Lies, damned lies, and statistics: Looking at re-exports
An increasing part of international trade involves re-exports. In the Netherlands, e.g., re-exports constituted 54 per cent of total exports of goods in 2015. Re-exports are products imported into and exported from an intermediary country. Activities in the intermediary country may involve, e.g. repackaging or adding user manuals in the language of the destination country, or trade mediation so as to reduce informational barriers. The intermediary country assumes short-term ownership of the products, but the goods are exported without having undergone any significant industrial processing. Thus the re-exporting country does not actually produce the exported products.
Re-exports are large in relative terms in office machinery and computers. Important re-exporting countries include the Netherlands, followed by Germany, Belgium, and the UK. Re-exporting in these countries can be explained by the presence of (sea) ports and attached transport and logistics infrastructure, and/or an overall openness to trade.
Re-exports present a main concern in trade statistics, because (a part of) trade is not registered to its proper origin or final destination. In Lankhuizen and Thissen (2019), we derive bilateral trade flows corrected for re-exports for a large number of countries at a detailed product level. Our method consists of estimating the most probable origin and destination of re-exports in the international trade system as a whole. Next, trade from an origin A to the re-exporting country B and trade from the re-exporting country B to the final destination C is changed to trade from A to C (see the figure below). We use a constrained non-linear optimization. This means that the resulting data should satisfy a number of definitions as closely as possible, given a set of restrictions or conditions.
a. Bilateral trade including re-exports b. Correction for re-exports
We illustrate the importance of properly taking into account re-exports. Not assigning trade to its proper origin and destinations causes estimates from gravity estimations to be biased, because the data partly reflect the characteristics of the re-exporting countries. In particular, we find that the effects of geographical distance, cultural distance and adjacency are underestimated. Furthermore, we illustrate the importance of properly taking into account re-exports for two current events in international trade policy, the renegotiation of NAFTA and Brexit.
The US administration perceives trade deficits as the result of unfair trade practices. The main focus of the administration’s trade policies is therefore to reduce them. Accordingly, the US have pulled out of the Trans-Pacific Partnership (TPP), pushed ahead with tariffs on billions worth of Chinese imports, while extending the application of steel and aluminum tariffs to Canada, Mexico, and the European Union. The administration has also renegotiated NAFTA (the free-trade agreement between the US, Canada and Mexico). Our results suggests that the US trade policy may be based on erroneous premises. Focusing on NAFTA, we show that the US actually runs a trade surplus with Canada, and the trade deficit with Mexico is reduced by half at least, if bilateral trade is calculated the right way. Taken together the US run a large trade surplus with its partner countries in NAFTA. We also show that there may be additional losses to the UK economy from Brexit that are not taken into account in existing studies. British re-exports total over $15 billion on average between 2000 and 2010 (approximately 5 per cent of total UK exports). A back-of-the-envelope estimate of these potential maximum direct costs of Brexit amounts to $539 million. Sectors that are likely to be affected include wholesale trade, road haulage, and storage and transport, but also business support services linked to wholesale, e.g. administration, computer services, distribution, and advertising. A Brexit may also impede exports to e.g. the US and China, as some of the flows are re-exported via the EU.
Lankhuizen, M. B. M., & Thissen, M. (in press). The implications of re-exports for gravity equation estimation, NAFTA and Brexit. Spatial Economic Analysis.
 The agreement is now the United States-Mexico-Canada Agreement, or USMCA.