HR capabilities and AI adoption both correlate with greater supply‑chain agility, which in turn correlates with stronger resilience among Indonesian rice logistics providers; agility mediates the effects of human and digital resources on resilience.

Main Finding

Both Human Resource (HR) capabilities and Artificial Intelligence (AI) adoption positively increase Supply Chain Agility (SCA), and higher SCA in turn significantly improves Supply Chain Resilience (SCR). SCA significantly mediates the effects of HR capability and AI adoption on SCR. In short: human and digital resources build resilience by enabling greater agility.

Key Points

• Context: Study of rice supply chain logistics in Indonesia using survey responses from 210 Logistics Service Providers (LSPs).
• Core relationships:
  • HR capability → positive effect on Supply Chain Agility (SCA).
  • AI adoption → positive effect on SCA.
  • SCA → positive effect on Supply Chain Resilience (SCR).
  • SCA mediates HR capability → SCR and AI adoption → SCR links.
• Theoretical framing: Extends Resource-Based View (RBV) and Dynamic Capabilities Theory (DCT) to rice logistics by showing how organizational resources (human and digital) are transformed into resilience through agile capabilities.
• Operational significance: Agility enables better responses to seasonal harvest swings, distribution variability, and environmental uncertainties common in rice logistics.

Data & Methods

• Data: Cross-sectional survey of 210 Logistics Service Providers involved in Indonesia’s rice supply chain.
• Analysis: Quantitative approach testing direct and mediated relationships among HR capability, AI adoption, SCA, and SCR. (The paper reports statistical tests of mediation; typical practice here is to use multivariate regression or structural equation modeling to test these paths.)
• Sector and scope: Single-country (Indonesia) and single commodity (rice) focus; respondents are logistics providers rather than producers/retailers.

Implications for AI Economics

• Returns to AI should include resilience benefits, not only efficiency/productivity gains. AI adoption generates an indirect welfare/productivity effect via increased agility that reduces disruption costs.
• Complementarity of human capital and AI: Investments in AI are more effective when paired with workforce capabilities (training, decision rights, cross-functional skills). Models of AI diffusion and firm investment should incorporate complementarities and skill-upgrading costs.
• Valuation and cost–benefit analysis: Economic evaluations of AI in supply chains should quantify avoided disruption costs, inventory smoothing benefits, and faster recovery times, in addition to throughput improvements.
• Policy considerations:
  • Targeted subsidies or tax incentives for AI adoption may be justified if public benefits from improved food security and reduced disruption spillovers are large.
  • Workforce development (training, reskilling) amplifies AI’s economic value; policy should combine digital adoption support with human-capital programs.
  • Data-sharing platforms and interoperable standards can magnify network-level resilience gains from AI.
• Research suggestions for AI economics:
  • Estimate the monetary value of the resilience premium from AI-enabled agility (panel or quasi-experimental designs to identify causal effects).
  • Model macro supply-chain resilience incorporating heterogeneity in AI adoption and HR capabilities to study systemic risk reduction.
  • Explore distributional impacts: which firms (by size, capital access) gain most from AI-driven resilience, and labor-market consequences for logistics workers.
  • Dynamic and longitudinal work to capture adaptation over time and potential diminishing returns or complementarities as AI diffuses.

Limitations to note (for economic interpretation): cross-sectional design limits causal claims; single-commodity, single-country sample limits generalizability; “AI adoption” is treated at an aggregated level—economic models would benefit from disaggregating AI types, functionalities, and adoption intensities.