“I Am Willing, but Acting Is Hard”: Unpacking the Intention–Behavior Gap in Carbon Inclusivity Mechanisms Using Social Practice and MOA Theory

Abstract

Carbon inclusivity (CI) mechanisms face weak low-carbon adoption despite strong consumer intention in China. Prior studies explain low-carbon behavior mainly from individual psychological perspectives yet rarely investigate the intention–behavior gap in the digital–physical scenarios of CI mechanisms. Furthermore, single theoretical lenses such as the Motivation–Opportunity–Ability (MOA) framework or Social Practice Theory (SPT) have clear limitations in interpreting this gap. This study integrates the MOA framework with SPT to construct a novel “Motivation–Intention–Context–Practice–Behavior” (MICPB) model, framing behavioral change as a dynamic interplay of internal motivation, external opportunities, and evolving social practices. Different from studies adopting MOA or SPT in isolation, the MICPB model bridges individual psychological processes and routine social practices to unpack the intention–behavior gap. Adopting grounded theory to analyze 42 interviews from China’s CI pilot regions, we identify three core dimensions of the gap: limited internal competencies and emotional conflicts, fragmented external incentives and platform defects, and misalignment between low-carbon values and daily routines. The MICPB model reveals that the gap stems from misalignment across motivation, context and practice, and enriches MOA theory by incorporating practice-based dynamics. This integrated framework deepens understandings of barriers to consumer low-carbon participation under CI schemes. This study suggests coordinated policy and corporate efforts to restructure social practices, bridge the intention–behavior gap, and align digital governance with low-carbon transitions.

1. Introduction

Carbon inclusivity (CI) mechanisms have emerged as a groundbreaking policy innovation in China’s climate governance framework, representing a novel approach to mobilize public participation in emission reduction through digitalized incentive systems [1]. As a voluntary carbon reduction mechanism integrating “Internet + Big Data + Carbon Finance”, CI quantifies and rewards low-carbon behaviors at the individual and household levels, marking a significant departure from traditional command-and-control environmental policies [2]. Unlike Personal Carbon Trading (PCT), which is market-based, and carbon management platforms for enterprises, which focus on enterprise-level emissions accounting, CI is characterized by its voluntary, incentive-driven, digitally mediated design that targets individuals’ daily low-carbon behaviors. While pilot programs in Beijing, Shanghai and other regions have demonstrated initial success in raising climate awareness, a persistent disconnect between environmental intentions and actual behaviors threatens to undermine their long-term effectiveness—a phenomenon encapsulated in the “intention–behavior gap” paradox [3,4].

The urgency of addressing this gap becomes evident when examining current CI participation patterns. Recent national surveys reveal that although public awareness of carbon inclusivity reached 36.8% nationwide in 2025—and exceeded 50% in pilot regions—the actual platform registration rate remained only 17.9% [5]. This discrepancy manifests across multiple behavioral domains, from intermittent use of public transportation to abandonment of energy-saving appliances, suggesting systemic rather than isolated challenges. Such behavioral attrition not only diminishes CI’s environmental impact but also represents substantial sunk costs in platform development and policy implementation, highlighting the critical need to understand the underlying mechanisms of participation drop-off.

Existing literature on pro-environmental behavior has identified several dimensions of the intention–behavior gap, including psychological barriers (e.g., self-efficacy deficits), economic constraints (e.g., high green premiums), and structural limitations (e.g., infrastructure gaps) [6]. However, these perspectives remain fragmented when applied to CI contexts, failing to account for (1) the unique digital–physical interaction dynamics of CI platforms; (2) the evolving social practices surrounding emerging low-carbon technologies; and (3) the multi-stakeholder governance structure characteristic of CI systems. This theoretical incompleteness is particularly striking given CI’s rapid scaling—from 18 pilot cities in 2021 to nationwide implementation plans by 2025.

The current research landscape reveals two critical knowledge gaps. First, while studies on mandatory personal carbon trading mechanisms abound [7], scholarly attention to voluntary mechanisms like CI remains disproportionately scarce, with fewer than 20 empirical studies published in international journals as of 2024 (Web of Science data). Second, extant research predominantly employs quantitative methods that overlook the lived experiences and contextual complexities shaping CI participation [8]. This methodological limitation obscures the processual nature of behavior change and the socio-technical transitions inherent in CI adoption. Recent scholarship on digital sustainability transitions underscores that platforms like CI are not merely technological tools but are becoming central to the future of energy provisioning, actively reshaping the relationship between users and energy systems [9]. A narrow methodological focus risks missing these critical dynamics.

Nevertheless, existing studies only examine individual or structural factors in isolation and fail to clarify the interactive mismatches between individual psychological drivers and socio-material social practices that cause the CI intention–behavior gap. Resolving this limitation can advance theoretical understanding of the intention–behavior gap beyond context-general models. Our study addresses these gaps by introducing a novel theoretical synthesis. We integrate the individual-focused Motivation–Opportunity–Ability (MOA) framework [10] with the practice-oriented Social Practice Theory (SPT) [11] to develop a multidimensional theoretical lens. The novel theoretical lens (namely, the proposed comprehensive theoretical model in this study) is specifically designed to overcome the fragmentation in existing literature as follows. (1) Capturing Agency and Structure: It accounts for individual psychological drivers (Motivation–Intention) while simultaneously examining the constraining and enabling forces of the socio-material context (Context) and routinized behaviors (Practice). (2) Illuminating the “Black Box” of Transition: It conceptualizes the journey from intention to sustained behavior not as a linear step, but as a dynamic process mediated by the alignment (or misalignment) between one’s motivations, the digital and institutional opportunities, and the embeddedness of the new behavior in daily practices. This integrated view is crucial for designing digital environmental platforms, as it shifts the focus toward empowering users and aligning interventions with their lived experiences and practice routines—a key factor for sustaining engagement often missed by narrowly designed feedback systems [12].

The primary objective of this research is to construct and apply a comprehensive theoretical model to elucidate the intention–behavior gap in CI mechanisms. Through a grounded theory analysis of 42 in-depth interviews, we seek to (1) delineate the specific factors that constitute the gap across motivational, contextual, and practical dimensions; and (2) demonstrate how this integrated framework provides a more holistic and actionable explanation than prior theoretical approaches.

The study’s significance is threefold. Theoretically, it advances sustainable consumption literature primarily through the development of a unified framework that bridges psychological and sociological perspectives on behavior change. Methodologically, it demonstrates the value of qualitative inquiry to uncover hidden dimensions of digital–environmental behaviors. Practically, the proposed theoretical model serves as a diagnostic and design tool, providing evidence-based recommendations for optimizing CI platform designs, calibrating policy incentives, and fostering cross-sector collaborations.

2. Literature Review

2.1. Theoretical Foundations of the Intention–Behavior Gap

The persistent disconnect between environmental intentions and actual behaviors constitutes one of the most significant challenges in sustainability science. This intention–behavior gap has been widely observed across diverse domains of pro-environmental behavior, ranging from energy conservation to sustainable transportation choices [3]. Meta-analytic evidence suggests that although a substantial proportion of consumers in developed economies hold positive environmental attitudes, only a small group consistently translates these attitudes into sustainable consumption practices [13]. The gap appears even more pronounced in emerging digital environmental platforms represented by Carbon Inclusion (CI) mechanisms. While this gap aligns with the broader research on the intention–behavior gap, it exhibits distinct features stemming from platform-specific barriers. Recent survey data from China reveal that while 81.5% of the public recognize the value of low-carbon actions and 68.3% express willingness to participate, the actual platform registration rate stands at only 17.9%, a gap of over 50 percentage points [5].

Drawing on Steg & Vlek [14], this study proposes an original three-generation classification based on shifts in the analytical focus of environmental behavior research. This evolutionary trajectory reflects the growing complexity of our understanding of environmental behavior and the increasing recognition of the multifaceted nature of the intention–behavior gap.

① First Generation: Cognitive Models (1990s–2000s)

The Theory of Planned Behavior (TPB) [15] lays the foundational framework for understanding the intention–behavior relationship. This influential model identified three core determinants of behavioral intention: attitudes toward the behavior, subjective norms and perceived behavioral control.

These cognitive factors were posited to directly influence behavioral intentions, which in turn were theorized to predict actual behavior. While this model achieved significant predictive success in some domains, subsequent research revealed that its application has encountered challenges, manifested as an inconsistency between high green consumption intentions and low actual behaviors [16].

② Second Generation: More Sophisticated Extended Frameworks (2000s–2010s)

The limitations of first-generation cognitive models prompted the development of more sophisticated theoretical frameworks during the 2000s–2010s, marking the second generation of research on the intention–behavior gap. This period witnessed significant theoretical diversification through three principal advancements. First, affective–cognitive models, particularly Behavioral Reasoning Theory [17], elucidated the dual pathways through which justification mechanisms explain a considerable share of variance in the gap: (1) global motives rooted in enduring environmental values, and (2) specific reasoning processes evaluating immediate product attributes and situational considerations. Second, contextual approaches embodied by the Attitude–Behavior–Context model [10] indicate that contextual factors can strongly moderate intention–behavior linkages in environmental contexts. Their work highlights three prominent barriers: economic constraints, limited availability of sustainable alternatives, and social network influences. Third, the integration of personality psychology through the Big Five framework revealed systematic individual differences in pro-environmental behavior, with conscientiousness consistently emerging as the most robust predictor and openness to experience demonstrating significant but more modest effects, mediated substantially by environmental concern [18]. This multidimensional expansion fundamentally enriched our understanding of the gap’s complexity beyond purely cognitive explanations.

③ Third Generation: Practice-Oriented Approaches (2010s–present)

The third generation of research (2010s–present) has witnessed a paradigm shift toward practice-oriented approaches, responding to fundamental critiques of earlier individualistic models. This emerging perspective emphasizes the crucial role of material and social contexts in shaping environmental behaviors, revealing three critical limitations in conventional frameworks. First, systematic reviews indicate that the majority of existing studies fail to account for the essential dependencies between behaviors and their supporting infrastructures, both physical and digital [11]. Second, longitudinal analyses remain strikingly underrepresented, with only a small proportion of investigations examining behavioral evolution across meaningful time periods, thereby neglecting the dynamic nature of habit formation and practice stabilization [19]. Third, of particular relevance to contemporary digital platforms, existing theoretical frameworks exhibit notable limitations in capturing digital–physical behavioral hybridity, as only a small fraction of models accounts for the distinctive features of app-mediated environmental actions. These insights collectively underscore the need for more holistic approaches that transcend individual-level analyses to incorporate the complex interplay between technologies, temporal dimensions, and socio-material contexts in environmental behavior research.

Although this three-generation classification facilitates analytical clarity, pro-environmental behavior theories do not have rigid boundaries and contain substantial conceptual overlaps. For instance, both the TPB and Behavioral Reasoning Theory treat environmental concern and behavioral intention as core constructs; situational constraints emphasized in the ABC model also intersect with contextual reasoning described in Behavioral Reasoning Theory. Such overlaps arise from divergent disciplinary roots (social psychology, economics, sociology) and varying focal mechanisms across theories.

This theoretical progression sets the stage for more integrative approaches that can better account for the complex realities of contemporary environmental behaviors, particularly in emerging digital platforms like CI mechanisms. Furthermore, evidence is mounting that the success of such digital–environmental systems hinges on their ability to align with existing social practices and to address the complex interplay of non-financial motivations and social norms, aspects that pure economic or technologically deterministic models frequently overlook [20].

Overall, the three generations of research reflect a gradual shift in environmental behavior scholarship: the research focus expands from individual cognition and intention formation toward contextual and socio-material perspectives. Each theoretical stream offers complementary insights into the intention–behavior gap yet emphasizes distinct components of behavioral processes. Cognitive models represented by the TPB primarily explain intention formation and its link to subsequent behavior; extended frameworks incorporate additional variables including behavioral reasoning, contextual conditions, and individual heterogeneity; and practice-oriented approaches stress how environmental behaviors are embedded within broader socio-material arrangements covering materials, competences and infrastructures. Nevertheless, relying solely on any single perspectives yields partial explanations for why individuals with environmental awareness and positive intentions fail to adopt or sustain carbon-inclusive consumption practices.

These separate theoretical strands remain largely disconnected, with few attempts to reconcile individual psychological mechanisms and socio-material constraints within a unified analytical framework. To address this issue, our study integrates the Motivation–Opportunity–Ability (MOA) framework with Social Practice Theory (SPT). The MOA framework provides an individual-level perspective for examining how motivation, opportunities, and abilities shape the translation of intentions into behavior, while SPT situates these processes within the broader social and material contexts in which practices are performed and reproduced. The two perspectives are therefore complementary: MOA helps explain individual-level mechanisms underlying behavioral enactment, whereas SPT highlights the role of socio-material conditions in shaping and constraining such behavior. Their integration provides a suitable theoretical basis for examining the knowledge–behavior gap in carbon-inclusive consumption, particularly in digital platform contexts where participation depends on both individual-level conditions and supportive social, material, and technological environments.

2.2. Integrative Frameworks for CI Mechanisms

2.2.1. MOA Theory: Tripartite Determinants

The Motivation–Opportunity–Ability (MOA) framework [10] offers a systems perspective for analyzing CI participation, examining psychological, social and material enablers. MOA theory has been widely adopted in environmental behavior research to interpret individual pro-environmental decision-making, confirming that internal motivation, external situational opportunity, and personal ability jointly drive behavioral implementation and effectively explain individual-level low-carbon behavioral activation mechanisms. Motivation components include intrinsic drivers (self-efficacy, personal norms), extrinsic incentives (rewards, social recognition), and emotional catalysts (pride, guilt reduction) [21]. Evidence shows financial incentives alone sustain participation poorly without intrinsic motivation [22]. Opportunity structures in CI platforms employ digital nudges and social benchmarking to make sustainable choices more visible and rewarding [23,24]. Ability factors—including technical literacy, carbon numeracy, and habit formation—account for a considerable share of long-term engagement variance [25,26], underscoring the need for built-in user education in stimulating CI participation. Together, these tripartite determinants reveal how CI systems must simultaneously address motivational complexity, opportunity creation, and capability development to effectively bridge the intention–behavior gap.

Compared with other theories, the MOA framework focuses on the conditions under which intentions can be transformed into actual behaviors. This perspective is par-ticularly relevant to the intention–behavior gap in CI mechanisms, as participation re-quires not only motivation to engage in low-carbon actions but also sufficient opportunities provided by digital platforms and the ability to perform and maintain such behaviors.

Nevertheless, MOA suffers from inherent individualism-oriented limitations. It emphasizes instantaneous individual decision-making and ignores the constraint of routinized daily social practices on behavioral translation. In the digital–physical integrated scenario of China’s CI mechanisms, the intention–behavior gap is not merely an individual cognitive deficiency problem. Single MOA theory cannot explain why consumers with sufficient motivation, opportunity and ability still fail to sustain CI participation due to rigid daily life routines.

2.2.2. Social Practice Theory: Three-Element Framework

Social Practice Theory [11] reconceptualizes sustainable behaviors as dynamic practices co-constituted by materials (technologies/infrastructures), competences (skills/knowledge), and meanings (norms/identities). Different from individual cognitive theories, SPT shifts environmental behavior research from individual-level analysis to socio-material structural perspective, and effectively explains behavioral inertia and low-carbon transition resistance caused by solidified daily social practices in low-carbon research. This framework reveals how practice evolution occurs through element reconfiguration—for instance, smart meter adoption (material) requires energy literacy (competence) and reshapes environmental identities (meaning) [27]. It effectively explains behavioral inertia in unsustainable systems, such as fashion waste perpetuated by washing technologies and ingrained laundry routines [28], while also identifying intervention points through material redesign (e.g., eco-labels [29]) or competence development (e.g., food waste reduction [30]). Originally applied to energy practices, its emphasis on socio-technical interactions makes it particularly apt for analyzing digital–environmental systems like Carbon Inclusivity platforms, where user engagement depends on the alignment of app interfaces, carbon accounting skills, and green status meanings.

However, SPT has obvious explanatory boundaries. It downplays individual subjective intention and motivational heterogeneity, and cannot interpret the divergent behavioral translation results of individuals under similar social practice contexts. For China’s CI mechanisms with widespread positive public low-carbon intentions, single SPT fails to answer why most intention holders cannot translate their willingness into continuous low-carbon behaviors.

Crucially, neither MOA nor SPT alone can fully illuminate the intention–behavior gap in CI scenarios. MOA captures individual behavioral drivers yet overlooks structural constraints embedded in daily social practices. SPT accounts for routinised behavioral inertia but offers limited analytical leverage to explain heterogeneous individual motivations and capabilities under identical socio-material contexts.

This mutual limitation creates an unresolved theoretical tension that cannot be addressed by simply applying one framework in isolation. This study proposes a theoretically coupled framework instead of a simple supplementary combination. Epistemologically, MOA and SPT are mutually compatible: MOA adopts an agent-centered perspective to analyze individual behavioral decision-making, while SPT offers a structural lens centered on socio-material arrangements, jointly bridging the agent–structure divide that limits isolated behavioral theories. The three socio-material elements of SPT (materials, competences, meanings) shape and set boundaries for the operation of MOA’s individual-level determinants. Specifically, platform hardware and digital interfaces (materials) define accessible behavioral opportunities within CI systems; shared social norms and green identity values (meanings) strengthen or weaken individuals’ low-carbon motivation; practical operation skills (competences) overlap substantially with the ability dimension in MOA. In turn, aggregated individual actions driven by motivation, opportunity and ability continuously reshape material infrastructure, reconstruct social environmental meanings and update public competences, thereby facilitating the evolution of low-carbon social practices. Within the digital–physical integrated context of China’s CI mechanisms, unpacking the intention–behavior gap necessitates this mutual embedding of MOA’s individual cognitive logic and SPT’s socio-structural logic.

2.3. Carbon Inclusivity Research Landscape

The emerging literature on Carbon Inclusivity (CI) mechanisms reveals three critical gaps when examined through the lens of practice–theoretical approaches. First, while mandatory personal carbon trading schemes have been extensively studied [7], voluntary mechanisms like CI remain under-theorized. This disparity persists despite CI’s rapid adoption across China since 2015, suggesting a misalignment between research focus and current policy developments.

Second, existing CI studies show a strong quantitative orientation—Li et al. [8] found that 76.5% of analyzed publications relied on survey methods, primarily capturing behavioral intentions rather than actual practice dynamics. This methodological tendency aligns with broader critiques in green consumption research, where cross-sectional designs dominate (accounting for 89% of studies according to ElHaffar et al., 2020 [3]). Such approaches often fail to capture the co-evolution of material infrastructures (e.g., CI app interfaces) and social meanings (e.g., carbon credit symbolism) in daily practices, a process central to Shove et al.’s [11] practice–theoretical framework.

Theoretical integration also remains fragmented. Although behavioral theories (mostly TPB) have been increasingly adopted in CI research, Liu et al. [31] highlighted that relatively limited research has integrated motivation-oriented perspectives (e.g., MOA) with practice-oriented approaches (e.g., SPT). This disciplinary siloing inhibits understanding of how individual motivations intersect with sociotechnical systems—a gap acutely relevant for CI platforms that simultaneously require personal engagement and infrastructural support.

To further clarify the research progress and unresolved limitations of existing empirical CI literature, this study systematically reviewed all international empirical studies focused on China’s carbon inclusivity mechanisms published before 2024. A comparative summary of core studies is presented in

Table 1. Summary of Empirical Studies on Public Participation Willingness in China’s CI Mechanisms.

Research Paradigm	Published Literature	Theoretical Basis	Methodology	Core Findings
Quantitative cognitive studies (MOA perspective)	Li et al. (2024) [32]; Weng et al. (2025) [33]; Li et al. (2023) [8]; Zhang et al. (2025) [34]; Wei et al. (2024) [35]; Tan et al. (2019) [1]; Wang et al. (2023) [19]; Liu et al. (2023) [31]	MOA, TPB, evolutionary game theory, incentive economics, policy evaluation framework	Questionnaire regression, quasi-natural experiment, evolutionary game simulation, systematic literature review	Regard residents’ low-carbon participation as intentional, rational decision-making. Identify economic incentives, perceived benefits, and psychological factors as key drivers of participation willingness; assess policy effectiveness of carbon inclusion schemes.
Qualitative social practice-oriented studies (Social Practice Theory perspective)	Papachristos (2017) [36]; Kloppenburg & Boekelo (2019) [9]; Shove et al. (2012) [11]	Social Practice Theory, socio-technical transition theory, digital platform practice framework	Qualitative case analysis, critical literature synthesis, comparative sociotechnical analysis	Low-carbon action is embedded, routinised social practice shaped by meanings, materials and competences. Digital platforms and infrastructure reshape everyday energy and consumption practices in low-carbon transitions.

. As illustrated in the table, existing empirical research on resident participation in carbon inclusivity mechanisms can be classified into two distinguishable paradigms. The first stream corresponds to quantitative cognitive and incentive-based studies, which align with the rational behavioral logic underpinning the MOA framework. This literature predominantly explores determinants of participation intention but rarely disentangles the intention–behavior gap. It frames low-carbon engagement as individual decision-making, overlooking routine, material and structural constraints embedded in daily life. The second stream is qualitative social practice-oriented studies grounded in social practice theory. These works argue that environmental behavior cannot be simplified to individual attitudes or willingness. However, empirical evidence contextualized within China’s carbon inclusivity schemes remains limited, and few studies integrate social practice theory with cognitive behavioral models such as the MOA framework. Existing empirical work thus falls into two relatively isolated research paradigms, lacking cross-paradigm theoretical integration to reconcile individual decision-making and routine practice dynamics. Against this backdrop, this study combines MOA theory and social practice theory to explain why residents hold positive willingness yet struggle to enact low-carbon behaviors, addressing the above research limitations.

2.4. Research Positioning and Contributions

Existing studies cannot fully explain the CI intention–behavior gap, as they fail to address the dynamic mismatches between individual psychological factors and socio-material practice constraints underlying the CI intention–behavior gap. Against this theoretical backdrop, this study seeks to achieve the systematic coupling of the MOA framework and Social Practice Theory to unpack the intention–behavior gap specifically within CI contexts. Relative to existing integrated theoretical models, the proposed MICPB framework may deliver two distinct incremental contributions. Firstly, it accommodates the unique digital–physical interaction characteristics of CI digital platforms, a feature rarely centralised within established behavior models. Secondly, it embeds a three-stage evolutionary trajectory of social practice into the transformation pathway from intention to behavior. This approach has the potential to move beyond static cross-sectional interpretations and delivers a dynamic, process-oriented perspective to explain sustained low-carbon participation and user disengagement.

This study tries to make the following contributions through addressing the aforementioned gaps: First, it proposes a theoretically coupled integration of MOA and SPT for CI research, bridging the agency–structure divide in environmental behavior analysis. Rather than merely juxtaposing two independent theoretical lenses, this research elaborates bidirectional interaction mechanisms between their constituent dimensions. While MOA effectively explains individual participation thresholds [21], and SPT captures practice institutionalization [11], their synthesis offers novel insights into how motivational factors (e.g., perceived policy incentives, social recognition pursuit) interact with material constraints (e.g., platform interoperability issues, limited low-carbon product availability)—an underexplored nexus in sustainability transitions. SPT’s materials and meanings construct the contextual boundary conditions for individual motivation and opportunity, whereas cumulative individual behaviors shaped by MOA further drive the reconfiguration of socio-material practice elements.

Second, the research advances methodological diversity in CI studies by employing grounded theory analysis of in-depth interviews. This approach responds to calls for “practice-sensitive methods” [37] that move beyond techno-centric metrics to capture the lived experience of energy transitions. By tracing how users articulate their CI engagement experiences, the study seeks to reveal the co-constitution of material infrastructures, cultural meanings, and individual competences—a tripartite dynamic often marginalized in attitude–behavior correlation studies.

Practically, this work contributes to designing more equitable CI systems. Existing platforms often implicitly target tech-savvy urban youth, whereas our focus on heterogeneous user groups (across age and income) identifies barriers and enablers neglected in current implementations. These findings will inform policy frameworks for equitable digital environmental governance, platform designs responsive to heterogeneous user competences and incentive systems synthesizing extrinsic and intrinsic drivers.

This study may provide a transferable model for analyzing digital sustainability tools beyond CI—particularly in Global South contexts where cross-cultural techno-social interactions remain understudied.

3. Methodology

3.1. Data Collection

From January to March 2024, the research team collected in-depth interview data from various CI pilot regions in China, as well as documentary materials and social media information dating back to 2015, when Guangdong Province took the lead in issuing the “Implementation Plan for CI Pilot Work in Guangdong Province” and initiating CI pilot work. These data sources were used to comprehensively analyze consumers’ experiences with CI. To effectively analyze this large-scale dataset, we first carried out relevant preprocessing, which lays a solid foundation for subsequent research. The research team, through word-by-word analysis, identified emerging themes in the dataset, such as motivations, intentions, and actual behaviors related to CI participation, as well as the contextual factors and different stages of dynamic changes in social practices that influence these behaviors. These findings were then discussed among the entire research team. This preliminary analysis not only allowed us to understand the actual operation of the CI mechanism from multiple dimensions, including consumers, government, businesses, and social organizations, but also focused our research on exploring the “intention–behavior” gap among consumers.

To examine consumers’ subjective experiences of participating in CI programs, we adopted the interpretive framework and acknowledged the importance of capturing participants’ subjective meanings and experiences related to such programs. This is because consumers’ subjective feelings and significance attached to green consumption form an integral part of their overall consumption experience. To ensure the comprehensiveness and diversity of feedback, we implemented theoretical sampling [38] through initial purposive selection followed by snowball sampling [39]: starting with a pre-selected target group, we progressively expanded the scope of in-depth interviews. Our interview team ventured into the ecological environment bureaus and eco-themed cafes of carbon-inclusive pilot cities, engaging in face-to-face conversations with interviewees to gain deep insights into their practices and experiences of reducing carbon emissions through choosing green transportation methods (such as public transit, cycling, walking, and using new energy vehicles), participating in green consumption and recycling activities, and converting these emission reductions into carbon credits for exchange or cash-in. Additionally, we examined participants’ frequency of using CI platforms and personal carbon accounts to assess the sustainability of their participation.

Table 2. Basic Information of Interviewees.

	Classification	Number of People	Percentage (%)
Region	Hangzhou	10	23.81
	Shanghai	8	19.05
	Beijing	6	14.29
	Guangzhou	6	14.29
	Wuhan	6	14.29
	Jiaxing	6	14.29
Age	18–25 years old	9	21.43
	26–35 years old	9	21.43
	36–45 years old	10	23.81
	46–60 years old	12	28.57
	61 years old and above	2	4.76
Gender	Male	19	45.24
	Female	23	54.76
Type of interview	In-depth individual interview	18	42.86
	Focus group interview	24	57.14
Type of occupation	University students and graduate students	10	23.81
	Public officials in the education sector and public institutions	10	23.81
	Corporate employees	6	14.29
	Freelancers and individual business owners	5	11.90
	Retirees	5	11.90
	Practitioners in the environmental protection industry	6	14.29
Monthly household income	Less than 5000 yuan	9	21.43
	5000 to 15,000 yuan	15	35.71
	15,000 to 30,000 yuan	10	23.81
	more than 30,000 yuan	8	19.05

presents the demographic information of the interviewees, including gender, age, region, and duration of participation in CI programs.

Over the course of three months of intensive survey, we conducted 22 interview sessions, comprising 18 one-to-one individual interview sessions and 4 focus group sessions, with each session lasting between 60 min and two hours. These sessions involved 42 participants in total: 18 individual interviewees and 24 participants across four focus groups. All interviews were semi-structured to ensure flexibility and depth. The sample size of 22 interviews follows the principle of theoretical saturation [40], as no new core categories emerged in later interviews. While snowball sampling may introduce network homogeneity bias, this risk was mitigated by combining initial purposive sampling with stratified recruitment. We deliberately diversified participants across regions, age, occupation, income and CI participation experience, and avoided over-reliance on a single social network.

Notably, after the first six individual interviews, we innovatively introduced a photo-elicited approach [41], where photographs related to CI activities served as visual cues to stimulate interviewees’ memories and emotional sharing. By integrating the strengths of in-depth interviews, observational methods, and artifact analysis, this approach made the interviews more vivid and intuitive, facilitating a deeper sharing of perspectives, feelings, and experiences. This technique employs photographs as visual stimuli to prompt interviewees to recall and elaborate on in-depth cognitions, emotions and contextual experiences concerning carbon-inclusive participation. It is adopted since visual cues facilitate participants’ reflection on daily practices and material contexts that conventional verbal interviews may fail to fully capture. Its strengths lie in mitigating interviewer-induced bias and uncovering implicit contextual information, whereas the main drawback is possible self-presentation bias when participants select photos. In this research, photographs and matched interview narratives were analyzed jointly. Textual data from photo-elicited interviews were coded using the same grounded theory protocols applied to other interview materials. It should be noted that photo-elicitation functioned merely as an auxiliary prompting technique rather than a fundamental revision to the semi-structured interview protocol. The core interview themes, research questions and analytical framework remained consistent across all interview materials. Visual cues were introduced to facilitate memory retrieval and enrich participants’ narratives without altering the primary lines of inquiry. Hence, the initial six interviews and subsequent interviews are not treated as two distinct datasets requiring separate comparative coding. The outline for the in-depth interviews is presented in

Table 3. Outline for In-Depth Interviews.

Topic	Outline of Contents
Interviewee’s Low-Carbon Cognition	What low-carbon behaviors have you engaged in before participating in the CI program? Do you agree with the low-carbon philosophy of the CI program?
Situation of Participation in CI Programs	When did you start participating in the CI program? What channels did you use to participate? What types of CI activities have you participated in and how frequently? What factors attracted you to participate in the CI activities?
Factors Hindering Continuous Participation in CI Programs	What do you think are the reasons why you can’t consistently participate in the CI program? What difficulties and inconveniences have you encountered during your participation? In your opinion, what aspects of the CI operating mechanism need improvement? What do you think are the reasons why others haven’t participated or don’t often participate in the CI program?

.

At the outset of each interview, we invited interviewees to narrate their ‘stories’ of engaging in carbon-inclusivity initiatives [38], serving as an entry point for the conversation. To mitigate interviewees’ reluctance to address sensitive or personal issues, we adopted projective techniques. As unstructured, indirect probing approaches, these techniques use ambiguous open-ended situational prompts to elicit respondents’ implicit motivations, beliefs, attitudes and emotions. The interviews continued until we reached theoretical saturation [40], ensuring the comprehensiveness and depth of our research findings. Prior to each interview, participants were briefed on audio recording, transcription, follow-up member-checking and the photo-elicited procedure, and informed consent was obtained. All data and photographs were anonymized to protect participant privacy.

To ensure research validity, we implemented a two-phase verification process. ① Proactive measures during data collection: Prolonged engagement (3-month core fieldwork) and real-time peer debriefing with 2 external qualitative researchers; and ② post hoc validation: Member check interviews [42] conducted via WeChat/phone 3 months post-interviews (covering 30% participants) and policy document triangulation for 20 participants for behavioral consistency assessment. This layered approach not only confirmed the accuracy of initial findings but also revealed temporal dynamics in participation behaviors, particularly in bridging the intention–behavior gap. Concretely, triangulation strengthened the internal logical relationships among categories within the MICPB model; member checking refined the conceptual boundaries of multiple subcategories concerning CI participation barriers; and independent coding reduced subjective interpretive bias and stabilized the overall categorical framework supporting the final model.

Through in-depth analysis of data from six initial interviews, combined with existing literature, we refined emerging theoretical constructs and elucidated how consumers construct meaning in their participation in carbon-inclusivity initiatives. These preliminary findings provided guidance for the focusing and optimization of subsequent 12 individual interviews and 4 focus group interviews. Throughout the research process, we accumulated over 200 pages of interview transcripts, which served as invaluable firsthand evidence supporting our deep understanding of the complex mechanisms and subjective experiences of consumer participation in carbon-inclusivity initiatives.

It is recognized that focus group data may be influenced by group dynamics including social desirability bias and dominant participant voices, which differentiates such material from one-to-one interview data. Several measures were adopted to reduce these biases. Moderators encouraged less vocal participants to share independent views and avoided prompting group consensus. During transcript annotation and open coding, researchers distinguished stable personal standpoints from reactive opinions shaped by group interaction. Constant comparison across the two interview formats indicated no substantial divergence in core categories; themes identified in focus groups were consistently supported by individual interview evidence, reinforcing the robustness of the analysis.

3.2. Data Analysis Based on Programmatic Grounded Theory

Straussian Grounded Theory, also known as Programmatic Grounded Theory [38,43], emphasizes research presuppositions and literature orientation. Its coding system encompasses open coding, axial coding, and selective coding, aiming to create theory through a rigorous process. This theory recommends using a skeletal theoretical framework, consisting of a set of relatively concise and fundamental concepts, models, and hypotheses that form the core of disciplinary theoretical systems, to guide the analysis of interview data. This approach differs from Glaserian Grounded Theory [44], which rejects pre-set theoretical frameworks and prioritizes the natural emergence of theory from empirical data, with its analytical coding system limited to substantive and theoretical coding. The Straussian approach places greater emphasis on the application of literature in research, helping to maintain the sensitivity and depth of theoretical construction.

Combining the MOA theory [10] and the theory of Dynamics of Social Practices [11,45], this study employs Straussian Grounded Theory to delve into the practices of consumer participation in CI programs. The analysis of interview data follows the three-level coding method. Firstly, during the open coding phase, the interview team reads the transcripts line by line, marking keywords, phrases, and significant points. By employing the constant comparative method [46], the team continuously compares similarities and differences within the interview data to extract categories and subcategories, ensuring accuracy and depth through iterative comparisons. Secondly, in the axial coding phase, preliminary concepts and categories are synthesized, and their internal relationships are compared to form higher-level categories or subcategories. The coding process produced 76 initial concepts (e.g., “Obtain social recognition”), 20 corresponding sub-categories (e.g., “Intrinsic motivation”) and 6 main categories (“Motivation for pro-environmental behavior”) (See

Table 4. The categorization of open coding.

Category	Original Statements (Initial Concepts)
Intrinsic motivation	A08 Through participating in CI programs, I feel a sense of responsibility as a global citizen, which fills me with pride. (Environmental responsibility) A16 Participating in CI programs turns low-carbon behaviors into visible carbon credits, giving me more motivation to persist. (Expected benefits) A24 Since participating in the CI program, I have received approving looks from my friends around me. (Obtain social recognition) A30 Participating in CI programs highlights a green lifestyle and reflects a certain attitude and way of life. (highlights a green lifestyle)
Extrinsic motivation	A08 Subsidies and tax exemptions are significant motivations for my participation in CI programs, as they provide tangible benefits when I purchase energy-efficient products and adopt low-carbon transportation. (The role of policy incentives) A12 The commercial incentives offered by some companies, such as reward points and discounts, have to some extent stimulated my enthusiasm for participation. (The role of commercial incentives) A17 Currently, major brands are introducing various types of electric vehicle models that cater to individualized needs. Moreover, with the reduction in battery costs and the emergence of scale effects, vehicle prices have become more reasonable. (Supply and pricing in the low-carbon market) A20 In my community, low-carbon living has become a widely advocated behavior, and this atmosphere has prompted me to embrace an environmentally friendly lifestyle. (Social norms) A29 In my social circle, low-carbon living has become a pressure that motivates me to actively participate in emission reduction. (Group pressure)
Low-carbon knowledge and skills	A22 I am willing to purchase low-carbon products, but they do not have an advantage in terms of performance and quality, and the prices are also high. It’s hard for me to make a decision to buy them, especially since I just graduated and am on a tight budget. (Understanding of low-carbon products) A35 Although I am personally inclined to purchase eco-friendly products, when making a decision, I still need to take my family’s opinions into account. Both my wife and I are environmentally conscious, but since our child is still very young, we tend to lean towards products that are both affordable and of good quality. (Ability to make low-carbon decisions) A10 I usually share energy-saving tips or recommend low-carbon products to my friends. (Ability to disseminate low-carbon information) A26 As I’ve grown older, it’s not easy for me to learn new things. For example, with the new smart thermostat and energy-saving light bulbs installed at home, I just can’t remember how to operate them in the most energy-efficient way. (Low-carbon living skills)
Resource and time management skills	A05 I really want to participate in CI programs, but with the elderly in my family needing medical care and my children needing to go to school, there are expenses everywhere. Every time I think about buying environmentally friendly products, I have to give up when I consider our household expenses. (Ability to control economic costs) A21 I’m too busy with work. Things like garbage classification and energy conservation often slip my mind when I’m caught up in work. I want to participate in some low-carbon activities on weekends, but they often get canceled due to work commitments. I feel powerless. (Time management skills) A26 I’m not proficient with smartphones and I’m not very good at using social media. I want to share my experiences of participating in low-carbon activities, but I don’t know how to do it. This has gradually diminished my enthusiasm for participating in CI initiatives, and I feel excluded. (Efficiency of information acquisition and dissemination)
Learning and adaptability skills	A02 When I can easily access and understand relevant low-carbon knowledge, I am more motivated to practice low-carbon behaviors and actively participate in CI initiatives. (Ability to acquire low-carbon knowledge) A25 Of course, relevant skills will affect participation enthusiasm! For example, when I register and log in through a CI platform, and use a carbon footprint calculator to track my low-carbon behaviors, I can better appreciate the joy of participating in CI initiatives. (Skills for participating in CI initiatives) A18 When I successfully changed my old, environmentally unfriendly habits, such as choosing to walk or bike instead of driving, I deeply felt the satisfaction of contributing to the Earth. (Ability to change consumption habits and behavioral patterns)
Positive anticipated emotions	A10 Every time I see my achievements, such as the amount of carbon reduced and the accumulated carbon credits, I feel proud and satisfied. (Sense of achievement and pride in environmental protection) A15 I am actually quite eagerly anticipating the opportunity to exchange carbon credits for goods or services, as well as cash rewards. (Expectation of obtaining tangible economic benefits) A30 Through participating in CI programs, I have found that I can gain a greater sense of identity and belonging within my social circle. (Social identity and sense of belonging)
Negative anticipated emotions	A19 I realized that I don’t know much about CI, and this lack of understanding makes it difficult for me to actively participate. (Lack of basic understanding/knowledge) A04 I hold reservations about the actual effectiveness of CI programs, and this sense of distrust gives me pause, as I worry that participating may not contribute to the environment. (Sense of distrust towards the CI mechanism) A23 Although I am aware of the benefits of participating, when considering the potential economic costs, such as purchasing high-priced low-carbon products, I still hesitate. (Concerns about the economic costs of participating in CI programs) A27 I realize that my lifestyle and consumption patterns have already been formed, and it will require a great effort to change them. (Inertia in lifestyle and consumption patterns)
Emotions related to personal values	A09 Using high-end, branded products better reflects my taste, even if they may not be as environmentally friendly. This makes it somewhat difficult for me to choose between them and low-carbon products. (Conspicuous consumption values) A24 I have a special emotional attachment to certain brands, and even though I know there are more environmentally friendly options available, that emotional bond makes it difficult for me to let go. This sometimes leaves me feeling conflicted. (Emotional consumption values) A33 I have always believed that low-carbon consumption requires sacrificing some aspects of quality of life in pursuit of environmental protection. This makes me feel that participating in CI means leading a less comfortable life, which is why I am somewhat resistant. (Cognitive bias towards low-carbon consumption)
Policy guidance and support	A01 I feel that participating in CI programs allows me to enjoy the tangible benefits brought by top-level policy support, which further motivates me to actively engage in various low-carbon behaviors. (Top-level policy guidance) A17 I believe that the standard system of CI has a dual impact on me. On the one hand, it enables me to clearly understand my carbon reduction contributions; on the other hand, the standards can sometimes be overly complex or inconsistent across different platforms, which leaves me somewhat confused. (CI standard system) A33 I have found that the current forms and magnitude of policy incentives are insufficient. For instance, the reward amounts are relatively small, or the thresholds for receiving benefits are quite high, which leaves me feeling unmotivated to persist in low-carbon behaviors. (The form or intensity of policy incentives) A28 The process of redeeming rewards is rather cumbersome, and the waiting time is also relatively long, which causes me some inconvenience. (The fulfillment of carbon credit rewards)
Construction and Technical Support of CI Platforms	A20 I have found that the functionality of CI platforms is not yet perfect. For example, the calculation method for carbon emissions reduction is not transparent enough, and the variety and quantity of reward options for exchange are relatively limited, which has gradually diminished my enthusiasm for participation. (Construction of CI Platforms) A13 I have found that the digital and intelligent technology support related to CI still needs improvement. For instance, delays or errors sometimes occur during the data collection and processing process, which affects the accurate measurement of emission reductions. (Related digital and intelligent technology support) A09 Some issues existing in the process of CI data standardization and mutual recognition, such as complex operations and lack of transparency in carbon emission measurements, can somewhat diminish my enthusiasm for participation. (Data standardization and mutual recognition) A35 Due to data exchange barriers and privacy concerns, collaboration among multiple departments is difficult, resulting in delays in receiving water, electricity, and gas data on the platform. The delay feedback from the platform will undoubtedly affect the user experience and participation enthusiasm. (Immediate feedback from the platform) A25 To my knowledge, most platforms basically adopt a single interaction mode of “generating carbon credits through low-carbon behaviors and exchanging them for goods.” This model lacks innovation and diversity, making it difficult to retain users in the long term. (Interactivity of the platform)
Supply and demand of low-carbon market	A07 The low-carbon products available in the market are quite diversified, such as electric vehicles, energy-efficient appliances, and bamboo furniture. I actively purchase them when economically feasible, as they are beneficial to health after all. (Types and quantities of low-carbon products) A14 Despite the improving quality of electric vehicles nowadays, some brands still face issues such as inadequate range, incompatible charging infrastructure, and insufficient after-sales service. For instance, it is reported that the actual range of some models of WM Motor’s vehicles is significantly lower than the official data. (Quality of low-carbon products) A27 I am willing to contribute to environmental protection, but my wallet is tight. The prices of low-carbon products on the market are still relatively higher compared to traditional products, and my purchasing power cannot keep up. (Purchasing power for low-carbon products) A14 I have had some experience with low-carbon products, such as biodegradable tableware and recyclable clothing. While I find these products environmentally friendly, to be honest, they are not as convenient and comfortable to use as traditional products. (Comfort and convenience of low-carbon products)
Social atmosphere	A19 I have observed that my friends around me are all practicing a low-carbon lifestyle, such as using eco-friendly shopping bags, cycling, or walking. I feel that I should also follow suit. (Social norms) A04 I often see public figures sharing their low-carbon lifestyles on social media, and these exemplary demonstrations make me feel that participating in CI initiatives is not only a responsibility but also a fashion trend. (The exemplary demonstration effect) A28 Although I frequently encounter promotions of CI initiatives on social media and in the news, these messages are often fragmented and one-sided, coming solely from official sources, and lacking interaction. (Social dissemination) A32 Although schools and some communities have carried out environmental education, such education often remains superficial, lacking genuine experiential and participatory elements. (Public education) A16 When I share my low-carbon lifestyle on social media, I always receive many likes, which makes me feel proud. (Face culture)
The stage of meaning construction	A22 I currently have only a preliminary understanding of CI initiatives and have not yet deeply recognized their profound significance for both individuals and society. (Understanding of the concept of carbon reduction initiatives) A32 Currently, I primarily opt for green modes of transportation such as walking, cycling, taking the bus, using the subway, and driving an electric vehicle to participate in CI initiatives. (Selection of CI Practice Strategies) A23 My understanding and recognition of the significance of “CI” still need to be enhanced. That is to say, my comprehension of the specific value of CI and its role in individual lives and at the societal level is not yet profound. (Recognition of the Value of CI) A22 Although I was initially very interested in CI, believing that it allowed me to contribute to environmental protection while earning rewards, upon actual participation, I found that these initial motivations alone were not sufficient to sustain my commitment. (Initial motivations for participating in CI)
The stage of adaptation	A19 I would love to participate in CI, but it’s really challenging. For instance, if I choose to take the bus instead of driving, it means I have to wake up early and spend a lot of time waiting for the bus and transferring, which requires considerable determination. (Behavioral change) A05 Every time I’m squeezed and sweating in a crowded subway or bus, I feel irritated and uneasy, and I can’t help but wonder, ‘Why am I putting myself through this?’ (Emotional distress) A24 Our CI platform has integrated over 20 emission reduction scenarios, and it can record low-carbon behaviors across different domains and award points, which can then be exchanged for coupons and other rewards. However, I find the entire process quite cumbersome. Consumers who are not familiar with the platform’s operation may have to undergo a relatively long adaptation process. (Habit formation) A28 At a promotional event organized by the environmental protection agency, I discovered that there are many other ways to accumulate points more quickly besides energy conservation at home, such as purchasing an electric vehicle, taking public transportation, and sorting garbage. (Feedback and Strategy Adjustment in CI Practices)
The stage of stabilizing	A09 Since the launch of the “Three Jin Green Life” mini-program, it has integrated multiple emission reduction scenarios and attracted a lot of participants. Overall, the platform is relatively stable and efficient. For example, taking public transportation and reducing the use of single-use plastic products can earn points on one’s personal carbon account, which can then be redeemed for rewards. (Stability of the CI Mechanism Platform) A14 The CI platform not only provides convenient channels for participation, but more importantly, it conveys a low-carbon lifestyle concept. This helps to create a strong low-carbon cultural atmosphere. (The entrenchment and stability of a low-carbon social culture) A25 Currently, the platform leverages cloud computing, big data, and other technologies to precisely quantify low-carbon behaviors. With just a simple registration, users can now accrue carbon credits by walking, using shared bicycles, driving electric vehicles, and participating in waste sorting, instantly redeemable for rewards. (The maturity of digital technology applications) A30 I feel that traditional reward points don’t motivate me much. It would be great if they could introduce loans or insurance based on carbon credits. (Innovation in CI Mechanisms) A22 As time went by, I began to promote the benefits of low-carbon electricity usage in my social circle and shared my carbon reduction achievements. My friends were also inspired by me. (Individual sharing and dissemination of low-carbon practices)
Age and Gender	A04 I think young people generally have a higher acceptance of new technologies. They are more likely to understand and operate mini-programs, and can participate more conveniently. (Age) A12 I feel that women tend to pay more attention to participating in CI initiatives through low-carbon lifestyles. Generally, women are more meticulous and disciplined, and they can persist in taking public transportation, reducing the use of disposable products, and so on. (Gender) A22 Men may be more inclined to participate in CI programs by purchasing low-carbon products. (Gender)
Occupation and income	A13 A friend of mine works in IT. Since their company launched a carbon account system, he has not only been recording and sharing low-carbon behaviors but also, through data analysis, discovered that commuting is the main source of employees’ carbon emissions. As a result, he suggested to his leaders the introduction of a remote work policy, which effectively reduced employees’ carbon footprint. (Professional background) A26 I have a high income, but I’m very busy. I found that purchasing high-end green products is also a way to participate in CI program, so I decisively bought an electric vehicle. (Income) A09 My income is not high, but I am willing to do my best for environmental protection. I start from my daily life, such as walking or cycling to and from work every day, turning off lights and saving water and electricity at home. (Income)
Educational background	A12 I feel that consumers with higher educational backgrounds are more likely to accept new things and put them into practice. (Educational background) A24 A high school classmate of mine, who holds a PhD in Environmental Science from the United States, initiated a “Challenge for a Green Lifestyle” competition in his community after returning to China. He guided residents to reduce their carbon footprint by walking, cycling, or taking public transportation, and offered both spiritual and material rewards. (Overseas educational background) A24 I have a buddy who works on environmental technology research and development. Not only did he purchase an electric vehicle, but he also developed a mobile app that records low-carbon behaviors in real-time and quickly converts them into carbon credits. (Professional background)
Regional differences	A17 In big cities like Shanghai, our community provides smart recycling bins that are not only convenient but also have a transparent reward system, thus effectively promoting waste sorting. (Convenient low-carbon living facilities and transparent reward mechanisms in big cities) A22 Our CI program across the province is doing quite well. We have a clear reward mechanism, incorporate diverse emission reduction scenarios, and can precisely quantify low-carbon behaviors while issuing carbon credits. What’s more impressive is that we can even generate officially certified “emission reduction certificates”. (Diverse emission reduction scenarios and official certification)
Family structure and internal interactions	A01 I myself am taking a wait-and-see attitude, but my wife and children are really enjoying the CI program. They record their points through a mini-program and discuss how to exchange them for desired items. Over time, I’ve been influenced by them. (Interactions within the family) A13 We both care about environmental protection, but it’s really hard for us to stick to it. Mainly because we’re too busy with work, our child is still in kindergarten, and we don’t have enough time and energy. Besides, we’re also under quite a lot of financial pressure. (Limitations of young nuclear families)
Behavioral intention	A18 After our mini-program was launched, many families downloaded and registered, earning points by recording green behaviors. However, as time went by, some residents found it difficult to keep up. Some were too busy with work, while others might have forgotten, found it troublesome, or deemed the points for certain behaviors too low to bother recording. (The contradiction between intention and behavior) A31 I strongly agree with the idea of low-carbon living, but it’s really hard to stick to it. Even though I know that driving to work increases carbon emissions, I can’t help it sometimes. For instance, when the company calls for an emergency meeting, I can’t just take the bus or transfer through multiple subway lines. (The contradiction between intention and behavior) A29 I am willing to prioritize purchasing green products; however, in actual decision-making, I often end up choosing regular products with higher cost-effectiveness due to reasons such as price, product performance, or lack of green certification. (Imbalance between intention and behavior)

Note: Codes such as A1, A08, A16, and A30 mean the original sentence of the respondent. The words in parentheses at the end of each sentence denote the initial concept obtained by encoding the original sentence.

and

Table 5. The process and results of axial coding.

Main Categories	The Corresponding Sub-Categories	The Connotation of Relationship
Motivation for pro-environmental behavior	Intrinsic motivation	It refers to the intrinsic needs that drive consumers to participate in carbon reduction due to their sense of environmental responsibility, expected benefits, pursuit of social recognition, and desire to demonstrate a green lifestyle, which influence their motivation for pro-environmental behavior. It refers to the external driving force for participating in carbon reduction, which is promoted by policy incentives and commercial incentives, influenced by the supply and prices in the low-carbon market, and shaped by social norms. This external driving force affects individuals’ motivation for pro-environmental behavior.
	Extrinsic motivation
Pro-environmental behavioral ability	Low-carbon knowledge and skills	It refers to the sum of consumers’ understanding of CI, their level of low-carbon awareness, and the abilities and skills demonstrated when making low-carbon consumption decisions and practicing a low-carbon lifestyle, which affect their pro-environmental behavioral ability. It refers to the comprehensive ability demonstrated by consumers in terms of economic cost control, time planning, and efficient information acquisition and dissemination, which influences their pro-environmental behavioral ability. It refers to the comprehensive ability demonstrated by consumers in acquiring low-carbon knowledge, mastering skills related to participating in CI, changing behavioral patterns, and embracing new technologies, which influences their pro-environmental behavioral ability.
	Resource and time management skills
	Learning and adaptability skills
Consumer emotions	Positive anticipated emotions	It refers to the sense of environmental achievement, economic benefit expectations, social recognition, and sense of belonging that consumers gain from participating in environmentally friendly actions, which constitutes a dimension of consumer emotions. It refers to consumers’ lack of awareness of CI, compounded by distrust in their effectiveness, concerns about economic costs, and inertia in changing established consumption patterns, which constitutes a dimension of consumer emotions. It refers to the intrinsic factors influencing consumer emotions, stemming from conspicuous consumption values, emotional consumption values, and biases in low-carbon cognition. These values jointly act on the consumer decision-making process, constituting a dimension of consumer emotions.
	Negative anticipated emotions
	Personal value emotions
External opportunities and conditions	Policy guidance and support	It encompasses the establishment of a CI standard system, specific forms and levels of rewards, as well as the efficiency of reward fulfillment, which constitutes one of the external opportunities and conditions.
	CI platform construction and technical support	It refers to the construction of a standardized CI platform, utilizing digital and intelligent technology to support data processing and standardization, achieving cross-platform data recognition and real-time feedback, and enhancing platform interactivity, which constitutes one of the external opportunities and conditions.
	Supply and demand in the low-carbon market	It refers to the variety, quantity sufficiency, quality reliability of low-carbon products in the market, as well as the purchasing power and feedback evaluations of these products, which constitutes one of the external opportunities and conditions.
	Social atmosphere	It refers to the external cultural orientation that promotes participation in CI, shaped by social norms, group pressure, the leading role of public figures, social dissemination, public education, and face culture. This constitutes one of the external opportunities and conditions.
Dynamic changes in social practice	The stage of meaning construction	It refers to the dynamic cognitive process in which consumers, through their recognition of the value of CI, form initial motivations to participate and choose suitable CI practice strategies. This constitutes the first stage of dynamic changes in social practice. It refers to the second process of dynamic changes in social practice, where consumers undergo changes in behavioral patterns, gradually develop environmentally friendly habits, face potential emotional challenges, and continuously adjust their strategies through practical feedback to long-term adapt and sustainably reduce carbon emissions during the process of practicing CI. It refers to the third process of dynamic changes in social practice, where consumers gradually integrate into the CI system and share their participation achievements, the CI mechanism platform gradually stabilizes, governments and enterprises introduce innovative incentive mechanisms, and a low-carbon cultural atmosphere gradually forms across the entire society.
	The stage of adaptation
	The stage of stabilizing
Consumer characteristics	Age and Gender	It refers to important demographic characteristics, which are also an important component of consumer characteristics. It refers to important demographic characteristics, which are also an important component of consumer characteristics. It refers to important demographic characteristics, which are also an important component of consumer characteristics. It refers to important demographic characteristics, which are also an important component of consumer characteristics. It refers to important family characteristics, which are also an important component of consumer characteristics.
	Occupation and income
	Educational background
	Regional differences
	Family structure and internal interactions

). Through ongoing validation, their representativeness and stability are ensured, facilitating the generation of new categories or refining existing ones [38,43].

Lastly, during the selective coding phase, core categories are selected from the coded categories and linked to subordinate categories based on information such as causal conditions, context, strategies, and consequences. Furthermore, based on the attributes and dimensions of each category, we systematically and organically link these categories together, validate the relationships among them, and continue to develop the categories until they reach a refined and comprehensive state. This process ultimately aims to construct a theoretical framework with explanatory and predictive power [38]. To ensure research quality, member-checking interviews were conducted three months after the in-depth interviews to verify the accuracy of the thematic categories obtained from the previous interviews and the validity of the interview conclusions [47].

In addition, based on in-depth interviews, the research team integrated various methods such as observation and secondary data analysis to collect data related to the interview themes. A text database was established, and the triangulation method [48] was employed to cross-validate the multi-source data, ensuring the consistency and reliability of the coding analysis and thus safeguarding the validity and accuracy of the conclusions.

All documentary materials and social media texts from 2015 to 2024 were incorporated into the unified grounded theory coding workflow rather than analyzed via separate thematic analysis. The same open, axial and selective coding protocols were applied to both interview transcripts and auxiliary texts. Following sequential coding procedures, theoretical saturation testing will be conducted after category formation.

Following the COREQ guidelines [49], the research team engaged in reflexive consideration of how researchers’ positional backgrounds and pre-existing assumptions could shape data analysis and interpretation. All interviewers and analysts possess academic backgrounds in sustainability, environmental behavior and consumer research, which inevitably informed the reading of interview narratives. To avoid imposing predetermined interpretations onto participants’ lived experiences, the team maintained ongoing reflexive awareness throughout coding. No prior personal or institutional relationships existed between researchers and interviewees. Tentative codes and emergent themes were discussed collectively within the team, and interpretations were further scrutinized via peer debriefing with two independent external qualitative researchers. To mitigate the risk that researchers’ prior familiarity with MOA and SPT might introduce analytical bias during coding, we implemented sensitivity re-coding and inter-coder reliability assessments, detailed subsequently in Section 4.6. These checks confirm that core categories emerge from empirical data rather than being artificially fitted to pre-established theoretical frameworks.

4. The Coding Analysis and Model Construction Based on the Programmatic Grounded Theory

After systematically organizing the interview recordings and verbatim transcripts, this study accumulated over 200,000 words of textual data generated from 22 interview sessions involving 42 participants. All participant utterances extracted from transcripts served as the basic unit for open coding. To conduct theoretical saturation testing, we set aside textual materials from a subset of participants and reserved these extracts for validation after initial model construction. Throughout the grounded coding analysis, we strictly followed the procedures of programmatic grounded theory [38,43]. We adopted a constant comparative strategy and continuously refined the theoretical framework until theoretical saturation was achieved, whereby newly examined participant extracts no longer yielded novel conceptual insights. In summary, as the core component of the research, this process focused on constructing a comprehensive theoretical model that fully elucidates the formation mechanism of the “intention–behavior” gap in consumer participation in CI programs, through standardized programmatic grounded coding and theoretical saturation testing.

4.1. Open Coding

During the open coding (known as the first-level coding), we conducted a word-by-word analysis of the raw data and assigned corresponding conceptual labels. Through continuous comparison and analysis of content within and across datasets, we systematically logged and refined initial concepts, subsequently identifying conceptual categories [38,43]. To minimize the interference of subjective preferences in the assignment of conceptual labels, this study directly named or extracted original concepts from the interviewees’ raw statements. In order to delve into the key factors influencing the “intention–behavior” gap in consumer participation in CI programs, we rigorously screened the raw data, eliminating sentences with low information content, vague expressions, or oversimplification. Following this process, the study collected and identified over 1000 original statements and their associated initial concepts. Given the large number of initial concepts and their overlaps, we employed a frequency screening method, retaining concepts that appeared at least three times while excluding those that appeared less than three times or were inconsistent. Table 4 details the entire process from initial conceptualization to categorization of the raw data. The final result of categorization reveals the core elements influencing consumer participation in CI programs. For brevity, this study selects only typical original statements and their corresponding initial concepts for each category.

4.2. Axial Coding

Axial coding, also known as secondary coding, aims to distill core concepts and categories from a large amount of data. Through integration and induction, it constructs a concise and comprehensive conceptual framework to elucidate the core content of the data. Specifically, axial coding establishes intrinsic links between categories and clusters, and integrates similar or related categories, thereby forming higher-level main categories. Meanwhile, based on the frequency of occurrence, coverage, and explanatory power of the categories in the data, categories are divided into main categories and sub-categories. Through repeated comparison and induction, this study identified six main categories underlying the intention–behavior gap. Among them, five categories constitute the core mechanistic dimensions of the theoretical model, while consumer characteristics serve as demographic moderating variables that explain heterogeneous behavioral responses rather than core causal components. The process of axial coding, which is also the formation process of the main categories, is shown in Table 5.

4.3. Selective Coding

Selective coding is a process based on axial coding that aims to refine a core category from a myriad of conceptual categories. Through deep analysis and integration, it establishes a theoretical framework with overarching and profound explanatory power. The core of selective coding lies in untangling the intrinsic connections between the core category and other categories, outlining the essential characteristics and logical thread of the research object with a clear “storyline”.

In this study, we focus on the “factors and mechanisms influencing the ‘intention–behavior’ gap in consumer participation in CI,” setting this as the core category. Revolving around this core, we have distilled six major categories as the pillars of the “storyline”: pro-environmental behavioral motivation, pro-environmental behavioral ability, consumer emotions, external opportunities and conditions, dynamic changes in social practices, and consumer characteristics. These six major categories are the primary factors explaining the “intention–behavior” gap in consumer participation in CI, but their pathways and mechanisms of action vary. Among the six categories, five constitute the core mechanistic dimensions supporting the sequential theoretical pathway of the MICPB model. The sixth category, consumer characteristics, functions as exogenous demographic moderators that capture inter-individual heterogeneity. It explains divergent responses to identical contextual conditions yet is not embedded within the core causal chain linking intention and sustained participation behavior.

Specifically, pro-environmental behavioral motivation indirectly influences actual participation in CI through the intention to participate, revealing the existence of an “intention–behavior” gap. Serving as internal contextual factors, consumer emotions and pro-environmental behavioral ability help explain the “intention–behavior” gap in CI participation; acting as external contextual factors, dynamic changes in social practices and external opportunities and conditions can further elucidate this gap. Based on this elaborate “storyline,” we have constructed and developed an innovative theoretical framework: the “Model of Influential Factors and Formation Mechanisms of the Intention–Behavior Gap in Consumer Participation in CI.” Through in-depth analysis, the typical relational structures among the major categories have been clearly delineated (see

Table 6. The result of selective coding.

Core Category	Typical Relational Structure
Factors and mechanisms influencing the ‘intention–behavior’ gap in consumer participation in CI	Pro-environmental behavioral motivation → Participation intention in CI	The transformation of consumers’ pro-environmental behavioral motivation into their intention to participate in CI is a process of internal value identification. Consumers’ pro-environmental behavioral motivation directly determines their participation intention in CI.
	Participation intention in CI → Participation behavior in CI	The participation intention for CI is the internal factor driving participation behavior in CI. The participate intention in CI directly determines consumers’ participation behavior.
	Pro-environmental behavioral ability Participation intention in CI–Participation behavior in CI	As an internal contextual factor explaining the “intention–behavior” gap in consumers’ participation in CI, pro-environmental behavioral ability influences the strength and direction of “intention–behavior” relationship in CI participation.
	Consumer emotions Participation intention in CI–Participation behavior in CI	As an internal contextual factor explaining the “intention–behavior” gap in consumers’ participation in CI, consumer emotions influence the strength and direction of “intention–behavior” relationship in CI participation.
	Dynamic changes in social practice Participation intention in CI–Participation behavior in CI	As an external contextual factor explaining the “intention–behavior” gap in consumers’ participation in CI, dynamic changes in social practice influence the strength and direction of “intention–behavior” relationship in CI participation.
	External opportunities and conditions Participation intention in CI–Participation behavior in CI	As an external contextual factor explaining the “intention–behavior” gap in consumers’ participation in CI, external opportunities and conditions influence the strength and direction of “intention–behavior” relationship in CI participation.

), providing a solid theoretical foundation for understanding and promoting consumer participation in CI.

4.4. Theoretical Saturation Test

Theoretical saturation was assessed to confirm the comprehensive development of our model, indicating no new conceptual elements would emerge from additional data [40].

Initial categories were developed using data from the first 28 participants. We then sequentially coded the remaining 14 reserved interview transcripts for independent saturation verification. Minor descriptive details within existing categories were observed in the first two reserved transcripts, yet no new main categories or distinct subcategories emerged starting from the 3rd transcript.

Supplementary stability checks were further implemented. We examined two distinctive cases: participants aged 68 and 72 with monthly incomes below ¥3000. Their behavioral patterns consistently aligned with the established categories without introducing new conceptual dimensions. These heterogeneous cases are used merely to test the stability of the coding framework, rather than to judge theoretical saturation.

The comprehensive analysis confirms three key findings: the six primary dimensions demonstrate conceptual completeness, all subcategories exhibit stable relational properties (See Table 6), and the model shows robust explanatory power across diverse behavioral patterns. These results collectively support our conclusion that theoretical saturation was attained for the intention–behavior gap framework, ensuring the model comprehensively captures the studied phenomenon.

4.5. The Construction of the Comprehensive Theoretical Model

After completing the processes of open coding, axial coding, and selective coding, this study identified the core category as well as the main categories of “pro-environmental behavioral motivation, consumer emotions, pro-environmental behavioral ability, external opportunities and conditions, dynamic changes in social practices, and consumer characteristics.” The relational structures of each main category were also delineated. All main categories and their relationships are induced from interviewees’ original quotations listed in Table 4. The MICPB model’s structure emerges firstly from these empirical categories, and subsequent theoretical integration serves to interpret the data rather than predetermine the analytical outcomes. Building on this foundation and integrating the Dynamic Theory of Social Practice and the MOA model, the MICPB model theoretically integrates the Motivation–Opportunity–Ability (MOA) framework with Social Practice Theory (SPT) through three synergistic linkages: (1) MOA’s individual-level motivational drivers (e.g., environmental responsibility in intrinsic motivation or policy incentives in extrinsic motivation) and capability assessments map onto the intention-formation phase, while (2) SPT’s tripartite practice elements (materials, competences, meanings) explain behavioral contextualization in the transition from intention to action. This synthesis is empirically grounded in our coding results, particularly the axial coding relationships between policy incentives (MOA’s opportunity) and value recognition in the meaning construction stage (SPT’s practice evolution). Along with a review of relevant literature, this study constructed a theoretical model (MICPB model) illustrating the “The formation mechanism of the ‘intention–behavior’ gap in consumer participation in CI,” as shown in Figure 1.

Within this MICPB model, motivation, ability and emotions constitute the core individual-level mechanisms shaping the intention–behavior relationship, while external opportunities and evolving social practice dynamics operate as critical contextual conditions. Consumer characteristics function as moderating factors that generate heterogeneous responses but lie outside the primary causal pathway. The intention–behavior gap arises fundamentally from mismatches between individual internal states and the constraints embedded within socio-material practice environments.

4.6. Sensitivity Test of Coding Results

To examine whether the pre-specified MOA and SPT frameworks unduly shaped emergent categories, a sensitivity check adopting Glaserian-style independent re-coding was conducted following the primary analysis. Approximately 20% of interview transcripts were randomly selected. The coder carried out substantive coding without referencing the MOA and SPT frameworks or the coding outcomes generated in the main analysis. The comparison of coding results reveals that nearly all six main categories derived from the original Straussian grounded theory analysis reappeared in the independent re-coding, though there are minor divergences in several subcategories. This outcome indicates that core categories originate primarily from empirical data. The predefined theoretical frameworks primarily serve to interpret themes emerging from data, instead of imposing artificial structures on raw interview material.

Inter-coder reliability was assessed via blind independent coding of 20% of randomly selected interview transcripts by an external researcher. Coding outcomes generated by this independent researcher were compared with the primary coding results from the research team. To overcome the limitation of raw percentage agreement, Cohen’s Kappa was adopted to account for chance agreement. The analysis yields an observed agreement of 92% and a Cohen’s Kappa of 0.84, indicating almost perfect inter-coder reliability. All coding discrepancies were systematically discussed and reconciled, and the refined coding criteria were applied to the full dataset.

5. Interpretation of the Comprehensive Theoretical Model

Through grounded analysis, the above constructed theoretical model of the formation mechanism of the “intention–behavior” gap in consumer participation in CI effectively elucidates the phenomenon of “knowing is easier than acting” in consumer participation in CI. Specifically, the seven main categories of pro-environmental behavioral motivation, participation intention, pro-environmental behavioral ability, consumer emotions, external opportunities and conditions, dynamic changes in social practices, and consumer characteristics are the primary factors influencing consumer participation in CI. However, the influence paths and mechanisms of each main category are not consistent.

5.1. Pro-Environmental Behavioral Motivation Indirectly Influences Actual Participation Behavior Through the Intention to Participate in CI

In the theoretical model of this study, consumers’ pro-environmental behavioral motivation serves as the intrinsic driver for their participation intention in CI. According to TPB, consumers’ pro-environmental behavioral motivation will first influence their behavioral intention, which in turn will affect their actual behavior. In other words, pro-environmental behavioral motivation indirectly influences consumers’ actual behavior through the mediation role of behavioral intention. This supports the viewpoint of this study: consumers’ pro-environmental behavioral motivation indirectly influences their participation behavior in CI through their participation intention. This mediation pathway aligns with MOA’s proposition that motivation operates through intention formation [10], while our grounded data reveal its contingency on SPT’s material competence dimension (e.g., A26’s smartphone literacy barriers). This finding extends static MOA reasoning by showing that motivation–intention linkage cannot be isolated from the material and practical context highlighted by practice theory.

5.2. There Exists an “Intention–Behavior” Gap in CI Participation

The TPB regards behavioral intention as the direct driving force for participation in CI programs. However, reality often deviates from theoretical presumptions: even if consumers have the intention to participate, they may not necessarily adopt actual behavior. This phenomenon reveals that there is not always a seamless connection between “intention” and “behavior,” but rather a gap exists [50]. Relevant studies have shown that there is a weak correlation between pro-environmental intentions and actual behaviors [51], as multiple contextual factors moderate the relationship between the two. When pro-environmental intentions conflict with the context, the predictive power of intentions on behaviors is significantly weakened; conversely, if they align, the predictive power of intentions is strongest [52]. Feedback from interviewees’ in-depth interviews vividly demonstrated this gap, as exemplified by one interviewee saying, “A31 I strongly agree with the idea of low-carbon living, but it’s really hard to stick to it. Even though I know that driving to work increases carbon emissions, I can’t help it sometimes. When the company holds an emergency meeting, I can’t just take the bus or transfer through multiple subway lines, can I?”

Further analysis of the interview data reveals that the “intention–behavior” gap manifests primarily in two aspects: First, genuine intentions are constrained by internal and external contexts and fail to translate into behavior, as exemplified by A31’s case of having to drive due to an emergency meeting at work. Second, contradictions within the behavioral intentions themselves lead consumers to seek excuses or rationalizations for their inability to consistently participate in CI programs, thereby creating the “intention–behavior” gap. For instance, “A18 After our mini-program was launched, many households downloaded and registered to earn points by recording green behaviors. However, as time went on, some residents found it difficult to persist. Some were too busy with work, while others might have forgotten, found it troublesome, or deemed the points for certain behaviors too low to bother recording.” In summary, due to the influence of internal and external contexts and contradictions within behavioral intentions, consumers’ participation intentions in CI programs are restricted in their conversion to actual behaviors, resulting in the gap. This finding not only enriches the application context of the TPB but also provides insights for promoting the effective implementation of CI policies: it is necessary to focus on and optimize the contextual conditions that affect behavior realization, while also guiding consumers to form stable pro-environmental intentions, in order to bridge the “intention–behavior” gap and effectively promote the implementation of CI programs. It further helps explain why intention exhibits weak predictive power in many field studies, by unpacking heterogeneous contextual constraints frequently overlooked in generic TPB frameworks.

5.3. Consumer Emotions and Pro-Environmental Behavioral Abilities Are the Internal Contextual Factors That Elucidate the “Intention–Behavior” Gap in CI Participation

Existing studies have focused on the direct effect of consumer emotions on pro-environmental behavior. However, research that examines emotions as an internal contextual factor and explores their moderating role in the “intention–behavior” gap in CI participation remains scarce. Consumers’ perception of their own pro-environmental behavioral abilities serves as a crucial bridge for translating intentions into behaviors [53]. In the context of consumer participation in CI programs, pro-environmental behavioral abilities can be broken down into three dimensions: low-carbon knowledge and skills, resource and time management abilities, and learning and adaptation abilities [54].

Through grounded research, this study dissects the dimensions of consumer emotions and pro-environmental behavioral abilities and their impacts. Consumer emotions, encompassing positive anticipated emotions, negative anticipated emotions, and personal value emotions, exert a significant influence on pro-environmental behaviors such as CI participation [55]. Positive anticipated emotions, such as a sense of environmental achievement, social identification, and anticipation of economic benefits, can stimulate consumers’ continued participation in CI programs. Conversely, negative anticipated emotions, such as distrust in mechanisms, concerns about economic costs, preferences for high-carbon consumption, and cognitive biases towards low-carbon practices, diminish consumers’ enthusiasm for participation.

Pro-environmental behavioral abilities encompass low-carbon knowledge and skills, the ability to purchase low-carbon products, and low-carbon consumption habits [56]. In the context of this study, many consumers fail to sustainably participate in CI programs due to their lack of understanding of low-carbon products, inability to make independent low-carbon consumption decisions, dearth of low-carbon living skills, and difficulty in effectively managing the economic and time costs associated with participation. Furthermore, limited ability to share low-carbon information and difficulty in changing ingrained consumption patterns often leave consumers with the intention but not the means to act; notably, the interaction between emotional ambivalence (e.g., A24’s brand attachment conflicts) and ability deficits (e.g., A22’s product knowledge gaps) exemplifies how MOA’s ability component and SPT’s teleoaffective structures co-constitute internal barriers, resulting in a “knowing is easier than acting” in their participation. This offers a new integrative perspective to move beyond separate examinations of emotion or ability found in existing intention–behavior gap literature.

5.4. The Dynamic Changes in Social Practice and External Opportunities and Conditions Constitute the External Contextual Factors That Elucidate the “Intention–Behavior” Gap in CI Participation

The MOA model reveals that opportunities, as a moderating variable, play a crucial role in the translation of consumers’ pro-environmental behavioral intentions into behaviors, serving as the key to understanding the potential gap between the two [10]. The theory of the dynamics of social practices [11] points out that social practices influence consumers’ environmental awareness, behavioral motivations, and other factors through three stages: meaning construction, adaptation, and stabilizing. And this, in turn, moderates the relationship between their pro-environmental behavioral intentions and behaviors, uncovering the challenges that may arise from awareness to behavioral stabilization. Furthermore, external conditions have moderating effects on the relationship between pro-environmental behavioral intentions and behaviors, and these external conditions encompass factors such as moral norms, policy incentives, and environmental technology innovations [57]. And these factors collectively constitute the external conditions for practicing pro-environmental behaviors and are crucial for bridging the “intention–behavior” gap.

Based on this, this study categorizes the external contextual factors influencing participation in CI programs into two types: external opportunities and conditions, and dynamic changes in social practice. This study finds that these two categories of factors exert a moderating effect on the “intention–behavior” relationship in CI participation. Specifically, external opportunities and conditions such as policy guidance, the development of CI platforms, support from digital and intelligent technologies, supply and demand in the low-carbon market, and social atmosphere all influence the enthusiasm for participating in CI programs. However, interview data reveal that inadequate incentive mechanisms, insufficient platform technical support, limited choices of low-carbon products, and the yet-to-be-formed habit of low-carbon consumption constitute major barriers hindering consumers’ continuous participation in CI programs. Meanwhile, dynamic changes in social practice, as a dynamic process encompassing three stages of meaning construction, adaptation, and stabilizing, further moderate the relationship between consumers’ pro-environmental behavioral intentions and actual behaviors by influencing their environmental cognition, motivations, and sociocultural factors. This theory provides a new perspective for deeply understanding the formation and maintenance of pro-environmental behavior.

In summary, external contexts such as external opportunities and conditions, as well as dynamic changes in social practice, not only confirm the existence of the “intention–behavior” gap in pro-environmental behavior but also demonstrate that when these factors significantly influence consumer behavior, they will promote or inhibit their participation in CI programs. Conversely, when these external factors have a weaker impact, pro-environmental behavior relies more on intrinsic behavioral intentions. This duality mirrors MOA-SPT synergies: where MOA’s opportunity structures (e.g., policy incentives) enable practice evolution, SPT’s socio-material configurations (e.g., A17’s smart recycling infrastructure) stabilize them. Such interconnection addresses the existing divide between individual cognitive models and socio-material practice theories within pro-environmental behavior scholarship.

The MICPB model remedies key limitations of prior theories. TPB concentrates on individual cognition but ignores dynamic socio-material practices; MOA separates motivation, opportunity and ability without capturing staged practice evolution; and Social Practice Theory emphasizes socio-material contexts yet understates heterogeneous individual emotions and capabilities. This integrated framework bridges these divides. Meanwhile, its generalizability across varied institutional settings awaits further verification.

6. Conclusions and Policy Implications

This study uncovers the “intention–behavior” gap in CI participation. Through in-depth interviews, we have identified that internal contextual factors (e.g., consumer emotions and pro-environmental competencies), external contextual factors (e.g., platform opportunities and social norms), and the dynamic evolution of social practices collectively moderate the conversion process. By integrating the Motivation–Opportunity–Ability framework with Social Practice Theory, we propose the Motivation–Intention–Context–Practice–Behavior (MICPB) model, which systemically deconstructs how misalignments among motivational, contextual, and practical spheres generate this gap. The MICPB framework advances existing literature by achieving a context-specific integration of MOA and SPT tailored to CI research. Unlike static integrated models such as the Comprehensive Action Determination Model (CADM) [58], this framework incorporates the staged dynamics of social practice and explicitly captures low-carbon behaviors mediated by digital CI platforms in Chinese pilot settings.

The theoretical novelty of the MICPB model lies in three aspects: (1) it bridges individual psychological drivers and socio-structural dynamics, revealing their interactive effects; (2) it reframes the intention–behavior transition as a non-linear, spiral process mediated by social practice evolution; and (3) it highlights how digital platform engagement and offline green behaviors can form interconnected routines that mutually reinforce each other, thereby enhancing the stickiness of low-carbon practices.

These findings offer three key contributions to understanding and optimizing CI mechanisms: (1) highlighting the critical link between digital platform engagement and the performance of offline green behaviors; (2) demonstrating how temporal mismatches between short-term incentive cycles (e.g., monthly rewards) and the long-term process of practice stabilization (e.g., 3-month habit formation) undermine sustained participation; and (3) revealing the critical trade-offs between leveraging extrinsic motivations (e.g., rewards) and fostering intrinsic pro-environmental competencies.

Rooted in the multi-dimensional mechanism of the MICPB model and the axial coding outcomes, this study puts forward a series of refined, actionable policy and design recommendations. Rather than general macro proposals, these targeted measures correspond to different functional pathways within the model and the three-stage evolutionary logic of social practice.

From the perspective of pro-environmental behavioral motivation, consumer participation in CI is influenced by a combination of intrinsic drivers (e.g., environmental responsibility, expected benefits, and pursuit of social recognition) and extrinsic drivers (e.g., policy and commercial incentives, low-carbon market supply, and social norms). Aligned with the MICPB model’s emphasis on phase-sensitive interventions, recommendations for improving policy incentives are as follows: Firstly, the government should strengthen low-carbon education through multi-channel promotions. Secondly, a carbon reduction reward mechanism should be established, offering material rewards, tax exemptions, or green credit policies based on CI. Furthermore, the government should support R&D and production of low-carbon products. Lastly, the government should advocate green lifestyles and create a low-carbon social atmosphere [59], collaborating with communities, enterprises, and schools to form a group effect [6]. These recommendations are directly derived from our axial coding (see Table 4) and can be further refined by tailoring them to practice evolution stages: symbolic rewards (e.g., carbon leader badges) in the meaning construction phase, material support (e.g., subsidized e-bike rentals) in the adaptation phase, and institutionalized recognition in the stabilization phase. In practical operation, authorities can implement tiered recognition systems to match different practice stages. Initial symbolic incentives cultivate public awareness in the meaning construction phase, continuous material support lowers participation barriers during adaptation, and long-term institutionalized benefits consolidate stable low-carbon routines.

Recommendations for improving commercial incentives, derived from our analysis of key barriers in platform design, market supply, and user competencies, are as follows: Firstly, enterprises should enhance R&D of low-carbon products to reduce costs and improve price competitiveness, directly addressing barriers identified in the “Supply and demand of low-carbon market” category. Secondly, enterprises should raise awareness and purchase intention through carbon credit systems and green marketing, targeting the shaping of “Positive anticipated emotions” and alleviating “Negative anticipated emotions”. Thirdly, enterprises should optimize supply chain management and select eco-friendly materials. Most critically, findings from the “CI platform construction and technical support” and “Pro-environmental behavioral ability” categories necessitate that firms develop adaptive platform interfaces and design long-term incentive schemes aligned with the “Dynamic changes in social practice”. Platform operators can optimize product experience according to the MICPB model’s context mechanism: simplify cumbersome carbon recording and reward redemption procedures, add automatic carbon accounting for daily commuting scenarios, and display real-time carbon reduction feedback to mitigate participants’ perceived participation costs.

From the perspective of internal contextual factors, this study reveals the moderating role of the three dimensions of consumer emotions (positive anticipated, negative anticipated, and personal value emotions) and the multifaceted nature of pro-environmental behavioral abilities (encompassing low-carbon knowledge/skills, resource/time management abilities, and learning/adaptability abilities). To bridge the gap, corporate actions should directly target these specific dimensions: Fostering positive anticipated emotions by showcasing green product advantages; mitigating negative anticipated emotions by providing detailed information and third-party certifications to build trust; and addressing personal value conflicts through personalized recommendations. To directly enhance users’ pro-environmental behavioral abilities—a core category encompassing low-carbon knowledge/skills, resource management, and learning adaptability—enterprises must implement adaptive platform interfaces that cater to these varying competency levels and leverage multi-channel education to bridge specific knowledge and skill gaps identified in the data. Furthermore, point rewards and membership systems serve as crucial tools to support consumers through the “adaptation stage” of social practice, helping them overcome emotional distress and behavioral costs while cultivating the sustained habits and competencies necessary for long-term CI participation. Considering heterogeneous user competencies reflected in the MICPB framework, platforms can launch differentiated interface modes. Simplified operation modes suit groups with limited digital capabilities, while social sharing functions satisfy identity demands among young participants.

Our findings clarify that the external context of CI participation is structured by four main categories: Policy guidance and support, CI platform construction and technical support, demand in the low-carbon market, and social atmosphere. Within these categories, our data highlight that effective economic incentives (e.g., subsidies, tax adjustments) must be carefully designed, as issues with reward forms and fulfillment processes can diminish motivation. Similarly, the analysis of digital platforms reveals that their effectiveness hinges not just on integration but on resolving core user concerns regarding platform functionality, data interoperability, and the lack of immediate feedback. Furthermore, technological innovation’s role extends beyond cost reduction to enhancing product quality and diversity to meet market expectations. Finally, the social atmosphere is shaped by a complex interplay of social norms, exemplary demonstrations, and public education, going beyond group pressure alone. Therefore, cross-sector collaboration should be precisely targeted at these identified gaps: (1) unifying data standards to resolve platform interoperability; (2) establishing dynamic incentives aligned with practice stages; and (3) fostering a low-carbon culture through community influence and systematic education. Governments can promote unified carbon accounting standards and mutual recognition of carbon credits across regional CI pilot platforms, removing institutional barriers stemming from inconsistent technical rules.

From the perspective of the dynamic changes in social practice, this study reveals how the three stages of social practice evolution—meaning construction, adaptation, and stabilization—directly moderate the intention–behavior gap by presenting stage-specific challenges. Our findings specify that in the meaning construction stage, the key is to address the core factors of “understanding of CI concepts” and “formation of participation motivations” through targeted education and value reinforcement. During the adaptation stage, interventions must focus on supporting users through the critical processes of “behavioral change,” “overcoming emotional distress,” and “habit formation,” which constitute the primary hurdles in this phase. Finally, the stabilization stage requires efforts to secure the “stability of the CI platform” and foster the “embedding of a low-carbon culture,” which are essential for maintaining long-term participation. The MICPB model integrates these practice dynamics with individual motivation and contextual factors, providing a systemic framework for diagnosing intervention points across the entire behavior change journey.

Consistent with the processual logic embedded within the MICPB model, policymakers and platform operators should adopt phased intervention portfolios. Concept popularization and symbolic incentives dominate the meaning construction phase; continuous tangible incentives sustain participants during adaptation; and diversified value realization mechanisms consolidate low-carbon routines once practices enter the stabilization phase.

In summary, the MICPB model offers a novel theoretical lens and actionable framework for addressing the intention–behavior gap in CI. Future research may quantitatively validate the contribution of each dimension and explore the model’s applicability in other sustainable consumption contexts.

Participants were recruited from ecological environment bureaus and eco-themed cafes, which tend to overrepresent environmentally engaged citizens and environmental practitioners. Limited inclusion of dormant “zombie users” and CI drop-off participants restricts the generalizability of results to broader urban consumers.