National Freight Demand Modelling: A Tool For Macrologistics Management

Purpose The purpose of this paper is to present the results of South Africa's national freight demand model and related logistics cost models, and to illustrate the application of the modelling outputs to inform macrologistics policy.Design/methodology/approach Spatially and sectorally disaggregated supply and demand data are developed using the input-output (I-O) model of the economy as a platform, augmented by actual data. Supply and demand interaction is translated into freight flows via a gravity model. The logistics costs model is a bottom-up aggregation of logistics-related costs for these freight flows.Findings South Africa's logistics costs are higher than in developed countries. Road freight volumes constitute 80 per cent of long-distance corridor freight, while road transport contributes more than 80 per cent to the country's transport costs. These challenges raise concerns regarding the competitiveness of international trade, as well as the impact of transport externalities. The case studies highlight that domestic logistics costs are the biggest cost contributor to international trade logistics costs and can be reduced through inter alia modal shift. Modal shift can be induced through the internalisation of freight externality costs. Results show that externality cost internalisation can eradicate the societal cost of freight transport in South Africa without increasing macroeconomic freight costs.Research limitations/implications Systematic spatially disaggregated commodity-level data are limited. There is however a wealth of supply, demand and freight flow information collected by the public and private sector. Initiatives to create an appreciation of the intrinsic value of such information and to leverage data sources will improve freight demand modelling in emerging economies.Originality/value A spatially and sectorally disaggregated national freight demand model, and related logistics costs models, utilising actual and modelled data, balanced via the national I-O model, provides opportunities for increased accuracy of outputs and diverse application possibilities.