Environmental Behavior Driving Household E-Waste Recycling in Emerging Urban Contexts

Abstract

Rapid electronic waste (e-waste) accumulation poses a critical challenge for urban sustainability in emerging economies. However, few studies have examined what motivates households to actively participate in formal disposal systems, particularly in contexts where infrastructure remains limited. This study investigates the determinants of e-waste recycling intention and behavior in Surabaya, Indonesia. A total of 168 active recyclers are surveyed and analyzed using structural equation modeling and importance–performance mapping. The findings reveal that awareness of environmental consequences significantly influences both recycling intention and actual behavior. Interestingly, while the perceived cost of recycling significantly shapes residents’ intention to participate, it does not translate into a significant effect on their actual recycling behavior. Similarly, the convenience of recycling services shows no significant influence on either intention or behavior. Mediation analysis confirms that environmental awareness indirectly shapes recycling behavior through its effect on intention. These findings suggest that, among early adopters of formal e-waste recycling in a developing-country context, cognitive drivers such as awareness outweigh structural barriers like cost and convenience in shaping long-term recycling engagement. For policymakers, this underscores the importance of highlighting awareness of e-waste impacts and the benefits of proper recycling, alongside efforts to remove physical and financial barriers for broader segments of the population.

1. Introduction

The rapid expansion of electronic consumption has created a critical waste challenge for many urban areas, especially in developing countries. As the volume of discarded electronic devices continues to grow each year, cities face increasing pressure to manage e-waste safely and sustainably. Global E-waste Monitor reported that 62 billion kilograms of electronic waste was generated globally in 2022, a volume five times greater than that documented in formal recycling [1]. Currently, only 22.3% of the total e-waste mass is properly collected and recycled. Improperly discarded products contain toxic chemical substances that damage the environment and endanger human health. In addition to health and environmental risks, the improper disposal of electronics leads to the loss of valuable resources (such as lithium, manganese, metals, plastic, and cobalt) that could otherwise be recovered and reused [2,3]. However, recycling rates in many emerging urban contexts remain low, largely due to weak household participation and underdeveloped formal collection systems [4].

Urban areas play a central role in e-waste generation and accumulation. Cities are hubs of electronic consumption and disposal, and, therefore, they are also key intervention points for establishing closed-loop material systems [5]. Despite this, formal e-waste recycling systems remain underdeveloped in many cities in emerging economies. Instead, informal recycling sectors often dominate, where unsafe dismantling practices and environmentally harmful techniques are commonly applied [6].

While technical infrastructure and regulatory frameworks are essential components of an electronic waste management strategy, they remain insufficient without corresponding behavioral changes among the households that must ultimately utilize these systems [4]. The success of formal collection schemes, recycling programs, and disposal facilities depends fundamentally on citizens’ willingness to participate, separate their waste, and divert materials from informal channels [7]. This behavioral dimension requires understanding the psychological, social, and contextual factors that shape individuals’ decisions to adopt pro-environmental practices, which will be explored further in this paper [8].

As the largest contributor of e-waste in Southeast Asia, Indonesia generates approximately 1.9 million tons of electronic waste [9]. This has several adverse effects on the individuals living around contaminated areas. People living in areas of chronic exposure are vulnerable to these hazardous effects [10]. The negative consequences of toxicants released from e-waste include cardiovascular, respiratory, and digestive illnesses [7,8]. Existing studies have revealed that heavy metal pollution from e-waste exposes children and workers to serious health risks [11]. Furthermore, exposure to polluted soil has been associated with a potential lifetime cancer risk in adults [12].

The issue is particularly visible in large cities such as Surabaya. Despite being one of the country’s major urban centers with growing environmental initiatives, the city continues to struggle with incomplete collection networks, reliance on informal recycling actors, and limited community engagement. Most households still dispose of electronic products together with regular waste or store them at home for long periods [13]. While national and local governments have introduced various regulations and public campaigns, these efforts have not yet translated into high recycling participation. Understanding the behavioral drivers that shape how households manage their electronic waste is therefore essential for designing effective interventions.

With the escalating demand for electronic products and rapid urbanization, concern regarding e-waste is a major challenge in the modernized world, including in Indonesia [13]. Problems such as a lack of proper recycling infrastructure and waste regulation exacerbate environmental contamination [14]. Addressing the e-waste problem remains difficult due to insufficient regulations and policies [15]. Moreover, poor individual participation in and low public awareness of waste management systems further hinder effective waste reduction [7], especially in Indonesia [13]. The absence of environmental education and the limited promotion of sustainability issues are primarily responsible for the low level of environmental awareness [16]. Financial issues and a lack of capacity building prevent sustainable management from being achieved [17]. These challenges pose limitations to the successful adoption of e-waste management practices. Hence, identifying supporting factors that can promote sustainable e-waste management may help address this issue.

Given that most e-waste is produced from the consumption of products that are no longer in use by consumers, it is essential to identify the key enablers of consumers’ active participation in e-waste management. Although previous studies have examined predictors of recycling intention [18,19,20,21], this study contributes new insights by focusing on actual e-waste recycling behavior among active recyclers rather than merely behavioral intention. The previous models used may have quite a lot of constructs, but we argue that the variables used must be more efficient and less complex to provide clear direction for specific urban contexts. There is a need for a more focused, empirically grounded approach that identifies the most influential factors and translates them into actionable recommendations. This study responds to that need by developing a simplified behavioral model that concentrates on five key determinants: consequence awareness, cost of recycling, convenience of recycling, recycling intention, and recycling behavior. By targeting individuals who have already engaged in formal recycling practices, this study provides a deeper understanding of the factors shaping responsible disposal behavior.

The Theory of Planned Behavior (TPB) serves as the primary theoretical framework used in this paper for understanding recycling behavior by positing that behavioral intention is the immediate antecedent of actual behavior [22]. In developed economies with well-established recycling infrastructure, TPB applications consistently find that attitudes and subjective norms dominate as predictors, while perceived behavioral control relates mainly to internal factors like knowledge and habit formation [23]. Structural barriers such as convenience and cost typically function as moderators rather than primary constraints, as infrastructure deficits are minimal [24,25].

In contrast, emerging market applications of TPB must account for fundamental infrastructure deficits that fundamentally reshape how behavioral constructs operate. In countries like Indonesia, where formal collection points are scarce and transparent information about recycling procedures is limited, perceived behavioral control is heavily influenced by external structural barriers [13,26]. Cost considerations carry different weights, as households face significant transaction costs, including transportation to distant facilities and uncertainty about procedures [27,28]. Convenience, rather than being a facilitator, becomes a constraining factor when facilities are simply unavailable [25,29]. These contextual differences explain why findings from developed countries cannot be directly transplanted to emerging markets and justify context-specific investigations such as the present study in Surabaya.

To analyze the relationships among these constructs, the study applies Partial Least Squares Structural Equation Modeling (PLS-SEM). This approach is suitable for behavioral research involving latent variables and allows for examining both direct and indirect effects, especially when data conditions are non-normal or sample sizes are moderate. The model is further complemented with Importance–Performance Map Analysis (IPMA) to move beyond statistical significance toward actionable prioritization. By identifying which factors combine strong influence with low current performance, IPMA enables concrete strategic direction for urban policymakers and waste management stakeholders. This methodological choice also represents a conceptual contribution. While advanced analytical approaches such as Necessary Condition Analysis (NCA) are increasingly discussed in sustainability research [30], their application in urban e-waste behavior studies remains rare and methodologically demanding. Given the data and validity constraints in the current study, NCA is acknowledged as a promising direction for future research rather than imposed prematurely. Instead, this paper establishes a robust and replicable baseline model that future studies can extend with necessity-based logic to explore threshold conditions for circular action. The combination of SEM and IPMA provides a more comprehensive understanding of both influential drivers and essential conditions, offering stronger guidance for designing targeted interventions.

Overall, this study seeks to contribute to the growing body of knowledge on e-waste management by presenting a parsimonious behavioral model tailored to an urban Indonesian context. By focusing on household-level dynamics and integrating complementary analytical methods, this study offers insights that can support more effective and context-specific strategies for improving e-waste recycling participation in Surabaya and similar cities. This paper is structured as follows: It starts with a literature review, including a detailed analysis of major concepts in e-waste management. Section 3 outlines the methods used in this article for replicability. Section 4 presents the results of the analysis. Section 5 interprets these findings to construct a framework of critical barriers and proposes policy models on the issue. Finally, Section 5 offers a concluding summary, including implications for future research.

Research Questions

It is important to clarify the distinction between the research questions (RQs) and the research hypotheses in this study. The research questions articulate the broad investigative objectives of the study, framing the core phenomena to be examined, including the influences of consequence awareness, cost, and convenience on recycling intention and behavior, as well as the mediating role of intention. The hypotheses, by contrast, represent the operationalization of these questions into specific, testable propositions derived from theoretical reasoning and empirical literature [31,32]. While the RQs remain open-ended inquiries, the hypotheses specify the expected direction and significance of relationships among constructs, enabling statistical testing through structural equation modeling.

• RQ1: Do the consequence awareness (CA), cost of recycling (CR), and convenience of recycling (COR) significantly influence e-waste recycling intention (RI) and actual recycling behavior (RB)?
• RQ2: Does recycling intention (RI) significantly influence recycling behavior (RB)?
• RQ3: Does recycling intention (RI) significantly mediate the association between individuals’ e-waste perceptions (CA, CR, and COR) and RB?

2. Literature Review

2.1. The Indonesian Context and Urban Importance of Surabaya

Indonesia is currently one of the largest producers of electronic waste in Southeast Asia [33]. At the regulatory level, Indonesia lacks a comprehensive and specific law governing e-waste management. The general legislation addressing waste management is Law No.18 of 2008 on Waste Management, which does not explicitly regulate e-waste treatment. Nearly two decades after the enactment, roughly 25% to 40% of waste in Indonesia remains unmanaged [34]. Although various programs and initiatives, such as the 3R (reduce, reuse, and recycle) framework, have become widely recognized within communities, weak enforcement and the absence of strict monitoring have undermined their effectiveness [35,36]. The situation also leaves households to rely heavily on informal recycling and private opportunities [37].

Surabaya, the second-largest city in Indonesia, represents an important urban case for examining e-waste management behavior due to its large population, rapid economic growth, and proactive environmental policies. Yet research on household-level behavioral drivers of e-waste recycling in this city remains scarce [13]. Most local studies have focused on technical or institutional aspects of waste management rather than micro-level psychological and behavioral factors. Furthermore, the municipal government has also initiated several waste management initiatives, including waste banks, recycling centers, and community-based waste sorting programs aimed at promoting household participation in waste reduction and recycling through behavior change. These initiatives are supported by the Surabaya Environmental Agency (Dinas Lingkungan Hidup, DLH), which coordinates waste management programs, environmental education, and partnerships with private recycling actors.

A temporary waste disposal site or Tempat Penampungan Sementara (TPS) is the most reliable waste collection location for people in Indonesia. Unlike waste banks, which are community-based systems that offer incentives to encourage recycling participation through the collection of valuable waste, TPS functions as a temporary storage site where all unsorted mixed waste is dropped off in one place before being transported to a final landfill. In contrast, TPS-3R (reduce, reuse, and recycle) refers to waste processing facilities that focus on sorting and managing waste to generate economic value. Across the city, there are approximately 173 official TPS transfer points that function as neighborhood-level sorting and consolidation facilities before waste is transported to larger treatment sites. In addition, the municipal system includes around 190 TPS locations, among which 9 facilities operate as TPS-3R centers, where waste is sorted and partially processed for recycling [38].

At the downstream level, residual waste from these collection points is transported to centralized facilities such as the Benowo landfill, the main disposal and waste-to-energy site in Surabaya. The city processes approximately 1600 tons of municipal waste per day, with a significant portion managed through waste-to-energy technology and recycling initiatives. Despite the presence of formal infrastructure, recycling activities in Surabaya remain strongly influenced by informal actors, including waste pickers and small-scale recyclers, who play a critical role in recovering recyclable materials from the waste stream [39,40].

Despite these initiatives, formal e-waste recycling infrastructure accessible to households remains limited. Unlike general recyclable waste that can be handled through community waste banks, e-waste requires specialized treatment due to the presence of hazardous materials. As a result, formal collection typically occurs through specific programs organized by government agencies, licensed waste management companies, or temporary collection initiatives conducted in collaboration with retailers and environmental organizations [41,42].

2.2. Theoretical Background and Hypotheses Development

This study employs the Theory of Planned Behavior (TPB) because of the importance of psychological factors in governing recycling behaviors. The Theory of Planned Behavior (TPB) is a widely used socio-psychological theory that explains social behavior [33]. TPB is also the preferred framework for assessing influences on recycling behavior [8]. Currently, only a few studies on consumer e-waste management behavior in Indonesia have employed a multi-theoretical approach [43,44,45]. This indicates a lack of a behavioral model to explain both recycling intention and actual behavior, highlighting the need for further investigation in Indonesia.

These variables are selected based on their consistent relevance in e-waste studies and their importance for understanding household decision-making in contexts where recycling infrastructure is limited. By narrowing the scope to essential behavioral and situational factors, the model aims to offer clearer explanations and more practical insights for policymakers and urban waste managers. Existing studies show mixed findings regarding recycling cost and recycling convenience factors. For instance, Juliana et al. [46] suggested that the cost of recycling significantly influences household recycling behavior in Malaysia, whereas Waheed et al. [47] reported a negative effect on e-waste recycling behavior among UAE residents, noting that residents hold negative perceptions when the costs are high. For convenience, Khan et al. [48] highlighted that the ease of convenience is a crucial contributor to recycling practices. Meanwhile, Waheed et al. [47] found that the proper recycling facilities moderate the relationship between recycling intentions and behavior. The availability of nearby recycling facilities can motivate individuals to participate in e-waste recycling [49].

The hypothesis development in this study is guided by several key theoretical interpretations that reflect both the TPB framework and the specific Indonesian context. These interpretations shape how each construct is expected to operate and inform the policy implications that follow. Four key theoretical interpretations guide the hypothesis development in this study. First, consequence awareness (CA) is interpreted as a foundational cognitive driver that may operate through dual pathways: directly by influencing behavior through internalized norms or habits and indirectly by shaping recycling intention [50,51]. This interpretation acknowledges that awareness may need to compensate for infrastructure deficits in contexts where structural supports are weak. Second, structural barriers, which are the cost of recycling (CR) and convenience of recycling (COR), are interpreted as conditional constraints whose influence depends on contextual conditions and prior participation. Among active recyclers who have already overcome initial barriers, these factors may show different patterns of significance for intention versus behavior [25,27].

Third, the intention–behavior relationship is interpreted as contextually variable, with the intention–behavior gap potentially widening in settings with significant structural barriers [51,52]. Among active recyclers, however, this gap may narrow as individuals have demonstrated the ability to translate motivation into action. Fourth, mediation pathways are interpreted differently, in which intention is expected to serve as a more effective mediator for psychological factors like awareness than for structural factors like cost and convenience, which may influence behavior through direct pathways that bypass conscious intention formation [22]. These interpretations collectively inform the study’s hypotheses and shape expectations about which relationships will achieve statistical significance in the Indonesian context. Below, this study will elaborate on each aspect of the behavioral determinants of e-waste recycling before moving on to the methodology.

• Consequence Awareness (CA)

Consequence awareness refers to an individual’s cognitive recognition of the negative environmental, social, and health impacts associated with improper e-waste disposal [52]. It represents the knowledge-based and ethical dimension of environmental responsibility. Higher awareness of environmental consequences has been consistently linked to stronger pro-environmental attitudes [46].

In the realm of e-waste management, numerous studies report that individuals who are aware of the toxicity and environmental risks of electronic waste are more likely to express intentions to recycle. A study by Ramayah et al. [23] found that environmental awareness significantly predicted recycling intention for electronic products. Similarly, Roy et al. [5] confirmed that knowledge about hazardous components in e-waste increases households’ motivation to seek responsible disposal methods.

Nevertheless, awareness alone does not guarantee behavioral implementation. Several studies have observed what is known as the “awareness–action gap,” where individuals acknowledge the problem but fail to act accordingly [50]. It is expected that consequence awareness could significantly affect students’ recycling intention toward e-waste. However, evidence on the role of awareness regarding the “awareness–action gap” requires further study to explore the role of consequence awareness on recycling behavior directly. The following hypotheses are formulated:

H1: 

Consequence awareness (CA) significantly influences recycling intention (RI).

H2: 

Consequence awareness (CA) significantly influences recycling behavior (RB).

2.

Cost of Recycling (CR)

Perceived cost refers to the perceived net economic costs or benefits associated with household e-waste disposal activities, as reflected in respondents’ own experiences and expectations. In e-waste recycling contexts, these perceived costs often act as a significant deterrent [53].

Empirical findings indicate that households are less likely to participate in recycling when the perceived effort outweighs the perceived benefit. For instance, a study by Ahmad et al. [27] found that transaction costs, such as travel distance to recycling centers and uncertainty about procedures, significantly reduce the probability of e-waste recycling behavior. Even when environmental concern is high, households may avoid recycling due to perceived inconvenience and inefficiency. These findings reinforce the need to minimize perceived cost in order to strengthen behavioral outcomes. Thus, with higher perceived costs, individuals would be less likely to recycle their e-waste. Hence, this study proposed the following hypotheses:

H3: 

Cost of recycling (CR) significantly influences recycling intention (RI).

H4: 

Cost of recycling (CR) significantly influences recycling behavior (RB).

3.

Convenience of Recycling (COR)

Convenience is regarded as one of the most powerful predictors of recycling behavior [24]. In the context of e-waste, convenience refers to the accessibility, simplicity, and user-friendliness of disposal or recycling systems. Factors such as proximity of drop-off points, availability of collection services, and clarity of procedures play crucial roles in determining participation.

Several studies demonstrate that increased convenience significantly boosts recycling rates. For example, Zhang et al. [52] found that households living within close proximity to collection facilities exhibited significantly higher recycling intention and behavior. Similarly, Nguyen et al. [54] reported that the availability of scheduled pickup services dramatically increased the likelihood of household participation in formal e-waste recycling programs.

Convenience also interacts with awareness and cost. While a household may possess high environmental awareness, limited convenience can undermine the translation of intention into actual behavior. This makes convenience a potential leverage point for practical policy implementation. Therefore, it is predicted that recycling convenience could substantially influence recycling behavior. This proposition will be examined in this study through the hypotheses proposed below:

H5: 

Convenience of recycling (COR) significantly influences recycling intention (RI).

H6: 

Convenience of recycling (COR) significantly influences recycling behavior (RB).

4.

Recycling Intention and Recycling Behavior

Recycling intention refers to the motivational readiness of an individual to engage in recycling behavior [22]. It captures the degree to which a person is inclined, planned, or determined to perform recycling in the near future. Therefore, if individuals have a positive intention toward e-waste recycling, they are more likely to perform such behavior. This suggests that recycling intention has a significant positive influence on recycling behavior and is supported by the literature. However, research highlights an intention–behavior gap, where psychological commitment fails to translate into execution due to barriers [51,52]. Despite this gap, intention remains a primary antecedent. Thus, the following hypothesis is formulated:

H7: 

Recycling intention (RI) significantly influences recycling behavior (RB).

Existing research on e-waste behavior found that, although intention is positively correlated with actual recycling behavior, the relationship is often weakened by situational barriers [52]. This is described as the intention–behavior gap, in which psychological commitment fails to translate into practical execution due to environmental, infrastructural, or economic constraints [51].

In urban e-waste contexts, this gap is often widened by poor accessibility to recycling facilities, a lack of time or competing priorities, the complexity of e-waste handling and data security concerns, and low perceived immediate benefit.

Additionally, this study further explored the indirect relationships of the three independent constructs (CA, CR, and COR) with RB via RI. RI is the motivational readiness and anticipated outcome of the valuation of external factors. Thus, this outcome is anticipated to mediate the association between stimuli (awareness, cost, and convenience) and response (behavior). The theoretical reason is that these stimuli may directly and/or indirectly influence RB through RI. The mediation function of intention has been documented in behavioral research. Interpreting how CA, CR, and COR influence RI, and subsequently RB, allows the model to capture both psychological and structural influences. Therefore, the following hypotheses were proposed:

H8: 

RI significantly mediates the association between CA and RB.

H9: 

RI significantly mediates the association between CR and RB.

H10: 

RI significantly mediates the association between COR and RB.

Figure 1 presents the research framework based on the proposed hypotheses.

3. Materials and Methods

3.1. Sample and Data Collection

A cross-sectional survey design was employed to investigate the determinants of responsible e-waste disposal behavior. This approach was considered appropriate because it allows researchers to examine the relationships among variables at a single point and has been used in previous studies investigating participation in recycling activities [33,55,56]. The survey was conducted from July to September 2025 (three months) in Surabaya, specifically in city areas (South Surabaya and Central Surabaya).

Respondents were selected through purposive sampling because the study focused on specific participants with relevant experiences. Data collection was conducted at collection sites during operational hours. The respondents were approached and gave informed consent prior to the study. After the nature of the study was explained, each respondent filled out a questionnaire, and the enumerator explained any matters that needed clarification. Respondents were identified as household residents of Surabaya who had prior experience in delivering electronic waste through organized collection mechanisms such as waste bank programs or other recycling initiatives. The targeted participants had to have experience in engaging in e-waste disposal by handing over electronic products such as smartphones, laptops, or household appliances to authorized waste management operators or local environmental services. These participants were initially identified based on their relevant knowledge and active engagement in waste management activities, including practices such as waste drop-off and delivery to authorized waste management services.

This study focused specifically on household-generated e-waste, such as mobile phones, small household appliances, and personal electronic devices. Corporate e-waste streams, which are typically managed through separate procurement, asset disposal, or contracted recycling services, were not included in the sampling frame. This distinction is important because such institutional electronic waste is often handled through different regulatory and logistical mechanisms to household waste.

Data were collected through structured questionnaires administered to the selected participants. By focusing on individuals with direct experience in existing take-back schemes, the study ensured that the analysis reflected actual behaviors and preferences that support the adoption of circular economy practices. This study successfully gathered 188 samples. However, after running a validity test, we removed 20 invalid samples to ensure data quality. Therefore, 168 respondents were examined for further statistical analysis.

This study examined 168 respondents, comprising 76 males and 92 females. As shown in

Table 1. Respondents’ profile.

Demographic Variables	Frequency	Percentage (%)
Gender
Male	76	45.24
Female	92	54.76
Age
18–24	23	13.69
25–34	112	66.67
35–44	29	17.26
45 or above	4	2.38
Educational level
High school or under	70	41.67
Bachelor’s degree or Sub-degree	91	54.17
Master’s degree or above	7	4.16
Monthly income (IDR)
Under 1,000,000	4	2.38
1,000,001–3,000,000	40	23.81
3,000,001–5,000,000	78	46.43
5,000,001 or above	46	27.38
Frequency of e-waste disposal
1–2 times/year	147	87.5
More than 3 times/year	21	12.5
Types of waste that have been submitted
Cellphone/tablet	42	25
Laptop/computer	31	18.45
TV	24	14.29
Electronic household appliances	46	27.38
Batteries/accumulators	25	14.88

, the majority of the respondents were aged 25–34 years, accounting for 72.61% of the sample. Additionally, 17.26% and 14.69% of the participants were aged 23–44 years and 18–24 years, respectively, and the smallest proportion of respondents was aged 45 years or above. Regarding waste engagement activities, 147 respondents disposed of e-waste 1–2 times per year, and the remaining 21 respondents disposed of it more than 3 times per year. Moreover, electronic household appliances were the most frequently submitted type of e-waste, with 46 respondents reporting this category. This was followed by cellphones/tablets (42 respondents), laptops/computers (31 respondents), batteries/accumulators (25 respondents), and lastly TVs (24 respondents). This sample size is sufficient, given that it exceeds the recommended tenfold rule based on the maximum number of arrowheads directed to the latent variable [32].

3.2. Instrument Development

The measurement of constructs was developed based on relevant literature and prior empirical studies. In this study, five constructs comprising a total of fourteen items were utilized to reflect the perceptions of waste management’s impact on climate change and to assess the implementation of circular economy practices at the individual level. Several items were adapted from existing validated scales to ensure content validity, while other items were self-developed to capture e-waste management behavior. The research instrument included the following constructs: consequence awareness (2 items), cost of recycling (2 items), convenience of recycling (4 items), recycling intention (3 items), and recycling behavior (3 items). All items were measured using a five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree), following the approach of several prior studies [57,58,59,60].

3.3. Data Analysis Approach

This study employed a multi-analytical approach involving Partial Least Squares Structural Equation Modeling (PLS-SEM) integrated with Importance–Performance Map Analysis (IPMA). To evaluate both approaches, this study used SmartPLS 4 to analyze the data. Prior studies mainly applied PLS-SEM [61,62,63], which focused only on the significance of the theoretical relationship between constructs and did not compare the importance and performance of antecedents on the target construct. To uncover potential constructs that need further attention, IPMA was used to enrich the path coefficient results from PLS-SEM, helping to identify areas for improvement and prioritize resources to achieve the best outcome [64]. This combination of methods enhances the robustness of theoretical testing and the relevance of practical implications. The integrated design enables the identification of significant predictors and high-priority areas for improving pro-environmental behavior among individuals.

PLS-SEM is known for its predictive power in identifying key target constructs within complex models, both in reflective and formative indicators [65]. Moreover, it can easily handle small sample sizes and single-item constructs with no identification problem [32]. Conceptually, the model is structured around two fundamental elements: measurement and the structural model. We first assessed the measurement model to examine the relationship between the construct and the indicator variables by analyzing the validity and reliability. Subsequently, a structural model was conducted to confirm the hypotheses. In addition, IPMA is utilized to evaluate both the total effects and rescaled latent construct scores, representing their importance and performance. This facilitates the identification of priority constructs that need high attention to enhance the target constructs [66]. By identifying the viable indicators of each construct, the targeted outcomes can be achieved. Given the complexity of the proposed relationship between constructs, this study applied PLS-SEM and IPMA, as has been done in previous studies in the environmental field [67,68,69].

4. Results

This study examined 168 respondents, comprising 76 males and 92 females. As shown in Table 1, the majority of the respondents were aged 25–34 years, accounting for 66.67% of the sample. Additionally, 17.26% and 13.69% of the participants were aged 35–44 years and 18–24 years, respectively, and the smallest proportion of respondents was aged 45 years or above. Regarding waste engagement activities, 147 respondents disposed of e-waste 1–2 times per year, and the remaining 21 respondents disposed of it more than 3 times per year. Moreover, electronic household appliances were the most frequently submitted type of e-waste, with 46 respondents reporting this category. This was followed by cellphones/tablets (42 respondents), laptops/computers (31 respondents), batteries/accumulators (25 respondents), and lastly TVs (24 respondents). These compositions offer varying insights into individuals’ participation in e-waste recycling.

To ensure that all items were measured in the same direction before assessment, reverse coding was conducted on the cost of recycling items. Following Hair et al. [32], examining the variance inflation factor (VIF) values was the first step in assessing the structural model. As presented in

Table 2. Reliability and convergent validity assessment.

Constructs	Items	Item Contents	Outer Loading	Full Collinearity VIF	CR	AVE
Consequence awareness [16,62]	CA1	Some electronic waste has valuable metals that can be recycled and sold	0.849	1.371	0.807	0.677
	CA2	Recycling e-waste helps conserve natural resources, such as metals and minerals.	0.796
Costs of recycling [25]	CR1	I feel that the money I get from selling recyclable e-waste is too little	0.791	1.230	0.826	0.704
	CR2	I feel that the costs for sorting or processing e-waste before selling are high	0.885
Convenience of recycling [25,62]	COR1	I have enough time to bring e-waste to the collection place	0.910	1.097	0.871	0.631
	COR2	I have an easy way to transport e-waste to the collection point	0.709
	COR3	I can send or deliver e-waste to the collection place when needed	0.751
	COR4	It is easy for me to get information about how and where to recycle e-waste	0.792
Recycling intention [25,71]	RI1	I plan to participate in e-waste recycling activities regularly	0.711	1.613	0.779	0.540
	RI2	I’m willing to allocate time to deliver my old electronic devices to designated recycling facilities	0.763
	RI3	When purchasing electronic products in the future, I prefer products that are designed to be recyclable	0.731
Recycling behavior (self-developed)	RB1	I am willing to recycle or donate old electronic devices instead of throwing them away	0.759	1.461	0.832	0.625
	RB2	I support the idea of using refurbished or second-hand electronic products	0.723
	RB3	I support brands or sellers that offer e-waste take-back or recycling programs	0.881

, all full collinearity variance inflation factors (VIFs) are lower than the threshold of 3.3, indicating that the model is free of multicollinearity issues [70]. Subsequently, the measurement model assessment was evaluated to ensure the validity and reliability of multiple items and their corresponding constructs. Table 2 shows the results of the outer loadings, with all item values ranging from 0.709 to 0.910, which are higher than the threshold value of 0.708 [31]. In addition, as Hair et al. [32] suggested, the minimum acceptable value for the average variance extracted (AVE) is 0.50. Likewise, the AVE values range from 0.540 to 0.704, further confirming convergent validity. This study did not use Cronbach’s alpha to assess internal consistency reliability due to the limitation in handling scales with a small number of items [31]. In this case, composite reliability was technically more appropriate for measuring reliability. As demonstrated in

Table 3. Discriminant validity.

Fornell–Lacker Criterion
Constructs	CA	CR	COR	RI	RB
Consequence awareness	0.823
Costs of recycling	−0.282	0.839
Convenience of recycling	0.048	−0.247	0.794
Recycling intention	0.477	−0.329	0.158	0.735
Recycling behavior	0.415	−0.292	0.185	0.534	0.791
HTMT Ratio
Constructs	CA	CR	COR	RI	RB
Consequence awareness
Costs of recycling	0.488
Convenience of recycling	0.089	0.405
Recycling intention	0.863	0.564	0.230
Recycling behavior	0.652	0.425	0.199	0.823

, the composite reliability values are between 0.779 and 0.871, which are greater than 0.70 [32]. Hence, this also proved that internal consistency of the latent variable was achieved.

According to Table 3, the square root of each construct’s AVE is higher than its squared correlation with any other constructs, confirming the Fornell–Lacker criterion [72]. Additionally, the HTMT values are all below the threshold of 0.90 [73]. These findings reveal that discriminant validity is also established. This indicates that no issues related to validity and reliability were found in the analytical findings. Therefore, the structural model could be assessed.

Before hypothesis testing, the f² effect sizes, predictive relevance (Q²), and coefficients of determination (R² values) are used to evaluate the quality of the proposed framework. Lastly, the significance of the path coefficients is examined to determine the strength and direction of the relationships between variables.

Table 4. Coefficient of determination (R²), predictive relevance (Q²), and effect size (f²).

Constructs	R2	Q2	f2 (RI)	f2 (RB)
CA			0.223	0.044
CR			0.042	0.009
COR			0.011	0.012
RI	0.263	0.133		0.173
RB	0.320	0.179

shows that CA, CR, and COR explain 26.3 percent of RI’s variance, while the four constructs (CA, CR, COR, and RI) predict 32 percent of the variance in RB. This indicates that all predictors have moderate explanatory power. Moreover, the predictive relevance (Q²) for RI (0.133) and RB (0.179) is higher than zero, demonstrating acceptable predictive accuracy. In addition, the effect size (f²) results reveal that CA has a medium effect on RI (f² ≥ 0.15) but a small effect on RB (f² < 0.15). However, the cost and convenience of recycling have a small effect on both outcomes (f² < 0.15), while RI has a medium effect on RB (f² ≥ 0.15) [74].

Hypothesis testing on

Table 5. Path coefficient and hypothesis testing.

Hypotheses	Path	Beta	T-Value	p-Value	Decision
H1	CA -> RI	0.419	4.432	0.000	Supported
H2	CA -> RB	0.197	2.395	0.017	Supported
H3	CR -> RI	−0.188	2.075	0.038	Supported
H4	CR -> RB	−0.083	0.938	0.348	Not Supported
H5	COR -> RI	0.092	0.878	0.380	Not Supported
H6	COR -> RB	0.092	1.034	0.301	Not Supported
H7	RI -> RB	0.399	4.014	0.000	Supported

presents mixed results. It reveals that three hypotheses (H1, H2, and H7) are supported, while four hypotheses (H3, H4, H5, and H6) are not supported. Specifically, CA significantly affects RI and RB, supporting H1 (β = 0.480, p < 0.001) and H2 (β = 0.220, p < 0.05), respectively. This finding suggests that consequence awareness plays a critical role in shaping both recycling intention and behavior. However, CR negatively influences RI (β = −0.188, p < 0.05) but has no significant effect on RB (p > 0.05), therefore supporting H3 but rejecting H4. This indicates that individuals do not perceive the cost of recycling as influential on their recycling behavior, but it remains a consideration when deciding whether to engage in recycling. Similarly, COR does not significantly affect RI or RB (p > 0.05), which indicates the rejection of H5 and H6, thus suggesting that convenience does not motivate the participants to actively recycle their e-waste. Nevertheless, RI exerts a significant influence on RB (β = 0.380, p < 0.001), supporting H7 and confirming that intention promotes actual recycling behavior. The result of the mediation analysis in

Table 6. Hypothesis testing through mediation analysis.

Hypotheses	Path	Beta	T-Value	p-Value	Decision
H8	CA -> RI -> RB	0.167	3.065	0.002	Supported
H9	CR -> RI -> RB	−0.075	1.786	0.074	Not Supported
H10	COR -> RI -> RB	0.037	0.821	0.412	Not Supported

shows that only one hypothesis (H8) is supported and the remaining two hypotheses (H9 and H10) are unsupported. This indicates that CR and COR do not influence RB, either directly or indirectly. In contrast, CA also significantly affects RB indirectly through RI. This indicates that consequence awareness has a direct and indirect relationship with recycling intention and behavior, while neither cost nor convenience of recycling does. The estimated path coefficients and significance levels of the structural model results are illustrated in Figure 2.

Furthermore, Importance–Performance Map Analysis (IPMA) is utilized to identify constructs that need improving to foster individuals’ recycling intention and behavior. As presented in

Table 7. IPMA analysis.

Constructs	Recycling Intention		Recycling Behavior
	Important	Performance	Important	Performance
CA	0.419	76.711	0.364	76.711
CR	−0.188	23.340	−0.158	23.340
COR	0.092	71.998	0.129	71.998
RI			0.399	78.947
Average	0.108	57.349	0.184	62.749

, the importance value of CA (0.419) for recycling intention is higher than the average importance value (0.108). Additionally, the results show that two constructs have performance values higher than the average value (57.349), namely, CA (75.092) and CR (76.660). Moreover, CA and RI are also crucial factors in achieving recycling behavior (RB), as their importance values of 0.364 and 0.399, respectively, are greater than the average value (0.184). The performance values of CA (76.711), COR (71.998), and RI (78.947) are satisfactory, as they are higher than the average performance value of RB (62.749). The importance value of CR for both outcomes is negative, reflecting an inverse relationship between recycling cost and the target constructs. The IPMA findings reveal that CA is a key driver that needs to be maintained to promote recycling intention and recycling behavior. Likewise, RI should also receive more priority. Meanwhile, COR is less critical in order to attain both targeted outcomes. The visualization of IPMA analysis on RI and RB is presented in Figure 3 and Figure 4.

5. Discussion

This study aims to understand the determinants of recycling intention and behavior among individuals who actively engage in formal e-waste recycling or organized or semi-formal collection channels. While existing studies mostly focused on investigating e-waste recycling intention [75,76], only a few scholars have studied actual behavior [77,78]. Therefore, this study attempts to fill this gap by examining the direct and indirect relationships to predict recycling intention and behavior among recyclers. For this purpose, three constructs, including consequence awareness (CA), cost of recycling (CR), and convenience of recycling (COR), are proposed in the study’s framework. The results show that six out of ten hypotheses are unsupported, and the remaining four are supported.

For instance, consequence awareness is found to have a significant influence on both recycling intention and behavior. In addition, the IPMA results show that consequence awareness is a major determinant shaping motivation to engage in e-waste recycling activities. This finding reinforces Nketiah et al.’s [76] and Wang et al.’s [25] previous studies. It demonstrates that individuals with a higher level of awareness have moral responsibilities and are more likely to help minimize the negative environmental repercussions of inadequate e-waste management by handing over their waste to formal collection services, or in Surabaya’s case, to organized or semi-formal collection channels. Nketiah et al. [76] found that people who were concerned about environmental degradation were willing to engage in pro-environmental actions. Recycling e-waste contributes to the conservation of natural resources (such as metals and minerals). Also, environmental protection behaviors can aid in preventing hazardous substances from e-waste from causing serious harm to humans [25]. In addition, an individual’s involvement may play a role in raising awareness among others. Danish et al. [79] revealed that individuals tends to participate in environmental initiatives after observing others’ activities within their social groups. Considerable efforts such as public awareness campaigns have also been found to promote a sense of environmental responsibility [80]. Given the critical role of this predictor, several approaches through a combination of stakeholders’ interventions, public education, and individual-based programs can be implemented to promote e-waste recycling activities. Additionally, collaboration among stakeholders in driving behavioral change, including enhancing public awareness campaigns, integrating sustainability into educational systems, and creating individual engagement programs, is crucial to promoting environmentally responsible initiatives.

In contrast, the cost of recycling is found to negatively influence recycling intention but have an insignificant effect on recycling behavior. A possible explanation for this is that the targeted respondents are individuals who already engage in e-waste recycling and do not perceive cost as a significant factor, as recycling has become a routine part of their lifestyle. Moreover, the most commonly submitted e-waste items are cellphones and electronic household appliances, which differ in size, handling requirements, and transportation needs. Nevertheless, the active recyclers do not consider these factors to be barriers to engaging in recycling activities. However, they still view financial factors as a strong motivation and consideration for participating in environmental activities. This result contradicts that of Ma et al. [81] and Cai et al. [28]. Households perceived cost as one of the main contributors to their waste separation behavior, as waste sorting required a lot of energy and money [81]. Likewise, people were unwilling to pay anything to recycle, and they showed a greater preference for selling e-waste to second-hand markets for money [28]. This indicates that the more individuals have to spend, the less willing they are to recycle e-waste [44]. To promote and maintain the willingness of individuals to engage in recycling activities, government provision of funding and incentives for each activity may help reinforce individuals’ motivations, especially among those who have not been previously conducted recycling activities. Furthermore, e-waste contains precious materials that offer financial benefits. Ravindra and Mor [82] reported that recoverable and recyclable components of e-waste are worth $65,000 per annum in Chandigarh, India. Similarly, in Indonesia, the economic value of recoverable materials could reach up to $14 billion [83]. Therefore, creating environmental programs that integrate economic incentives can increase public participation in e-waste management and promote the implementation of better circular economy practices. Stakeholders should strengthen the enforcement of extended producer responsibility (EPR) regulations to ensure that waste electrical and electronic equipment is processed into new resources through effective take-back systems. Enterprises involved in the production and distribution of electronic products are encouraged to implement pro-environmental initiatives, such as manufacturing products using reused spare parts and offering rewards to recyclers who return their old electronic devices. These measures can contribute to minimizing environmental degradation and promoting circular economy practices.

In addition, the convenience of recycling (COR) does not influence recycling intention and behavior. This result is inconsistent with previous studies that found convenience was a significant driver in motivating recycling initiatives [84]. The lack of proper waste infrastructure and limited access to waste treatment facilities lead some people to consider recycling e-waste as a difficult process that requires substantial effort and time. Although formal service facilities in Indonesia are often distant and hard to find, especially for people living in remote areas, active recyclers have developed adaptive strategies to overcome infrastructural limitations. These respondents reported recycling various forms of electronic waste regardless of size and handling difficulty. Their action is primarily motivated by moral obligation and environmental commitment rather than situational convenience.

Unlike general populations that require convenience to act, this study’s respondents of active recyclers demonstrate that high consequence awareness (CA) acts as a motivational buffer, rendering structural inconveniences statistically insignificant. However, developing more accessible recycling facilities remains a strategic approach to promote the habit of recycling, even among non-active recyclers, and to gradually mitigate environmental degradation [29]. Therefore, government involvement in supporting sufficient collection services and enforcing waste policies can further support the individual’s responsibility for environmental protection [81]. In general, Indonesia still relies on open dumping and conventional landfill systems, which often result in inefficient and unsustainable waste management. Stronger government involvement is required to invest in waste-to-energy technologies and to develop modern waste management facilities across regions in order to decrease waste volume while converting it into renewable energy and valuable resources. On the other hand, stricter regulations should be enforced through fines and penalties for individuals who improperly dispose of waste in public spaces, supported by surveillance systems and social accountability mechanisms.

In addition, recycling intention is evidenced to have a significant effect on recycling behavior. Moreover, the IPMA result further shows that recycling intention is also one of the key drivers of e-waste recycling behavior. These findings reveal that individuals who have the intention to recycle are more likely to take action to appropriately dispose of e-waste. The responses demonstrate that they are willing to allocate time to deliver old electronic products and participate in e-waste recycling activities regularly. This result is consistent with that of Shaharudin et al. [85], who found that a strong intention for proper e-waste disposal led to practical implications for environmental practices. Nevertheless, the mediation analysis shows that recycling intention does not mediate the effects of the cost of recycling or the convenience of recycling on actual behavior. These results support the evidence regarding their direct relationships, while consequence awareness has a significant influence on recycling behavior through recycling intention. Individuals who understand the environmental consequences of inadequate e-waste disposal develop a sense of responsibility and motivation that helps promote their engagement in sustainable practices. Moreover, the availability of recycling facilities and the provision of financial rewards would support the development of sustainable communities, especially in the context of the municipal city of Surabaya, which only has organized or semi-formal collection channels for e-waste. Building better infrastructure provides a more convenient pathway for individuals to engage more in recycling activities, whereas incentives serve as an impetus to increase the economic motivation for pro-environmental initiatives, especially among non-recyclers [64]. However, without sufficient education on recycling programs, individuals may not develop higher environmental awareness or understand proper e-waste practices [86].

Overall, the findings of this study support Sustainable Development Goals (SDGs) 11 and 12, particularly in achieving responsible consumption and sustainable communities. By understanding the determinants of participation in recycling behavior, this study offers valuable insights for improving public engagement in sustainable practices and promoting circular resource utilization. On the other hand, increased consequence awareness and stronger citizen involvement in e-waste recycling initiatives would improve resource recovery and help reduce environmental degradation.

6. Conclusions

Although previous studies have examined various predictors of recycling intention, this study offers a new theoretical contribution to understanding the influence of factors on actual e-waste recycling behavior among active recyclers. By combining SEM and IPMA, this study not only identifies significant determinants of recycling behavior but also highlights the relative performance and importance of these factors, thereby offering clearer priorities for improving e-waste recycling practices in the urban context. The results show that consequence awareness (CA) is the most crucial factor directly stimulating e-waste recycling behavior and exhibits an indirect effect through recycling intention. However, recycling cost (CR) has a negative effect on RI but does not significantly influence RB. In contrast, recycling convenience (COR) does not have a significant influence on either e-waste recycling intention or behavior, and mediation analysis of CR and COR confirms the absence of indirect effects on the outcome. Moreover, recycling intention (RI) is found to significantly influence recycling behavior. These findings highlight important factors that motivate active recyclers to participate in pro-environmental activities and thereby contribute to promoting Sustainable Development Goals (SDGs) 11 and 12. Therefore, to develop more sustainable communities and promote responsible consumption in Indonesia, several policy implications are suggested: enforcing extended producer responsibility (EPR), implementing take-back systems, imposing fines and penalties for non-compliance, improving waste management facilities, investing in modern technologies, and providing financial rewards for environmental initiatives.

Despite the contributions, this study has several limitations. This study explored the key drivers of sustained recycling behavior among recyclers. Therefore, as individuals who already engage in recycling practices, cost and convenience factors do not play critical roles. Therefore, future studies should compare active and inactive recycling individuals and examine how these relationships differ across socio-demographic groups. Subsequently, this study used a limited number of predictors to explore e-waste recycling intention and behavior. For instance, social influence from family and peers may encourage participation through perceived social pressure. In addition, this study did not explicitly differentiate between types of e-waste, which may influence perceptions of recycling activities. Lastly, the use of a cross-sectional survey limits the participants from explaining the reason behind their answers and does not allow for observation of behavioral changes over time. Hence, future work should investigate more diverse elements and incorporate interview-based data collection methods to gain a more comprehensive understanding of the context.