Although the costs associated with the international transport and insurance of merchandise trade (also referred to as CIF-FOB margins) are an important determinant of the volume and geography of international trade, remarkably little (official) data exist. Combining the largest and most detailed cross-country sample of official national statistics on explicit CIF-FOB margins to date with estimates from an econometric gravity model, and using a novel approach to pool product codes across World Customs Organization Harmonized System (HS) nomenclature vintages, the OECD has developed a new Database on International Transport and Insurance Costs (ITIC) that aims to fill this gap. The Database shows that distance, natural barriers and infrastructure continue to play an important role in shaping regional (and global) value chains.
International Transport and Insurance Costs can be significant
With an estimated trade-weighted average of 6% for all countries over 1995-2014, the ITIC shows that the costs of transport and insurance are significant, even without taking account of the fact that the international fragmentation of production that characterises global value chains can multiply these costs. Notwithstanding the role of other factors such as relative costs of production, government policy, and just-in-time production methods, geographical distance between trading partners, and the associated transportation costs, help to explain why global value chains still retain strong regional dimensions, as witnessed in ‘Factory Asia’, and the production hubs in Europe and in NAFTA. The ITIC data show for example that inter-continental trade increases transport and insurance costs by between 2 to 4% as compared to comparable intra-continental trade.
Figure 1 further illustrates this point for imports of electrical machinery and parts thereof (chapter 85 in the 2007 HS classification). It shows that CIF-FOB margins are significantly lower for Chinese imports from Vietnam and Hong Kong than from other Asian countries (reflecting in large part proximity but also the mix of products imported) and from Brazil and South Africa. Similarly, US imports from Mexico and Canada have much lower CIF-FOB margins than those from other trading partners, as do French imports from European partners.
The resulting CIF-FOB margins also highlight how natural geographical barriers, such as the Andes mountain range in Latin America, and poor intra-regional infrastructure, such as in Africa, may impose cost barriers to the development of regional production chains (although again, other factors, including ‘behind-the-border’ constraints such as domestic regulations, will also play a significant role). The data in Figure 1, which show relatively high margins for German imports from Italy (separated by the Alps), further illustrate this. German imports from Italy have a CIF-FOB margin equal to that of imports from the US for example. Similarly, (cross-Andean) Chilean imports from Brazil, Colombia and Argentina also have relatively high margins; higher indeed than imports from Germany.
* Calculations based on the OECD ITIC database
Oil prices matter
Oil prices (fuel) represent an important component of overall transport costs, and the data (and the model) also confirm the anticipated positive effect of crude oil prices on CIF-FOB margins, as shown in Figure 2. For example, the model shows that globally, a rise in oil prices from 25 to 75 USD per barrel increases the estimated CIF-FOB margin by 1.4 percentage point (all other variables remaining constant). Similarly, a reduction in oil prices from, for example 100 USD per barrel to 50 USD (which is approximately what happened in 2015, when oil prices dropped from 93 to 48 USD per barrel) reduces the CIF-FOB margin by close to 1 percentage point.
Fragmented production can multiply these costs
International fragmentation of production means that product parts and components cross borders many times during the course of (global) production, each time accumulating international transport and insurance costs. Typically, the longer the chain the higher the cumulative costs of transport and insurance. Consider for example an intermediate good of value 100 USD exported from country A to country B, with 6% transportation costs (i.e. an import value of 106 USD in country B) that is subsequently processed and exported to Country C for 156 USD (with an additional 4% of transportation costs), where the product is finalised and exported to country D for 237 USD (with an 8% transportation cost). The import price in country D will be 256 USD, consisting of a cumulative total of 31 USD in transport and insurance costs (or 12% of the final product value), as illustrated in figure 3.
The measure explained
In the absence of detailed data on transport and insurance costs for international merchandise trade, existing research has necessarily used analytical approaches to produce estimates. Typically, either information from one or a few countries is generalised to cover all global merchandise trade flows, or bilateral mirror data from UN’s COMTRADE database of bilateral merchandise trade statistics by productis used.
The new OECD ITIC database partly follows in these footsteps. However, one of the main improvements is that the underlying model is based on the largest and most detailed cross-country sample of official national statistics on explicit CIF-FOB margins to date; covering 16 countries (reflecting nearly 20% of global imports) that currently publish or have published detailed bilateral product-level information on CIF-FOB margins. The gravity-type model that was developed takes into account the effects of distance, geographical situation (contiguous partners, partners on the same continent), infrastructure quality, oil prices, product unit values and time. A variety of robustness tests were conducted to test (and confirm) the validity of the results.
The OECD ITIC database is available on OECD.Stat, as part of the wider set of information on International Trade and Balance of Payments Statistics. The database contains bilateral international trade and insurance costs for more than 180 countries and over 1000 individual products, covering the period 1995-2014, and provides a powerful new tool to further our understanding of global value chains. The new OECD ITIC dataset also has an important direct application in the context of the OECD-WTO Trade in Value Added (TiVA) initiative.
For further information on the methodology and underlying source data used, see the forthcoming OECD Statistics Working Paper, Miao and Fortanier, 2016, Estimating the Transport and Insurance Costs of International Trade