A Dynamic Theoretical Framework for Agricultural Value Chain Resilience

Thesis Abstract

This study addresses the persistent vulnerability of agricultural value chains to systemic shocks—such as extreme weather events, market volatility, and policy disruptions—by developing a dynamic theoretical framework that integrates resilience, complexity, and learning processes within value chain governance. The aim is to construct a parsimonious yet robust model that explicates how adaptive capabilities, information feedback loops, and institutional arrangements interact over time to sustain performance and reduce disruption costs in agricultural value chains. Specific objectives are (i) to conceptualize a dynamic resilience framework that merges the SCRC (supply chain resilience and complexity) lens with organizational learning and institutional theory; (ii) to identify key levers (risk sensing, adaptive capacity, diversification, and collaborative governance) and their temporal effects on resilience outcomes; (iii) to test the hypothesized relationships using longitudinal data from multiple commodity value chains; and (iv) to derive policy and managerial recommendations for strengthening resilience in both smallholder and aggregate-level supply networks. The methodological approach adopts a mixed-methods, longitudinal research design. The population comprises agricultural value chain actors in three diversified agro-ecological zones, including smallholder farmers (n ? 1,200), aggregator firms (n ? 60), processors (n ? 25), and retailers (n ? 40). A stratified random sampling technique yields a sample of 400 farmer respondents, 50 firms, and 20 interviews with industry intermediaries, complemented by purposive sampling of 15 key informants from regulatory agencies and farmer cooperatives. Primary data are collected through structured household surveys, firm-level questionnaires, and semi-structured interviews conducted over a three-year period to capture dynamic responses to shocks. Secondary data include weather indices, commodity price volatility measures, and policy indicator series aligned to the study region. Instruments encompass a multi-item resilience scale grounded in the dynamic capabilities and adaptive governance literatures, survey items for sensing and responding to disruptions, and metrics for supply chain performance (lead time, fill rate, throughput, and profitability). Validity and reliability are established through content validity by subject-matter experts, confirmatory factor analysis (CFA) for construct validity, and Cronbach’s alpha coefficients exceeding 0.70 for all major scales. Data analysis employs a combination of structural equation modeling (SEM) to test the dynamic relationships among sensing, learning, adaptation, and performance; vector autoregression (VAR) to capture temporal interdependencies and impulse response functions; and thematic analysis of interview transcripts to elucidate contextual mechanisms and governance arrangements. Model specification draws on a dynamic resilience framework that integrates the Resource-Based View (RBV), Dynamic Capabilities, and Institutional Theory, with explicit incorporation of feedback loops and time-lagged effects. Key expected findings include (i) evidence that enhanced sensing capabilities and information flows significantly improve adaptive capacity, which in turn reduces disruption costs and improves performance under shock scenarios; (ii) demonstration that learning-induced routines and bilateral trust-based governance amplify resilience by accelerating response times and enabling productive diversification; (iii) identification of threshold effects where the benefits of diversification and collaboration exhibit diminishing returns beyond certain scales or institutional contexts; (iv) confirmation that policy stability and supportive regulatory environments strengthen resilience via reduction of transaction costs and facilitation of collective action. The study also anticipates heterogeneity across actors, with smallholders benefiting disproportionately from targeted information services and cooperative governance. The contribution to knowledge lies in delivering a dynamic, theory-driven framework that unifies resilience, learning, and governance in agricultural value chains, offering a measurable model with time-variant parameters that can be applied across commodities and regions. Practical implications include guiding policymakers and practitioners in designing adaptive institutional arrangements, information infrastructure, and collaborative platforms to bolster resilience. The study concludes that resilience is not a static attribute but an evolving capability shaped by sensing, learning, and governance dynamics; robust resilience arises when dynamic capabilities are embedded in inclusive institutions and facilitated by transparent information exchange. Recommendations emphasize investment in digital sensing technologies, cooperative support structures, scenario-based planning, and policy tools that incentivize short- and long-term adaptive investments by all value-chain actors.