. GMM.GPL (SLF = 0.90) contributes most to GMM, followed by GMM.GPM (SLF = 0.85) and others. For GPI, GPI3 (SLF = 0.76) has the highest contribution, followed by GPI2 (SLF = 0.66) and others. GPB is most influenced by GPB4 (SLF = 0.79), followed by GPB1 (SLF = 0.74). All dimensions provide effective metrics for hypothesis testing in the research design.

4.6.3. Structural Equation Model (SEM) Analysis

Table 6. SEM goodness of fit indices (final model).

GOFI	Criteria	Value	Result	GOFI	Criteria	Value	Result
NCS	≤2	0.93662	Good Fit	NFI	≥0.90	0.99	Good Fit
p-value	>0.05	0.70	Good Fit	NNFI	≥0.90	1.00	Perfect Fit
RMSEA	≤0.08	0.000	Perfect Fit	CFI	≥0.90	1.00	Perfect Fit
SRMR	≤0.05	0.0140	Good Fit	IFI	≥0.90	1.00	Perfect Fit
GFI	≥0.90	0.88	Marginal Fit	RFI	≥0.90	0.99	Good Fit

shows that all goodness of fit indices for the final structural equation model perform well, with the exception of GFI (0.88), which is considered a marginal fit. Overall, the SEM fit is very good, with RMSEA, NNFI, CFI, and IFI all showing suitable values.

Regarding validity and reliability,

Table 7. Summary of SEM validity and reliability (final model).

Latent Variable	Dimension	SLF *	Error	Construct Reliability	Variable Extracted	Conclusion
				>0.70	>0.5
ENK	ENK			0.76	0.51	Reliable
	ENK1	0.7	0.5			Valid
	ENK2	0.75	0.44			Valid
	ENK3	0.7	0.52			Valid
GTR	GTR			0.82	0.50	Reliable
	GT1	0.76	0.42			Valid
	GT2	0.75	0.44			Valid
	GT3	0.53	0.72			Valid
	GT4	0.72	0.48			Valid
	GT5	0.69	0.52			Valid
GMM	GMM			0.90	0.74	Reliable
	GMM.GPO	0.8	0.35			Valid
	GMM.GPR	0.83	0.32			Valid
	GMM.GPL	0.9	0.19			Valid
	GMM.GPM	0.85	0.27			Valid
GPI	GPI			0.80	0.44	Reliable
	GPI1	0.62	0.62			Valid
	GPI2	0.66	0.57			Valid
	GPI3	0.76	0.42			Valid
	GPI4	0.64	0.59			Valid
	GPI5	0.64	0.59			Valid
GPB	GPB			0.84	0.57	Reliable
	GPB1	0.74	0.45			Valid
	GPB2	0.74	0.45			Valid
	GPB3	0.73	0.4			Valid
	GPB4	0.79	0.37			Valid

Note: SLF: Standardized Factor Loadings, showing item–construct relationships. * Values obtained from confirmatory factor analysis (CFA).

shows that all variables (ENK, GTR, GMM, GPI, GPB) have good validity and reliability results. Therefore, our research model has high reliability for the analysis of research hypotheses.

Path diagram analysis revealed that ENK influences GPB indirectly through GTR and GPI. ENK significantly predicted GTR (β = 0.90, t = 10.77) and GMM (β = 0.73, t = 3.46) and that GTR significantly influenced GPI (β = 0.70, t = 2.04). GPI also significantly predicted GPB (β = 1.02, t = 2.21). On the other hand, paths from ENK to GPI, and from GMM to GPB, were not significant. Figures S1 and S2 in Supplementary Materials provide the visual representation of these path diagrams, showing standardized solutions and t-values.

4.7. Summary of Hypotheses

Table 8. Hypotheses test summary (final research model).

Hypotheses	Path	β	t-value	S.E	Margin of Error	CI (95%)	Conclusion
H1	ENK → GTR	0.90	10.77	0.0835	0.163	0.737, 1.063	Accepted
H2	ENK → GMM	0.73	3.46	0.2107	0.412	0.318, 1.142	Accepted
H3	ENK → GPI	−0.05	−0.13	0.3846	0.756	−0.806, 0.706	Rejected
H4	GTR → GMM	0.27	1.27	0.2126	0.416	−0.146, 0.686	Rejected
H5	GTR → GPI	0.70	2.04	0.3431	0.673	0.027, 1.373	Accepted
H6	GMM → GPI	0.34	0.63	0.5397	1.057	−0.718, 1.398	Rejected
H7	GMM → GPB	−1.45	−0.93	1.5591	3.053	−4.505, 1.605	Rejected
H8	GPI → GPB	1.02	2.21	0.4611	0.904	0.116, 1.924	Accepted
H9	ENK → GPB	1.50	1.08	1.3889	2.722	−1.222, 4.222	Rejected

Note: β: Standard Coefficient; S.E: Standard Error; CI: Confidence Interval 95% (Lower, Upper).

presents the results of the structural path analysis conducted using structural equation modeling (SEM), assessing the hypothesized relationships among latent constructs. Each path includes a standardized coefficient (β), t-value (equivalent to the critical ratio, C.R.), standard error (S.E.), and a 95% confidence interval to determine the strength, direction, and statistical significance of the relationships.

The standardized coefficient (β) reflects the magnitude and direction of influence between constructs. Positive values indicate direct relationships, while negative values suggest inverse effects. A t-value greater than ±1.96 signals statistical significance at the 95% confidence level [125]. Confidence intervals further support interpretation by showing the plausible range of the effect; intervals that exclude zero confirm the significance of the relationship. Smaller standard errors indicate more precise estimates.

For example, the path from environmental knowledge (ENK) to green trust (GTR) (H1) produced a strong, significant effect (β = 0.90, t = 10.77, 95% CI (0.737, 1.063)). Conversely, the path from ENK to green purchase intention (GPI) (H3) was not significant (β = −0.05, t = −0.13, CI (−0.806, 0.706)), as the confidence interval included zero and the t-value fell below the threshold.

Among the nine hypothesized relationships, four paths were statistically supported: H1 (ENK → GTR), H2 (ENK → GMM), H5 (GTR → GPI), and H8 (GPI → GPB). Each showed significant t-values and confidence intervals that did not contain zero. In contrast, H3, H4, H6, H7, and H9 did not reach statistical significance, primarily due to low t-values and wide confidence intervals.

These results highlight the uneven influence of latent constructs within the model. Specifically, environmental knowledge and green purchase intention emerged as key predictors of sustainable consumer behavior, aligning with the theoretical framework of this study.

Moreover, the research analysis of the total standard coefficients of latent variable pathways, as shown in

Table 9. Standard coefficient analysis of research model.

No	Effects	Pathways		β Calculation		β Total
		DE *	IE **	DE *	IE **
1	ENK → GMM	β2	n. a	0.73	n. a	0.73
2	ENK → GTR → GPI	β3	β1* β5	−0.05	0.63	0.58
3	ENK → GTR → GPI → GPB	β9	β1* β5* β8	1.50	0.6426	2.1426

Note: β = Standard coefficient; (*) DE = Direct Effect; (**) IE = Indirect Effect; n. a. = Not applicable, as no indirect effect exists for this pathway.

, reveals that the pathway from ENK to GPB has the highest total coefficient (β = 2.1426), followed by ENK to GPI (β = 0.58). The ENK to GPB relationship includes two paths, β₉ (1.50) and β₁–β₅–β₈ (0.6426), resulting in a total coefficient of β = 2.1426. The ENK to GPI relationship involves two paths, β₃ (−0.05) and β₁–β₅ (0.63), yielding a total coefficient of β = 0.58. This analysis highlights the presence of indirect influences between variables, in addition to direct relationships.

Figure 2 below illustrates the relationships between the key constructs in the research model, highlighting the structural coefficients (β)* and t-values for each pathway. This visualization helps to clearly depict the strength and direction of the interactions between the constructs in the model.

5. Discussion

This study advances the understanding of green purchasing behavior within the context of Greater Jakarta by examining the interrelationships among environmental knowledge (ENK), green trust (GTR), green marketing mix (GMM), green purchase intention (GPI), and green purchase behavior (GPB). The findings offer both empirical validation and theoretical contributions, particularly relevant to emerging economies where sustainable consumption patterns are still evolving.

5.1. Demograpic Profile

The data indicate that 93.89% of respondents have purchased organic food, reflecting a strong level of consumer engagement with sustainable products. This aligns with previous studies emphasizing the rising consumer interest in environmentally responsible consumption [160,161,162]. This behavioral trend is particularly evident among individuals aged 36–45 and those with higher education levels; 73.3% hold at least a bachelor’s degree. This suggests that higher education and financial stability correlate with green purchasing behavior, reinforcing prior research linking socio-economic factors to sustainable consumption [163,164,165]. Moreover, 27.5% of participants report monthly expenditures exceeding IDR 7,500,000, emphasizing the role of financial capacity in facilitating green purchases [166,167]. These demographic insights align with established theories positing education and income as enablers of pro-environmental behavior. Higher educational attainment tends to foster environmental awareness and critical evaluation of green claims, while financial stability reduces the economic barriers often associated with premium-priced sustainable products.

5.2. Accepted Hypotheses

This study contributes to the green consumer behavior literature by empirically validating the mediating role of green trust (GTR) in converting environmental knowledge (ENK) into green purchase intention (GPI) and ultimately into green purchase behavior (GPB). The strong relationship between ENK and GTR (β = 0.90, t = 10.77) suggests that environmentally literate consumers are better equipped to assess green claims, reducing susceptibility to greenwashing and enhancing brand credibility [168,169,170,171].

By confirming the sequential pathway ENK → GTR → GPI → GPB, the findings provide empirical support for an extended Theory of Planned Behavior (TPB). Specifically, green trust enhances ethical alignment and perceived brand value, thereby increasing purchase intention and behavioral execution [50,172,173,174]. The direct effect of GPI on GPB (β = 1.02, t = 2.21) reinforces TPB’s emphasis on perceived behavioral control, particularly when conditions such as fair pricing and product accessibility are favorable [161,175].

Additionally, the dual effect of ENK on GPB (β = 2.14) through both direct and mediated pathways highlights the strategic importance of environmental education as a long-term driver of sustainable behavior [176,177,178,179]. These results extend existing research by illustrating that affective mechanisms, such as trust, are essential for transforming knowledge into action in emerging markets.

5.3. Rejected Hypotheses

5.3.1. Environmental Knowledge Requires Mediation

The non-significant direct effects of ENK on both GPI and GPB (β = −0.05, t = −0.13; β = 1.50, t = 1.08) confirm the Knowledge–Intention–Behavior (KIB) gap framework [180]. Although knowledge enhances awareness, it does not independently generate intention or action. Barriers such as price sensitivity, skepticism, and perceived inefficacy limit its direct impact [181,182,183,184,185,186,187]. This highlights the necessity of mediating constructs like trust and perceived behavioral control in influencing green behavior.

5.3.2. Trust Does Not Shape Green Marketing Strategy

The rejection of the GTR → GMM path (β = 0.27, t = 1.27) indicates that consumer trust does not directly inform marketing mix strategies. Although trust enhances loyalty, marketing strategies may still be driven by business objectives such as profitability, compliance, and feasibility rather than consumer trust [188,189]. This may also reflect methodological limitations in capturing strategic responsiveness to consumer sentiment [190,191].

5.3.3. Green Marketing Mix Alone Is Insufficient

Contrary to expectations, GMM did not significantly influence either GPI (β = 0.34, t = 0.63) or GPB (β = −1.45, t = −0.93), challenging assumptions about the efficacy of green marketing stimuli. This aligns with recent critiques of low-impact green advertising, especially in emerging markets where consumers often prioritize price and convenience over environmental values [61,192,193,194,195]. Weak targeting, inconsistent messaging, and greenwashing may further erode the persuasive power of green promotions. These findings reinforce the view that marketing effectiveness is contingent on trust, values, and perceived behavioral control, rather than on message content alone [99,101,196].

5.4. Research Contributions

5.4.1. Theoretical Contributions

This study advances the theoretical understanding of sustainable consumer behavior by extending the Theory of Planned Behavior (TPB) with two pivotal constructs: green trust (GTR) and the green marketing mix (GMM). Empirical findings confirm that environmental knowledge (ENK) indirectly influences green purchase behavior (GPB) through GTR, reinforcing the cognitive–attitudinal pathway of TPB [49,63]. This positions trust as a key psychological mechanism that translates knowledge into intention, a contribution particularly relevant in emerging, high-context, and price-sensitive markets [49,197].

By identifying GTR as a significant mediator, this study enriches the attitudinal dimension of TPB, demonstrating how the perceived credibility of environmental claims directly shapes pro-environmental intentions [50,51]. The analysis also highlights green price (GPR) as the most influential GMM dimension, underlining how economic constraints influence behavioral control, especially in urban markets like Greater Jakarta, where affordability is a central concern [198,199]. This finding suggests a refinement of TPB by integrating financial limitations into the perceived behavioral control construct.

Furthermore, the limited direct effect of GMM on behavior suggests that marketing strategies alone are insufficient without supportive contextual enablers, such as trust and socio-cultural norms. This aligns with broader calls in behavioral research to include external, context-specific moderators [61,200], offering a more ecologically valid and comprehensive behavioral model for sustainable consumption.

5.4.2. Managerial Contributions

The findings offer actionable guidance for marketers, businesses, and policymakers operating in emerging economies, particularly in metropolitan areas like Greater Jakarta, where environmental concern is rising but purchasing behavior remains constrained. Key strategic implications include the following:

• Trust building is essential: Firms should increase transparency in sustainability claims through the use of third-party certifications, credible eco-labels, and clear communication of environmental benefits to reinforce green trust and reduce skepticism.
• Consumer education: Targeted awareness campaigns can enhance environmental knowledge, thereby indirectly boosting purchase intention through heightened trust.
• Affordability as a critical enabler: Given that price is the most influential element of the green marketing mix, firms and governments must address the trust–affordability gap through the following:
• Tax incentives or subsidies for sustainable products [201].
• Value-based pricing strategies that emphasize long-term savings.
• Loyalty programs and cost reduction innovations such as eco-friendly packaging or localized production.
• Targeted marketing strategies: Promotions should integrate TPB elements, including subjective norms and perceived behavioral control, to reduce the intention–behavior gap. Strategies might include the following:
• Leveraging social influence via green endorsements and peer recommendations [202].
• Designing green products that fit seamlessly into consumers’ existing routines and lifestyle preferences.
• Enable supportive ecosystems: Policymakers should complement firm-level efforts by creating enabling environments through public awareness campaigns, CSR-linked tax reliefs and infrastructure investments that facilitate sustainable consumption (e.g., green logistics, accessible eco-products).

In summary, this research contributes a theoretically grounded and context-sensitive model of green consumer behavior, applicable to urban emerging markets. By addressing the trust–affordability–behavior nexus and incorporating cognitive, attitudinal, and contextual dimensions, this study provides a robust foundation for both advancing TPB and informing effective green marketing and policy strategies.

5.5. Research Limitations

Although this study applies a rigorous methodological framework and robust SEM analysis, several limitations warrant consideration.

First, the geographic scope is confined to Greater Jakarta, an urban region characterized by relatively high environmental awareness and better access to green products [203]. Therefore, the findings may not generalize to rural or underserved regions in Indonesia, where consumer behavior and infrastructure differ significantly.

Second, the cross-sectional design captures consumer attitudes and intentions at a single time point, limiting the ability to assess temporal changes in behavior. Longitudinal research could better illuminate evolving consumer responses to environmental policies and socio-economic developments [204].

Third, reliance on online surveys introduces potential sampling and self-reporting biases. Limited digital access and literacy among certain socio-economic groups may exclude segments of the population, while social desirability bias may compromise data accuracy [205]. Future studies should consider mixed-methods approaches to improve inclusivity and data validity.

Fourth, data collection occurred between March and November 2022, with the primary online survey administered between October and November 2022. This period coincided with the ongoing enforcement of the Community Activities Restrictions Enforcement (PPKM). Indonesia’s COVID-19 restrictions (PPKM) during the data collection period constrained in-person fieldwork. Travel limitations and health protocols restricted the use of qualitative methods such as focus groups and interviews [137]. Future research should incorporate hybrid or on-site methodologies to gain richer insights and triangulate quantitative findings.

Finally, this study focuses on green purchase intentions rather than observed behavior. Although intentions are widely used proxies, a persistent attitude–behavior gap exists [99,200]. As shown by Paul et al. [206] and Chen and Tung [207], situational and contextual barriers often prevent the translation of pro-environmental intentions into action. Future research should apply behavioral tracking or experimental designs to assess alignment between stated intentions and actual behavior.

5.6. Suggestions for Future Research

Building on this study’s findings and limitations, several future research directions merit attention:

• Broaden Geographic and Cultural Scope: Future studies should investigate green consumer behavior across diverse Indonesian regions, particularly rural areas, to capture socio-economic and cultural variability. Cross-national comparisons between emerging and developed economies may further contextualize behavioral determinants [203].
• Examine Mediators and Moderators: The indirect relationship between environmental knowledge (ENK) and green purchase behavior (GPB) calls for the inclusion of mediators such as environmental concern or perceived consumer effectiveness. Moderators like income, education, and environmental literacy may clarify boundary conditions for behavioral activation [204,205].
• Apply Longitudinal and Experimental Designs: To observe behavioral change over time, researchers should adopt longitudinal methods. Experimental approaches, such as A/B testing or simulated shopping scenarios, can establish causal links between interventions (e.g., pricing, labeling) and outcomes [202,203].
• Deconstruct the Green Marketing Mix (GMM): Given GMM’s non-significant direct effects, future research should analyze individual elements (product, price, place, promotion) to determine their relative influence across consumer segments and contexts.
• Explore Dimensions of Trust: As green trust (GTR) significantly mediates green behavior, future studies should distinguish between institutional, brand, and product trust. Understanding how trust is built, through certifications, CSR, or peer reviews—can inform both practice and policy [204].
• Evaluate Policy and Incentive Mechanisms: Considering price sensitivity, researchers should assess how subsidies, tax breaks, and financial nudges influence green product adoption, particularly among lower-income groups [134,137].
• Address the Knowledge–Intention–Behavior (KIB) Gap: The absence of a direct link between ENK and intention/behavior underscores the need to examine psychological (e.g., self-efficacy, habit) and contextual (e.g., availability, social influence) barriers [99,206,207].
• Incorporate Identity and Lifestyle Factors: Further investigation into eco-identity, values congruence, and lifestyle fit may uncover more nuanced predictors of green consumption. Segmenting consumers by eco-lifestyle profiles could improve targeting effectiveness.

Pursuing these directions will strengthen theoretical models and support evidence-based strategies for promoting sustainable consumption in emerging markets.

6. Conclusions

This study contributes to the expanding literature on sustainable consumption by examining the complex interplay among environmental knowledge, green marketing strategies, and green trust in shaping green purchase intentions and behaviors within the urban context of Greater Jakarta, Indonesia. Employing a structural equation modeling (SEM) approach, the findings reveal that green trust plays a pivotal mediating role, translating environmental knowledge into purchase intention. Although environmental knowledge does not directly influence green purchase behavior, its indirect effect, via trust, highlights the centrality of credibility and consumer confidence in promoting sustainable consumption [168,169].

By addressing the underexplored role of green trust in emerging Southeast Asian markets, this study fills a critical gap in the sustainable consumption literature, which has predominantly focused on Western or high-income contexts.

This research extends the Theory of Planned Behavior [63] by integrating green trust [46,50] and the green marketing mix [150] as additional constructs. It demonstrates that environmental knowledge indirectly influences green purchasing behavior through the attitudinal pathway of trust, enhancing TPB’s explanatory power. Additionally, the findings emphasize the influence of contextual and economic variables, particularly pricing sensitivity, suggesting a need to adapt TPB to better reflect the socio-economic realities of emerging urban markets [101, 199].

These findings suggest that an expanded TPB model, incorporating trust and pricing sensitivity, could offer a theoretically generalizable framework for future research in diverse emerging market contexts.

Notably, the green marketing mix exerted no direct effect on either green purchase intention or behavior. This implies that traditional marketing strategies may not be sufficient to trigger eco-conscious consumption in developing economies. These findings align with previous research, indicating that price sensitivity, economic uncertainty, and cultural norms may overshadow pro-environmental messaging [175,208]. These insights emphasize that psychological and contextual variables, such as perceived benefits, social influence, and economic constraints, must be factored into more effective green marketing strategies [209,210].

Despite its contributions, this study has several limitations. First, this study’s focus on Greater Jakarta, a densely populated and economically developed urban region, limits the generalizability of its findings to rural or culturally distinct areas of Indonesia [203]. Second, the cross-sectional design constrains the ability to observe changes in behavior over time, particularly in response to evolving environmental policies or economic conditions [204]. Third, reliance on online surveys introduces potential sampling bias, especially in a country with unequal digital access [137]. Lastly, this study measures behavioral intention rather than observed behavior, leaving room for discrepancy due to the well-known attitude–behavior gap [99,200].

To address these limitations, future research should adopt longitudinal and experimental designs to capture behavioral evolution over time and test causal mechanisms. Longitudinal studies could examine how sustained exposure to green marketing campaigns, eco-labels, or policy incentives cultivates trust and gradually shifts consumption patterns, while experimental designs could evaluate the impact of targeted interventions such as subsidies, CSR communications, or labeling schemes.

Expanding the geographic and cultural scope and applying mixed-methods approaches would enhance external validity and triangulate insights across contexts. Researchers are also encouraged to investigate additional mediating and moderating variables, such as perceived consumer effectiveness, identity alignment, lifestyle compatibility, and trust typologies, to better understand how green purchase intentions are transformed into action. Furthermore, disaggregating the green marketing mix and assessing the effectiveness of economic and policy instruments tailored to price-sensitive consumers would refine communication strategies and offer practical guidance for policy and marketing implementation.

Beyond its theoretical contributions, this study offers actionable insights for advancing planetary health, defined as the interdependent well-being of people and the planet. By showing how environmental knowledge and trust jointly shape behavior, the findings emphasize the collective impact of individual choices on ecological and public health outcomes [42,43,46]. This aligns with global systems-based frameworks that advocate for integrated behavioral, ecological, and policy-based solutions to mitigate climate and consumption challenges [42,44].

In practical terms, this study provides actionable implications for policymakers and marketers. Strategies that build consumer trust while addressing economic constraints are vital to bridging the knowledge–intention–behavior gap.

Governments and NGOs could prioritize trust-building mechanisms such as third-party certifications, transparent labeling, and targeted subsidies to boost green product adoption, especially among price-sensitive consumers.

As the urgency of climate action intensifies, empowering consumers in emerging economies with credible knowledge and trust-based interventions offers a pragmatic and context-sensitive pathway to accelerating sustainable consumption across the Global South.