Maritime Blockades, the Ukraine War, Tariff Trade Tensions, and Supply Chain Parameters as Drivers of Humanitarian Supply Chain Performance: The Mediating Role of Stochastic MILP Optimization Capability

Abstract

Humanitarian supply chains have been facing a growing accumulation of disruption events of war conflict, maritime blockade and trade protectionism and humanitarian aid delivery cannot rely anymore on secure channels. This research aims at examining the impact of maritime blockades, Ukraine war, supply chain parameters and tariff trade tension on humanitarian supply chain performance and tests the mediation effect of stochastic MILP optimization capability. Anchored in the Dynamic Capabilities Theory, the research proposes optimization capability as the ability of humanitarian organizations in dealing with unfavorable environment. Data was collected using a quantitative cross-sectional survey from 157 humanitarian logistics, supply chain, procurement and disaster relief professionals of Finland, Sweden, Norway and Estonia using purposive sampling method and analyzed by IBM SPSS 27 and SmartPLS 4.1 version with PLS-SEM. The proposed measurement model achieved construct’s validity with regard to internal consistency reliability and convergent validity with a total of six measurement constructs and thirty-seven indicators. Structural findings revealed that maritime blockades, Ukraine war, supply chain parameters, and tariff trade tension have significant impact on stochastic MILP optimization capability and humanitarian supply chain performance, and also confirm its mediation role between them. From theoretical perspectives, the study contributes to Dynamic Capability Theory by defining stochastic optimization as one particular reconfiguring capability at the forefront of the business model disruption process. Regarding practical perspectives, the finding recommends humanitarian agencies to use optimization tools, forecast in scenarios, conduct trainings for the staff, and employ quantitative support tools to assist decision-making. It is recommended in the discussion to address limitations based on the cross-sectional design, use of a non-random sample, relying on self-reports and exploring only four external influences. It also suggests, for future studies, to employ large and longitudinal sample, study further countries and disruption sources, explore alternatives like AI-based methods to improve efficiency or look for alternative solutions for dynamic modeling optimization.