Logistics costs in Norway: comparing industry survey results against calculations based on a freight transport model

Abstract

This paper presents estimates of logistics costs in Norway. Two different methodological approaches are taken: (i) use of the national freight transport model for Norway, in which logistics costs as a share of gross domestic product (GDP) are obtained from national freight flows between municipalities in Norway and from foreign trading partners and (ii) use of a survey of industry representatives, where the results are aggregated to the macro level and yield logistics costs as a share of GDP. The transport model includes detailed cost functions for transport and other logistics cost components along with a module for optimal shipment size, frequency and mode choice. Although the two approaches are quite different, we find almost identical estimates of transport costs. For the other logistics cost components (warehousing, capital costs, insurance, wastage, packaging and administration), the survey-based approach yielded slightly higher estimates, indicating that the freight transport model does not cover all logistics cost components.

Keywords:

• logistics costs
• freight transport model
• survey
• transport costs

Acknowledgements

This study was funded under the SMARTRANS Programme of the Research Council of Norway and co-funded by the Norwegian Public Roads Administration through the project “Logistics in Norway” (183742). We have benefited enormously from the extensive comments of two anonymous referees of this journal. We also thank the following persons for contributing to the project: Olav Eidhammer, Stein Erik Grønland, Harald Hjelle, Øystein Hop, Bjørn Gjerde Johansen, Kjell Werner Johansen, Jan Audun Larsen, Jarkko Lehtinen, Kent Lumsden, Heidi Christine Lund, Anne Madslien, Morten Steen Petersen, Toril Presttun, Hans Silborn and Øystein Strandli. We acknowledge Lauri Ojala and Tomi Solakivi at the Turku School of Economics for giving us access to the main results from Finland State of Logistics. All remaining errors and omissions are entirely our own responsibility.

Notes

1. The supplier (exporter) reports freight (trade flow) costs exclusive of freight and insurance (fob), while the purchaser (importer) reports freight (trade flow) costs inclusive of freight and insurance (cif). Thus, comparing fob and cif from trading partners yields an indirect measure of transportation costs (Yeats 1978).

2. OD are the places where the freight is loaded, unloaded and eventually reloaded. For example, in a transport chain where rail is used and the distribution is with truck at both ends, the PC flow is divided into three OD flows: the first is from place of production to the first rail terminal by truck; the second is between two rail terminals by rail and the third is from the second rail terminal to the place of consumption by truck.

3. Each cost component is based on a bottom-up approach.

4. The definition of capital cost in our survey is equivalent to the sum of the two capital cost components in the freight transport model: the capital costs of goods in transit and the capital costs of the inventory. In NOU (1988), the logistics costs of exported goods (exempting offshore exports of oil and gas) were defined as comprising transportation (57%), warehousing (33%), packaging (5%), insurance (3%) and inventory/capital costs (2%) and were found to constitute 16.4% of the export value. Bjørnland and Lægreid (2001) estimated logistics costs in Norway in 1997 (exempting offshore activity) comprising transportation (64%, that is 6% transportation time costs and 58% direct transport costs), warehousing (18%), packaging (14%) and insurance (4%), finding that logistics costs constituted 10.4% of GDP (compared to 12.2% in 1990).

5. One might argue that in-house logistics costs are not directly comparable to 3PL prices, if the latter includes a profit element. However, assuming a competitive market, where firms do consider buying a service or carrying it out by themselves, we still find that the comparison has relevance.