2.1. Integrasi Theory of Reasoned Action (TRA) and Status Quo Bias (SQB)

The TRA, a foundational framework in behavioral science, posits that behavioral intention arises from an individual’s attitude toward a behavior, shaped by underlying beliefs and evaluations of its potential outcomes (Fishbein & Ajzen, 1975). Within sustainable consumption research, pro-environmental attitude consistently emerges as a strong predictor of the ITS, particularly among younger consumers (Madinga et al., 2024; Rūtelionė & Bhutto, 2024). In this study, attitude is conceptualized as a central mediating construct.

However, the explanatory power of TRA—and its successor, the TPB—has been the subject of extensive scholarly debate. Both rely on the assumption of rational action, which assumes that individuals logically evaluate the consequences of their behavior. This rationalist view often fails to account for non-rational factors, such as emotions, ingrained habits, and complex situational contexts, which significantly influence real-world consumer decision-making, particularly in ethical and sustainable consumption domains.

To address these limitations, recent behavioral research increasingly integrates classical models with constructs from behavioral economics and social psychology to enhance predictive accuracy and ecological validity. Aligned with this trend, the present study integrates TRA with SQB. SQB provides a complementary theoretical lens that directly addresses the “rationality gap” in TRA by explaining why individuals often maintain existing consumption patterns due to cognitive comfort and perceived risk, resulting in inertia—a latent reluctance to change (Samuelson & Zeckhauser, 1988). Incorporating SQB enables the model to account for psychological friction that hinders intention formation, even when pro-environmental attitudes are strong. Within this framework, perceived inertia serves as a moderating factor, weakening the relationship between pro-environmental attitude and ITS (Kim & Kankanhalli, 2009).

Two antecedent constructs are introduced to broaden the model’s explanatory scope: social media influence and perceived product durability. Both align with TRA’s belief-based evaluative structure, functioning as external referents that shape individuals’ judgments toward sustainable fashion (Fishbein & Ajzen, 2011). Social media has become particularly relevant among Gen Z and Millennials, as platforms such as Instagram, TikTok, and YouTube not only disseminate information but also cultivate lifestyle identities, reinforce symbolic values, and normalize sustainability-oriented behaviors (Lin et al., 2023; Kim et al., 2024). Empirical evidence shows that repeated exposure to sustainability-related content fosters value internalization, making social media influence a culturally resonant predictor of pro-environmental attitude (Andika et al., 2026; Nabivi, 2025).

Similarly, perceived product durability represents a tangible attribute aligned with younger consumers’ expectations for quality and longevity. In this sense, sustainability encompasses ethical and functional dimensions (Madinga et al., 2024). Durable products are often viewed as efficient, economical, and environmentally responsible, enhancing long-term product valuation and reinforcing ideals of sustainable consumption (Aakko & Niinimäki, 2022; Ngo et al., 2024). Thus, durability transcends its technical definition to become a meaningful evaluative criterion in fashion choice.

Figure 1Research model 

ITS serves as the primary outcome variable. In behavioral research, intention is widely recognized as a proximal indicator of future behavior, particularly when actual behavior remains latent or constrained by contextual factors (Hino, 2017; Vermeir & Verbeke, 2006). In emerging markets such as Indonesia—where behavioral transitions tend to be gradual—intention is a more sensitive measure of psychological readiness.

By integrating TRA’s motivational mechanisms with the psychological constraints articulated in SQB, this study develops a conceptual model that captures both the aspirational and inertial dimensions of consumer decision-making. This integrated framework offers a more nuanced lens for examining the dynamics of sustainable fashion adoption among young consumers in emerging economies. Figure 1 presents an overview of the conceptual model.

2.2. Social Media Influence and Pro-environmental Attitudes

The inclination of younger consumers to engage with sustainable fashion is frequently underpinned by pro-environmental attitudes, which are pivotal in shaping ethical consumption trajectories. Within the framework of TRA, attitudes represent the evaluative consequence of behavioral beliefs formed through direct experience and mediated information (Fishbein & Ajzen, 1975). In this context, social media functions not merely as an informational medium but as an affective and normative space where sustainability narratives are circulated and internalized. Platforms such as Instagram, TikTok, and YouTube serve as socio-digital arenas in which sustainability ideals are increasingly curated, consumed, and performed (Truc, 2024). Content related to sustainable fashion often evokes moral resonance and emotional salience, reinforcing values associated with ecological responsibility (Schiaroli et al., 2024). The algorithmic structuring of digital exposure further personalizes these encounters, subtly intensifying the cognitive and affective dimensions of environmental belief systems (Eg et al., 2023).

Within the logic of TRA, such beliefs are understood as informational appraisals of behavioral outcomes, processed through internal value systems and expressed through attitudinal shifts. A growing body of evidence indicates that frequent exposure to sustainability discourse on social media is associated with heightened eco-consciousness and attitudinal alignment (Meng et al., 2023). Moreover, the symbolic authority of influencers contributes to the normalization of sustainability as a social value, amplifying its emotional resonance and perceived desirability (Singh, 2024). Against this conceptual and empirical backdrop, we propose the following hypothesis:

2.3. Perceived Durability and Pro-environmental Attitudes

Perceived durability refers to the belief that a product retains its functionality and aesthetic appeal over time (De Giovanni, 2024; Garvin, 1987). Beyond its technical connotation, durability carries symbolic weight as a representation of product integrity, economic efficiency, and environmental responsibility (Aakko & Niinimäki, 2022; Cooper et al., 2013). This attribute often intersects with broader ethical concerns among Gen Z and millennial consumers, reflecting an emerging sensibility that prioritizes long-term utility and reduced waste (Calculli et al., 2021). Within the architecture of TRA, perceived durability is positioned as a behavioral belief—an evaluative filter through which consumers appraise the ecological merit of fashion products (Fishbein & Ajzen, 1975). The perception of durability may also evoke associations with conscious purchasing and personal responsibility, reinforcing pro-environmental value orientations.

Empirical studies substantiate this link. Agost and Vergara (2020) suggest that durability perceptions correlate with behavioral tendencies to preserve and care for garments, signaling deeper attitudinal commitments to sustainability. Similar patterns are evident in recent work by Madinga et al. (2024) and Ngo et al. (2024), who report that favorable durability perceptions consistently align with pro-environmental dispositions, particularly among younger consumers attuned to sustainability cues. These insights support the inclusion of perceived durability as a key cognitive determinant within the proposed framework. The following hypothesis is thus formulated:

2.4. Pro-Environmental Attitude and Intention to Switch

Within TRA, attitudes toward a given behavior are considered proximal indicators of one’s intention to engage in that behavior (Fishbein & Ajzen, 1975). In sustainability research, pro-environmental attitudes reflect evaluative orientations grounded in ecological awareness, ethical considerations, and value coherence. These attitudes are not merely expressions of preference but also indicators of cognitive readiness to depart from unsustainable consumption patterns.

In this study, ITS refers to a purposive inclination to abandon conventional fashion in favor of environmentally responsible alternatives. Unlike adoption intention, which denotes initial engagement, ITS implies an evaluative break—a realignment of consumption identity anchored in internalized moral reasoning (Maslikhah et al., 2024).

Evidence from Rausch and Kopplin (2021) and Rūtelionė and Bhutto (2024) underscores a consistent relationship between pro-environmental attitudes and intention to adopt sustainable fashion, particularly among value-driven youth segments. Further, studies by Ngo et al. (2024), Mao et al. (2025), and Hasan (2025) highlight the mediating roles of cultural values, product meaning, and social norms in translating attitudes into intentions.

In emerging contexts like Indonesia, where sustainable consumption is still emergent, the attitudinal foundations of behavioral change merit closer attention. Recognizing this, the following hypothesis is advanced:

2.5. Mediation Role of Pro-Environmental Attitude

Within the architecture of TRA, behavioral intention is not shaped in isolation but often unfolds through layered cognitive appraisals. Attitudes emerge from these evaluations, bridging the terrain between externally provided information and internally held values (Fishbein & Ajzen, 1975). In settings where sustainability remains emergent—such as among young consumers in developing economies—the formation of such attitudes may serve as a critical conduit through which beliefs acquire motivational force (Anisah et al., 2024).

Two external referents—exposure to sustainability narratives via social media and perceptions of product durability—lie along this evaluative continuum (El-Shihy & Awaad, 2025; Madinga et al., 2024). While both constructs may exhibit direct associations with behavioral intention, a growing body of scholarship points to the mediating strength of attitudinal development in translating these stimuli into purposive action (Luthfiana et al., 2024; Wang et al., 2021). The attitudinal path offers a psychologically anchored mechanism through which meaning, relevance, and emotional congruence are forged before intent takes form.

Though sustainability content on digital platforms has been linked to both attitudinal and intentional outcomes (Jin et al., 2024; Singh, 2024), few models formally conceptualize attitudes as mediators in this process. Clarifying this intermediary role may yield richer insights into how social and informational stimuli are internalized. Accordingly, we put forth the following hypothesis:

Similarly, perceived product durability reflects not just an appraisal of longevity, but also an interpretive stance toward quality, responsibility, and long-term utility. When aligned with sustainability values, these beliefs influence behavioral intentions more reliably when filtered through attitudinal alignment. While durability is recognized as a predictor of sustainable behavior (Madinga et al., 2024; Ngo et al., 2024), its indirect operation via attitudinal mediation remains underexplored. Understanding how perceptions of durability are cognitively and affectively processed before manifesting in behavioral commitment may enhance both theoretical nuance and practical relevance. Therefore, we propose:

2.6. Moderation Role of Perceived Inertia

Within the TRA, attitudes are situated as proximal antecedents to behavioral intentions, formed through reflective appraisals of subjective beliefs and perceived social norms (Fishbein & Ajzen, 1975; Tiwari et al., 2024). The framework presupposes a decision-making process guided by rational evaluation, in which individuals are expected to align their beliefs and actions coherently. However, empirical research reveals a persistent disjunction between pro-environmental attitudes and their behavioral enactment, particularly in the context of sustainable consumption (Ogiemwonyi & Jan, 2023). This divergence—often conceptualized as the attitude–behavior gap—signals the presence of both psychological and contextual frictions that inhibit the translation of internal convictions into observable behavior.

Perceived inertia stands out among the various psychological barriers in this disconnection. It refers to a tendency to adhere to prior choices, often motivated by aversion to cognitive disruption, affective attachments to routine, and apprehension about the implications of alternative courses (Polites & Karahanna, 2012). This view is supported by behavioral psychology, which holds that habits are the cumulative result of repeated stimulus-response cycles—patterns that tend to reinforce existing behaviors while reducing receptivity to change (Wood & Neal, 2007). Within the fashion domain, inertia may manifest as habitual loyalty to fast-fashion brands, recurring impulsive purchasing, or lingering skepticism toward the aesthetic appeal of sustainable alternatives.

Empirical evidence increasingly positions inertia as a salient inhibitor of sustainable product adoption (Kemi & Zilahy, 2024; Khan et al., 2023; Wiederhold & Martinez, 2018). In conceptual modelling, inertia is often conceptualized as a contingency factor that moderates the interplay between psychological drivers and behavioral intent (Hu et al., 2024; Wang et al., 2020). This suggests that the inclination to preserve the status quo may dilute the otherwise positive influence of pro-environmental attitudes on ITS. In a fashion ecosystem shaped by symbolism, affect, and routinized consumption (Song & Bai, 2020), the explanatory power of inertia becomes particularly salient in illuminating the attitude–intention divide.

Despite its frequent citation as a barrier to behavioral change, the moderating role of inertia in the link between pro-environmental attitudes and ITS within sustainable fashion remains empirically underexamined. In light of these dynamics, it is anticipated that the influence of pro-environmental attitudes on intention may be attenuated when perceived inertia is high. Based on these conceptual arguments and empirical gaps, the following hypotheses are proposed:

3.1. Sampling Procedure and Data Collection Process

This study uses a quantitative, online survey approach to map the tendency of Indonesia’s young generation to shift their preferences from fast fashion to sustainable fashion. The target participants are 17 to 43 years old, representing Generation Z and millennials who are socially active and digitally active in the social media ecosystem. The selection of this age range considers the legal limits of consumption decisions in Indonesia and the high exposure of this group to digital fashion trends and sustainability content (IDN Research Institute, 2025).

Before the main data collection, the instrument was piloted with 30 respondents to ensure the clarity of the items and the consistency of the measurements. A panel of experts reviewed the construct’s validity, and internal reliability was assessed using Cronbach’s alpha, which yielded values above the minimum threshold of 0.70 (Hair et al., 2019).

Participants were recruited using purposive non-probability sampling based on predefined criteria, including residence in Indonesia, active engagement with social media platforms, and experience in purchasing fashion products. While not intended for statistical generalization, this sampling method enhances contextual relevance and alignment with the research objective.

The minimum required sample size was calculated using G*Power, yielding at least 138 respondents for five predictor variables, a statistical power of 0.95, a significance level of 0.05, and a medium effect size (f² = 0.15) (Memon et al., 2020). Data was collected over three months, from September to November 2024, utilizing Google Forms distributed via Instagram and WhatsApp to maximize geographical diversity. Responses from the Indonesian archipelago included Java, Sumatra, Kalimantan, Sulawesi, Bali, Nusa Tenggara, Maluku, and Papua. All participants provided voluntary consent before taking the survey. Of the 371 responses collected, 306 were deemed valid and included in the analysis, based on age-range and outlier checks conducted after the screening process.

3.2. Measurement Scales

The measurement instruments in this study are adapted from scales that have been empirically tested in prior literature, with contextual adjustments that preserve the conceptual meaning of each construct. The adaptation process follows a systematic approach to scale development, as suggested by DeVellis (2017), to ensure construct validity and consistency with the theoretical framework underlying the research model.

The social media influence construct is measured through four statements adapted from Li et al. (2024) and Xie and Madni (2023). These statements reflect how exposure to digital content shapes perceptions of sustainability and the orientation toward eco-friendly consumption. Pro-environmental attitudes are measured using four indicators from Hageman et al. (2024), which examine the affective and cognitive dimensions of awareness and commitment to environmental conservation through consumption behavior.

Perceived durability is measured by four items adapted from Jin et al. (2024) and Lopes et al. (2024) that reflect perceptions of product durability and long-term use value in sustainable fashion. Perceived inertia was evaluated using three items from Akbari et al. (2023) that capture the tendency to maintain fast-fashion consumption habits despite the availability of more sustainable alternatives. Meanwhile, the ITS is measured by four statements, adapted from the instrument by Yang et al. (2024), with the scope of the concept expanded from second-hand fashion to ITS to the overall practice of fashion consumption that is more environmentally responsible.

The entire construct uses a seven-point Likert scale, ranging from one (strongly disagree) to seven (strongly agree). This scale was chosen because it offers greater precision in capturing respondents’ gradations of attitudes and preferences and allows for a more nearly normal distribution of data for quantitative analysis (Joshi et al., 2015). Details of each item used are in Appendix 1.

3.3. Data Analysis

Data analysis in this study was carried out using the PLS-SEM approach in SmartPLS 4. This approach was chosen because it provides methodological flexibility for handling conceptual models involving multiple latent constructs, non-normal data distributions, and small-to-medium sample sizes. As a variance-based method, PLS-SEM prioritizes predictive aspects and aims to maximize the variance explained by endogenous constructs, making it relevant for exploratory-oriented studies and conceptual model development (Dash & Paul, 2021; Hair et al., 2019; Sarstedt et al., 2022).

The analysis procedure follows the two-stage approach commonly used in PLS-based studies, including the evaluation of measurement models and structural models. The first stage involves testing the reliability and validity of the construct, including internal consistency, convergent validity, and discriminant validity, to ensure the quality of the instruments used. The second stage focuses on testing the relationships between constructs in structural models and assessing the model’s predictive strength and overall fit. To test the significance of the model’s pathways, a bootstrapping technique with 5,000 resamples was applied, enabling reliable estimation of direct, mediating, and moderating effects within a single analytical framework.

4.1. Profile of the Sample

The description of the respondents is listed in Table 1. A total of 306 valid data points were analyzed, consisting of 166 women (54.2%) and 140 men (45.8%). Generation Z (61.8%) was the most significant proportion, followed by the millennial generation (38.2%). Most have a diploma or bachelor’s degree (59.5%) and a monthly income of less than IDR 2,000,000 (39.9%). By employment status, the student group is the largest (36.3%), followed by private employees (23.2%). Regarding social media usage, WhatsApp and Instagram are the most frequently accessed platforms (28.8% each), followed by TikTok (21.6%). Online shopping activity over the last six months shows that most respondents make fashion purchases 1–2 times (41.8%) and 3–5 times (36.9%), indicating relatively high engagement in digital consumption.

4.2. Common Method Bias Control and Evaluation

In survey-based quantitative studies with self-report measures, common method bias (CMB) is a potential systematic bias that can lead to inaccurate estimates of relationships among constructs, particularly when all data are obtained from a single source (Podsakoff et al., 2003). Several ex-ante procedures have been implemented to minimize potential bias, including item sequencing and respondent anonymity guarantees, to encourage honest responses and reduce the possibility of social desirability bias (Bartell & Spyridakis, 2012; Chang et al., 2010).

In addition to the procedural approach, the potential of CMB is also statistically evaluated through two primary methods. The first step is to use Harman’s Single-Factor Test to detect the dominance of one common factor in the data (Harman, 1960). The analysis showed that a single factor explained only 38.26% of the total variance, well below the 50% threshold, indicating that CMB did not dominate the data structure. Nevertheless, as Podsakoff et al. (2003) noted, the approach has limitations because it cannot detect method bias that spreads across constructs.

The complete collinearity assessment approach (Kock, 2017) was also employed to further assess potential common method bias. The overall VIF ranges from 1.022 to 1.481, well below the conservative threshold of 3 (Hair et al., 2019), indicating the absence of collinearity or significant general-method bias.

By combining procedural control strategies and statistical evaluation, general method bias does not significantly influence the overall validity of the structural model, thus supporting the integrity of this study’s empirical findings.

4.3. Evaluation of Measurement Models

As a first step in the PLS-SEM approach, assessing the measurement model is essential to ensure the construct’s reliability and validity. The reliability of the construct was evaluated through two leading indicators, namely Cronbach’s Alpha (CA) and Composite Reliability (CR), with a minimum value of 0.70 as the statistical feasibility limit (Hair et al., 2019). As shown in Table 2, the entire construct recorded CA values between 0.744 and 0.837 and CR values between 0.855 and 0.901. This range not only meets the minimum requirements but also indicates a stable and acceptable internal consistency in the context of latent measurement.

The construct’s validity is analyzed along two dimensions: convergent and discriminant. Convergent validity was assessed using factor loading values and Average Variance Extracted (AVE). Of the 19 initial indicators, one indicator (SMI1: 0.690) was eliminated because it did not meet the minimum loading threshold of 0.70. Other maintained indicators showed loading values ranging from 0.774 to 0.896, which were statistically significant and stable. The AVE for the entire construct ranges from 0.617 to 0.752, above the minimum threshold of 0.50, which is generally accepted as an indicator of convergent validity (Hair et al., 2019).

The discriminant validity was tested using three approaches: cross-loadings among indicators, the Fornell-Larcker criteria, and the Heterotrait-Monotrait Ratio (HTMT). Of the three, HTMT is recognized as the more sensitive method for detecting construct overlap (Henseler et al., 2015). An HTMT value below 0.85 indicates adequate conceptual differentiation (Hair et al., 2019). As summarized in Table 3, all HTMT values range from 0.049 to 0.829. This shows that each of the constructed measures is not only internally valid but also theoretically separate from the others.

After fulfilling all reliability and validity requirements at the measurement model stage, the analysis can proceed to structural model evaluation to examine the strength of the relationships within the theoretical framework.

4.4. Predictive Feasibility and Structural Model Fit

Structural model evaluation in the PLS-SEM approach not only considers the significance of pathways between constructs but also includes the model’s ability to explain endogenous construct variance, predictive accuracy, and overall model fit. To assess this quality, four leading indicators were used: the determination coefficient (R²), Q² Predict, PLS Predict, and the Standardized Root Mean Square Residual (SRMR).

R² reflects the magnitude of the proportion of endogenous construct variance that exogenous constructs in the model can explain. Based on the classification proposed by Hair et al. (2019), the R² values of 0.418 for ITS and 0.520 for pro-environmental attitude are moderate, indicating a theoretically feasible model’s explanatory capacity.

Meanwhile, Q² Predict assesses the model’s predictive relevance beyond the sample data. A Q² value greater than zero indicates that the model has adequate predictive power over new data (Hair et al., 2022). In the context of this study, Q² values of 0.357 for ITS and 0.498 for pro-environmental attitudes indicate that the model has stable predictive relevance at moderate levels.

Further assessment of predictive accuracy was carried out using the PLS Predict approach, which compares the performance of the PLS-SEM model against a linear regression (LM)- based benchmark model, as suggested by Shmueli et al. (2019). The analysis showed that most of the PLS-SEM model’s Root Mean Square Error (RMSE) and Mean Absolute Error (MAE) values were below those of the LM models. These findings suggest the PLS-SEM model has consistent predictive advantages, although not extreme. A summary of the comparison results is presented in Table 4.

As a complement to the structural evaluation, the Standardized Root Mean Square Residual (SRMR) value is used to assess the overall fit of the model (model fit). According to the tolerance limits set by Hu and Bentler (1998), an SRMR ≤ 0.08 is considered an acceptable indicator of model fit. This study reported an SRMR of 0.08, indicating that the model met the structural fit threshold for advanced interpretation.

4.5. Direct Effects

Before testing the hypothesis, the data distribution was evaluated by analyzing skewness for each construct. All values are within ±2, indicating that the data distribution meets the assumption of acceptable normality in the context of PLS-SEM (Hair et al., 2022). Based on these findings, a bootstrapping procedure with 5,000 subsamples was applied using a percentile technique at a 5% significance level (critical t-value = 1.96) to test the direct relationship, the mediating effect, and the role of moderation in structural models (Hair et al., 2022). A summary of the test results is presented in Table 5, and a visualization of the path is included in Figure 2.

The results indicated that SMI’s influence on PEA was statistically significant (β = 0.404; t = 4.923), supporting H1. The magnitude of the effect is assessed using the effect size (f² = 0.229), which falls in the medium range according to Cohen’s (1988) guidelines. In addition, PD was shown to have a positive and significant effect on PEA (β = 0.410; t = 5.333), supporting H2, with similar effects in the moderate category (f² = 0.236). Furthermore, the association between PEA and ITS was significant (β = 0.584; t = 11.828), providing empirical support for H3. In this case, the effect size was classified as large (f² = 0.574), indicating the central role of PEA in mediating the influence of external factors on ITS to sustainable fashion products.

Figure 2Structural Model 

4.6. Mediation Analysis

The mediation analysis was conducted to evaluate the PEA’s role as a mediator in the relationships among SMI, PD, and ITS. As shown in Figure 2, the analysis results indicate that the indirect pathway from SMI to ITS via PEA is significant (β = 0.236; t = 4.769), supporting the H4 hypothesis and demonstrating a full mediating effect.

Similarly, a full mediating effect was found for the PD → PEA pathway → ITS (β = 0.239; t = 4.170), supporting the H5 hypothesis. These findings confirm that PEA plays a crucial psychological role in bridging the influence of SMI and PD on individuals’ ITS to more sustainable fashion choices.

4.7. Moderation Analysis

The effect of PI moderation on the relationship between PEA and ITS was analyzed using a variable-interaction approach within the PLS-SEM structural model. The bootstrapping results revealed that the interaction between PEA and PI on ITS was significant and negative (β = –0.176; t = 2.267; p = 0.023; f² = 0.057), providing empirical support for H6. These findings suggest that psychological inertia can reduce the power of attitudes towards ITS to influence the adoption of more sustainable fashion alternatives.

The moderation effect is visualized in Figure 3. At low PI levels (–1 SD; indicated by the solid line), the slope of the relationship between PEA and ITS appears more pronounced, suggesting that under conditions of minimal habitual inhibition, pro-environmental attitudes exert a stronger influence on ITS. In contrast, at high PI levels (+1 SD; indicated by the dashed line), the regression line becomes flatter, indicating that higher inertia weakens the influence of pro-environmental attitudes on ITS. The line representing the average PI level shows a moderate slope between these two conditions.

Although the relationship between PEA and ITS remains positive across all PI levels, the differences in slope gradients indicate that PI functions as a negative moderator. In other words, PI does not reverse the direction of the relationship but reduces the strength of the influence of pro-environmental attitudes on ITS. These findings highlight that psychological resistance to change remains an important barrier that should be considered when designing interventions aimed at encouraging sustainable consumption behavior.

Figure 3Simple slope analysis 

6.1. Theoretical Contribution

This study contributes theoretically by proposing and empirically validating an integrated framework that combines the motivational logic of the TRA with the constraint-based perspective of SQB. Through this integration, the study offers a more nuanced explanation of the attitude–intention gap in sustainable fashion consumption by showing that favorable attitudes alone are insufficient to generate ITS when inertia remains high. In this sense, the study extends TRA by demonstrating the value of incorporating behavioral resistance into models of sustainable consumption, particularly in contexts where change requires consumers to depart from familiar routines.

A further contribution lies in refining how inertia is interpreted in emerging-market settings. The present findings suggest that inertia should not be understood solely as an internal cognitive tendency detached from consumption conditions. Instead, its observed role should be interpreted with attention to the economic context in which consumers make choices, especially when sustainable alternatives are perceived as less financially accessible. This does not mean that inertia is reducible to economic constraints; rather, it suggests that inertia may be amplified under affordability-related pressures. In this respect, the study indicates that the attitude–intention gap may reflect not only reluctance to abandon familiar consumption routines, but also difficulty in switching when sustainable alternatives are perceived as financially demanding.

The study also extends the interpretation of attitude formation among young consumers. While the model is grounded in TRA and does not explicitly incorporate identity-based constructs, the findings suggest that external stimuli such as social media influence and perceived durability may derive part of their motivational force from the social and symbolic meanings attached to sustainable consumption. In this sense, the evaluative processes underlying ITS may reflect not only functional or ethical judgments, but also concerns related to self-expression, social belonging, and the projection of a responsible identity. Together, these insights suggest that sustainable consumption among young consumers is shaped not only by what products do, but also by what they mean within socially visible consumption contexts.

Taken together, these findings provide a more context-sensitive framework for understanding sustainable fashion transitions in developing economies, where ITS are shaped simultaneously by evaluative processes, social-symbolic meanings, and behavioral resistance.

6.2. Practical Implications

The findings offer practical insights for stakeholders seeking to promote sustainable fashion among young consumers in emerging markets such as Indonesia. More specifically, the results suggest that efforts to encourage switching should address not only positive attitudes toward sustainability, but also the frictions that prevent those attitudes from translating into behavioral intention.

For marketers and fashion brands, sustainable fashion should be positioned not only as an ethical and practical choice, but also as an identity-relevant and socially resonant one. Because young consumers may respond not only to product value but also to symbolic and peer-related meanings, communication strategies should move beyond emphasizing durability and long-term usefulness alone. Campaigns are likely to be more effective when they present sustainable fashion as a way to express personal values, signal responsibility, and connect with a desirable peer community. At the same time, practical arguments—such as durability, cost-per-wear, versatility, and timeless design—remain important for reducing skepticism about value. Social media campaigns can strengthen this message by combining aspirational identity cues with concrete demonstrations, such as styling challenges, wardrobe rotation ideas, or garment-care tutorials that show how sustainable fashion can be both socially meaningful and functionally worthwhile.

For product developers and retailers, reducing switching burden is critical. Since inertia weakens the translation of favorable attitudes into ITS, business models that lower perceived risk, effort, and upfront commitment may facilitate behavioral transition. Illustrative examples from the Indonesian market include resale-oriented platforms such as Tinkerlust, which broaden access to preloved fashion through a dedicated digital marketplace, and lifecycle-oriented models such as SukkhaCitta, which offers lifetime repair and return-based circularity for its products. These approaches can lower experimentation costs, reduce uncertainty about product value, and reinforce the perception that higher-quality products are more accessible and easier to integrate into contemporary youth lifestyles.

For policymakers and sustainability advocates, the findings highlight the importance of improving both access to and public legitimacy for sustainable choices. Measures that support local production, circular fashion initiatives, or greater price competitiveness may help reduce affordability-related barriers in the market. Public campaigns and consumer education should also move beyond environmental awareness and product value alone by helping to normalize sustainable fashion as a socially relevant form of consumption. In Indonesia, community-based initiatives such as Sustainable Fashion Fest 2025—which featured a Repair Corner, Clothes Swap activities, and the launch of Rekynd Hub for local preloved and textile circularity—illustrate how sustainability can be made more visible and easier to adopt in everyday practice.

Taken together, these implications suggest that addressing the inertia trap requires a dual strategy: strengthening the practical, symbolic, and social appeal of sustainable fashion while simultaneously reducing the habitual and affordability-related barriers that make switching difficult. Such an approach is particularly important in developing-market contexts, where pro-environmental attitudes may already be present, but the transition to sustainable consumption remains constrained by both everyday purchasing realities and the social meanings attached to fashion choices.

6.3. Limitations and Suggestions for Future Research

Although this study makes a meaningful contribution to understanding young consumers’ ITS toward sustainable fashion, several methodological and conceptual limitations should be acknowledged. First, the use of a cross-sectional design limits the ability to capture how ITS evolve. Future studies could adopt longitudinal or experimental designs to examine the temporal stability of the observed relationships and to better track the transition from favorable attitudes to actual behavioral change.

Second, although the purposive sampling strategy was appropriate for reaching digitally active Gen Z and millennial consumers in Indonesia, it limits the generalizability of the findings to other demographic groups and national contexts. In addition, the economic profile of the sample should be considered when interpreting the role of inertia, as a substantial proportion of respondents reported monthly incomes below IDR 5,000,000. Under such conditions, the observed effect of inertia may partly overlap with affordability-related concerns. Future studies may therefore benefit from comparing consumer groups across income levels, age segments, or national settings to examine whether the role of inertia varies across different structural conditions.

Third, the study relies on self-reported perceptual measures, which may be subject to social desirability bias and common method bias. Beyond these concerns, this approach also limits the ability to clearly distinguish between resistance to change as a psychological tendency and constraints arising from affordability. While the findings indicate that perceived inertia weakens the relationship between attitude and intention, the current design does not allow for a clear differentiation between a lack of willingness to switch and a limited ability to do so. Future research could address this limitation by incorporating more nuanced measures, such as perceived affordability, switching costs, budget constraints, and willingness to pay.

Fourth, the study focuses on intention to switch rather than actual economic commitment or observed behavior. Given the well-documented gap between intention and action in sustainable consumption, future research would benefit from moving beyond self-reported intention toward more behaviorally grounded measures. This includes directly assessing consumers’ willingness to pay for sustainable fashion. Such efforts could involve experimental choice tasks, discrete choice or conjoint designs that incorporate price trade-offs, or simulated purchase scenarios that reflect realistic budget constraints. In addition, the use of incentive-compatible willingness-to-pay measures and behavioral data, such as actual purchase decisions, transaction records, or digital purchase tracking, would provide stronger evidence of whether consumers who express pro-environmental attitudes and ITS are indeed willing to pay a premium under real market conditions.

Finally, although integrating TRA and SQB strengthens the model’s explanatory capacity, further refinement remains possible. Future studies may incorporate additional psychological and contextual variables—such as perceived risk, social influence, environmental guilt, affordability perceptions, and access-related barriers—to build a more comprehensive account of sustainable fashion transitions. Alternative theoretical perspectives, such as the Value-Belief-Norm Theory or the Protection Motivation Theory, may also complement the present framework. Taken together, these directions would provide a stronger empirical basis for distinguishing psychological resistance from structural constraints on consumption in sustainable fashion behavior.