Planning Sustainable Smallholder Rubber Supply Chains: Green Intellectual Capital, Green Transformational Leadership, Digital Transformation, and Sustainable Value Co-Creation

The agricultural sector plays a pivotal role in the economies and livelihoods of developing countries, particularly in rural regions where smallholder farmers dominate production [1-3]. In Indonesia, smallholder rubber farming remains a critical source of income for rural households, yet it is increasingly exposed to market volatility, environmental degradation, and social vulnerabilities [4-6]. These intertwined challenges hinder progress towards several Sustainable Development Goals (SDGs), including poverty reduction, decent work and economic growth, responsible consumption and production, climate action, and sustainable rural communities [1]. At the same time, global buyers and emerging regulatory frameworks are raising expectations for traceability, environmental performance, and social responsibility along agricultural supply chains, placing new pressures on smallholder-dominated systems to improve their planning and management practices [7-9].

Fragmented relationships among farmers, intermediaries, and processors limit access to information and technology, and weak environmental management capabilities are typically characteristic of smallholder rubber supply chains [10-12]. This fragmentation undermines the coordination of production, quality management, and environmental practices, making it challenging to design and implement coherent, sustainable supply chain strategies. Although the literature on sustainable supply chain management has expanded rapidly, much of it has focused on large firms, export-oriented agribusinesses, or manufacturing systems, while offering relatively limited micro-level evidence on how sustainability is co-created among dispersed smallholder actors. Existing studies often concentrate on formal certification schemes, upstream-downstream governance, or specific environmental interventions, rather than on the underlying relational and knowledge-based mechanisms that enable collaborative sustainability in smallholder contexts [13, 14].

Three interrelated enablers are particularly relevant for planning sustainable smallholder rubber supply chains. First, green intellectual capital (GIC) reflects the stock of environmental knowledge, skills, and routines embedded in individuals, organizations, and their relationships. It shapes how actors recognize environmental issues, generate green innovations, and integrate sustainability into everyday decisions and coordination processes [15]. Second, green transformational leadership (GTL) encompasses the ability of leaders to articulate a compelling green vision, inspire followers to embrace sustainability goals, and support behavioral change through coaching, role modeling, and empowerment. Such leadership is crucial in smallholder-dominated networks where formal structures and incentives are often weak, and relational influence plays a central role [16]. Third, digital transformation (DT) is reshaping how information is shared, transactions are coordinated, and performance is monitored across agricultural value chains. Digital tools can reduce information asymmetries, enhance transparency, and facilitate more inclusive participation of smallholders in planning and decision-making [17, 18].

Despite these developments, an important gap remains in explaining why sustainability-oriented coordination failures persist in smallholder rubber supply chains. In South Sumatra, sustainability planning is constrained not only by environmental risks and market pressures, but also by fragmented intermediation, information asymmetry, weak traceability, uneven environmental practices in latex handling and processing, and limited cooperative coordination among farmers, traders, and processors. Existing studies rarely explain how these coordination problems are shaped by the interaction of knowledge-based resources, leadership, digital connectivity, and adaptive capacity at the micro level. As a result, the literature still offers limited guidance on why some smallholder networks are better able than others to translate sustainability pressures into shared practices and co-created outcomes. In particular, the role of dynamic capabilities (DC) remains underexplored in such settings, where capabilities are likely to be uneven, actor-specific, and still emergent rather than fully institutionalized [19-21].

This study addresses these gaps by proposing and empirically testing an integrated model of sustainable value co-creation (SVC) in smallholder rubber supply chains in South Sumatra, Indonesia. Drawing on the Resource-Based View (RBV), Dynamic Capabilities Theory (DCT), and Service-Dominant Logic (SDL), the model links GIC, GTL, DT, and DC to SVC among farmers, intermediaries, and processors [22-24]. Survey data from 129 supply chain actors are analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM) to assess both the measurement and structural models and to evaluate the predictive performance of the proposed framework [25].

This study is novel not merely because it applies an established model to a new empirical setting, but because it reframes SVC as a response to persistent coordination failures in a fragmented smallholder rubber supply chain. Specifically, the study brings together GIC, GTL, and DT as complementary enablers of coordination, while examining whether DC operates as a direct driver in a context where adaptive capacity is unevenly distributed across actors. In doing so, the study offers a more context-sensitive explanation of how sustainability-related value is co-created in smallholder rubber systems and why certain enabling mechanisms are more immediately consequential than others. Accordingly, the study addresses the following research questions:

1. How do GIC, GTL, and DT influence SVC in smallholder rubber supply chains?
2. What is the role of DC in shaping SVC in this context?
3. How can these micro-level mechanisms inform SDG-oriented planning and policy strategies for smallholder-dominated rural supply chains?

The contributions of this paper are threefold. First, it extends the literature on sustainable supply chain management and agricultural development by providing micro-level evidence from a fragmented smallholder rubber supply chain, where sustainability outcomes depend on coordination across actors with unequal resources and capabilities. Second, it advances theory by moving beyond an additive use of RBV, DC, and SDL and by positioning SVC as the relational mechanism through which knowledge assets, leadership, and digital connectivity are translated into sustainability-oriented coordination. Third, it offers context-sensitive planning and policy implications for governments, cooperatives, and supply chain leaders by linking the empirical findings to traceability gaps, information asymmetry, fragmented intermediation, environmental management practices, and weak cooperative coordination in South Sumatra. The remainder of the paper is organized as follows: Section 2 reviews the relevant literature and theoretical foundations; Section 3 describes the research methodology; Section 4 presents the empirical results; Section 5 discusses the findings and their implications for planning and policy; and Section 6 concludes with limitations and directions for future research.

3.1 Research design and context

This study adopted a quantitative, cross-sectional survey design to examine how GIC, GTL, DT, and DC influence SVC in smallholder rubber supply chains. The empirical context is South Sumatra, Indonesia, one of the central rubber-producing regions where smallholder farmers dominate production and interact with intermediaries and processors through largely informal arrangements. This context is particularly relevant to SDG-oriented planning because environmental risks, income vulnerability, and limited digital connectivity coexist with growing external pressure to achieve sustainable and traceable supply chains.

The conceptual model developed in this study links four exogenous constructs—GIC, GTL, DT, and DC—to a single endogenous construct, SVC. All constructs are specified as reflective latent variables and estimated using PLS-SEM, which is suitable for complex models, prediction-oriented research, and data that may deviate from multivariate normality.

3.2 Population, sampling, and respondents

The target population comprises key actors in the smallholder rubber supply chain in South Sumatra, namely smallholder farmers, intermediaries (collectors and traders), and processing firms. These actors are directly involved in latex production, collection, transport, and processing, and are therefore central to the planning and implementation of sustainability initiatives in the rubber sector.

A purposive sampling approach was used to ensure that respondents had sufficient experience and decision-making responsibility in rubber production and marketing. Eligible respondents had been involved in rubber-related activities for at least three years and participated in decisions related to production practices, marketing, or coordination with other supply chain actors. Potential respondents were identified through local cooperatives, farmer groups, and industry contacts.

An a priori power analysis using G*Power [57], assuming a multiple regression model with four predictors, a medium effect size (f² = 0.15), a 5% significance level, and 95% statistical power, indicated that a sample size within the range typically recommended for PLS-SEM would be sufficient to detect medium effects. The final sample consists of 129 respondents (farmers, intermediaries, and processors), which exceeds the minimum requirements suggested by the power analysis and standard PLS-SEM rules of thumb. It is therefore adequate for estimating the proposed model and conducting predictive assessments.

3.3 Measurement of constructs

All constructs were operationalized as reflective indicators measured using a structured questionnaire. Items for GIC, GTL, DT, DC, and SVC were adapted from established scales in the literature on GIC, green leadership, DT, and value co-creation, with wording adjusted to fit the smallholder rubber supply chain context.

Respondents were asked to indicate their agreement with each statement on a five-point Likert scale ranging from 1 = “strongly disagree” to 5 = “strongly agree”. GIC items captured the extent of environmental knowledge, skills, shared routines, and relationship-based environmental collaboration within and across organizations. GTL items reflected the degree to which leaders articulated a green vision, inspired pro-environmental behavior, and encouraged collaborative environmental initiatives. DT items assessed the use of digital communication tools, information systems, and traceability-related practices in managing supply chain activities. DC items measured organizations’ perceived ability to sense, seize, and reconfigure resources in response to sustainability-related changes. SVC items captured joint problem solving, shared planning, and collaborative initiatives that generate economic, social, and environmental value among farmers, intermediaries, and processors.

The questionnaire was drafted in Indonesian, the local language of the respondents. Before the primary survey, the instrument was pre-tested with a small group of practitioners and academics knowledgeable about rubber supply chains to assess the clarity, relevance, and cultural appropriateness of the items. Feedback from the pre-test was used to refine wording and layout.

3.4 Data collection procedures

Data were collected through face-to-face and assisted self-administered questionnaires in selected smallholder rubber-producing areas of South Sumatra for 3 months. Enumerators trained in research ethics and familiar with the local context approached eligible respondents through farmer groups, cooperatives, collection points, and processing facilities. After explaining the purpose of the study and assuring respondents of confidentiality, they were invited to complete the questionnaire. For participants with limited literacy or a preference for oral communication, enumerators read the questions aloud and recorded the responses.

Participation in the study was entirely voluntary. Respondents were informed that they could decline to answer any question or withdraw from the survey at any time without any consequences. No financial incentives were offered. The research protocol, including the questionnaire and recruitment procedures, was reviewed and approved by the Institutional Review Board of the Islamic University of Indonesia. All data were anonymized before analysis to protect the identity of participants and organizations.

3.5 Data analysis procedures

PLS-SEM was employed to estimate the measurement and structural models using SmartPLS version 4.1.1.4. This approach was chosen because it is appropriate for prediction-oriented studies, can handle complex models with multiple latent constructs and indicators, and imposes relatively few distributional assumptions on the data. The analysis followed the recommended two-step procedure in which the measurement model was assessed before the structural model [25, 58].

First, the measurement model was evaluated using indicator loadings, internal consistency reliability, convergent validity, and discriminant validity. Indicator reliability was assessed through outer loadings, internal consistency reliability through Cronbach’s alpha, rho_A, and composite reliability, and convergent validity through the average variance extracted (AVE). Discriminant validity was examined using both the Fornell–Larcker criterion and the heterotrait-monotrait ratio (HTMT). In addition, indicator-level collinearity was assessed using the Variance Inflation Factor (VIF) values. Detailed outer loadings, VIF values, and item-retention decisions are reported in Table A1. During the indicator purification stage, all initial outer loadings were examined to assess indicator reliability. Indicators with loadings below 0.70 were evaluated for potential removal, particularly when their deletion improved composite reliability, AVE, and the overall measurement quality of the construct. Indicators with negative loadings were removed because they were inconsistent with the theoretical direction of the construct. After the purification process, the model was re-estimated and only indicators with satisfactory loadings, acceptable VIF values, and adequate theoretical relevance were retained. The initial outer loadings and deletion decisions are reported in Table A2, while the final retained indicators are reported in Table A1.

Second, the structural model was evaluated by examining path coefficients, their significance levels, the coefficient of determination (R²) for SVC, effect sizes (f²), and predictive relevance. Bootstrapping was performed using 5,000 subsamples, with a two-tailed test, and the percentile bootstrap confidence interval method was used to assess the significance of the hypothesized relationships. Out-of-sample predictive power was assessed using PLSpredict in SmartPLS 4, with predictive performance evaluated using Q²_predict values and by comparing PLS-SEM prediction errors with those of benchmark models at the indicator level. Discriminant validity was primarily assessed using the HTMT ratio and the Fornell–Larcker criterion.

To reduce common method bias, several procedural remedies were implemented, including assuring respondents of anonymity, separating construct blocks in the questionnaire, and randomizing item order where appropriate. All items were measured on a five-point Likert scale, and the wording was refined during pre-testing to improve clarity and reduce ambiguity among respondents. Statistically, collinearity diagnostics were also examined to ensure that collinearity did not pose a serious threat to the model estimation.

5.1 Interpretation of key findings

The empirical results highlight the central role of GIC in enabling SVC in smallholder rubber supply chains. The significant and positive path from GIC to SVC suggests that environmental knowledge, shared routines, and collaborative relationships are crucial intangible resources for driving joint sustainability initiatives. In practical terms, when farmer groups, intermediaries, and processing firms possess a better understanding of environmentally friendly practices, have procedures that institutionalize green behaviors, and maintain trust-based relationships with each other, they are more likely to engage in collective actions such as adopting safer coagulants, improving waste management, or meeting sustainability standards. This finding is consistent with the notion that knowledge-based assets constitute a core foundation for both competitive advantage and sustainable development in agrarian contexts [36, 63, 64].

The positive and significant effect of GTL on SVC underscores the importance of leadership behaviors in mobilizing collaborative sustainability efforts. Leaders who communicate a clear green vision, provide inspiration, and encourage participatory decision-making appear to foster a climate in which actors are willing to engage in joint environmental projects and share responsibility for improving supply chain practices. In smallholder networks characterized by informal governance and social norms, such leadership can compensate for weak formal institutions and help bridge the interests of farmers, traders, and processors. The result extends prior evidence on the influence of transformational leadership on pro-environmental behaviors by showing its relevance for cross-organizational collaboration in rural supply chains [65-67].

The significant relationship between DT and SVC indicates that digitalization is not merely a technical trend but a key enabler of collaborative sustainability. The adoption of digital communication tools, basic information systems, and simple traceability solutions appears to support more efficient information exchange, reduce asymmetries between actors, and increase transparency regarding quality and environmental performance. When information flows are more timely and reliable, actors are better positioned to co-plan harvesting, collection, processing, and environmental controls, which strengthens their capacity to co-create sustainable value. This finding reinforces the idea that digital infrastructures, even in relatively simple forms, can be strategically leveraged to support inclusive and environmentally responsible rural development [68-70].

In contrast, DC do not show a significant direct effect on SVC, despite exhibiting a positive coefficient. Rather than treating this result as a null finding with limited meaning, it may indicate that DC in fragmented smallholder rubber networks do not operate as an immediate stand-alone driver of co-created sustainability outcomes. In such contexts, adaptive capacity is likely to be unevenly distributed across actors, weakly institutionalized, and dependent on intermediation structures, cooperative arrangements, and access to usable information. This helps explain why actors may possess emerging green knowledge, leadership support, or basic digital tools, yet still struggle to translate them into coordinated resource reconfiguration at the network level. The result, therefore, suggests that DC may be indirect, contingent, or actor-specific in this setting, becoming influential only when supported by stronger coordination mechanisms, more stable organizational arrangements, and sustained collective learning. In this sense, the non-significant direct path is theoretically informative because it reveals a capability gap between the emergence of sustainability-oriented enablers and the consolidation of higher-order adaptive capacity in smallholder supply chains [71-74].

Overall, the model explains a substantial proportion of the variance in SVC and demonstrates satisfactory predictive performance, indicating that the identified enablers capture important mechanisms shaping collaboration for sustainability in smallholder rubber supply chains. At the same time, the non-significant effect of DC and the still moderate effect sizes suggest considerable room to strengthen the underlying capacities of smallholder networks to support SDG-oriented transformation.

5.2 Contributions to sustainable development and planning

This study makes several contributions to the literature on sustainable development and planning. First, by focusing on a smallholder-dominated rubber supply chain in Indonesia, it addresses a context that has received relatively limited empirical attention compared to more formal and capital-intensive agri-food chains. The results provide micro-level evidence on how environmental knowledge, leadership, and digital practices influence collaborative mechanisms in a setting characterized by pronounced informal relationships, resource constraints, and environmental risks. This contextualized evidence enriches broader debates on how global sustainability agendas, such as the SDGs, intersect with the realities of rural production systems in the Global South.

Second, the study advances theory by integrating the RBV, DCT, and SDL into a single framework centered on SVC. Rather than treating sustainable development as the outcome of isolated technological interventions or top-down regulations, the proposed model conceptualizes it as emerging from the configuration of intangible resources and relational processes across multiple actors. GIC and digital resources are seen as strategic assets; GTL acts as a behavioral catalyst; and SVC represents the mechanism through which these elements are combined in practice to generate economic, social, and environmental value. This integrated perspective offers a more nuanced understanding of the micro-foundations of SDG-oriented planning in rural supply chains.

Third, the findings generate actionable insights for planners and policymakers tasked with designing sustainable development strategies for smallholder regions. By empirically demonstrating the importance of GIC, GTL, and DT, the study identifies specific capability domains that can be targeted through public programs, cooperative initiatives, and private-sector partnerships. Rather than relying solely on infrastructure investment or regulatory enforcement, planning efforts can be directed toward building environmental knowledge, fostering leadership that champions sustainability, and expanding digital connectivity among smallholder actors. These directions align with the SDG targets related to decent work and economic growth (SDG 8), industry, innovation, and infrastructure (SDG 9), responsible consumption and production (SDG 12), climate action (SDG 13), and partnerships for the goals (SDG 17).

Finally, the limited direct role of DC highlights the challenges of implementing transformative change in fragmented smallholder networks. This nuance tempers overly optimistic assumptions about the ease with which smallholder systems can adapt to new sustainability requirements. It suggests that planning approaches need to account for the time and support required to develop higher-order capabilities. The study thus contributes to a more realistic and grounded understanding of the conditions under which SVC can effectively underpin long-term SDG-oriented transformation.

5.3 Policy and planning implications

The findings offer several implications for sustainable development planning in smallholder rubber regions. First, the strong effect of GIC suggests that planning interventions should not stop at generic training programs, but should specifically address the coordination problems that weaken sustainability implementation in South Sumatra, including inconsistent quality handling, limited diffusion of environmentally safer latex-processing practices, and weak knowledge sharing across farmers, collectors, cooperatives, and processors. Local governments and cooperatives can therefore prioritize structured peer-learning forums, practical guidance on cleaner processing routines, and simple shared operating procedures that reduce avoidable variation and improve collective environmental awareness.

Second, the significant role of GTL implies that leadership development should be targeted at the actual coordination nodes of the supply chain, particularly farmer-group leaders, cooperative managers, and intermediary coordinators who influence whether collective rules are followed in practice. In fragmented networks, leadership matters not only for motivation but also for aligning actors around traceability discipline, fairer information exchange, and more consistent environmental practices.

Third, DT should be translated into context-specific planning instruments rather than broad calls for digital infrastructure. In this supply chain, the most immediate priorities are low-cost systems that reduce information asymmetry and fragmented intermediation, such as digital recording of deliveries, simple traceability logs, shared price and quality information, and accessible communication channels linking upstream actors with cooperatives and processors. Such tools can strengthen transparency, support compliance with sustainability requirements, and improve the visibility of environmental and quality performance across transactions.

Fourth, the non-significant direct effect of DC indicates that long-term transformation cannot be assumed to emerge automatically once green knowledge or digital tools are introduced. In practice, adaptive capacity in smallholder rubber systems may depend on gradual institutional strengthening, repeated collective learning, and stronger cooperative coordination. Planning strategies should therefore combine immediate coordination fixes with longer-term efforts to build organizational routines that allow actors to sense sustainability pressures, respond jointly, and progressively reconfigure their practices.

The authors would like to thank the smallholder farmers, traders, and processing firms in South Sumatra who generously shared their time and insights during the survey. We are also grateful to the leaders of farmer groups, cooperatives, and local agencies for their assistance in facilitating access to respondents and field logistics. Institutional support from the authors’ home departments and universities is gratefully acknowledged. Any remaining errors or omissions are the sole responsibility of the authors.