Why are some of the WTO stats different than those in the GTP?

There are often small differences between the Global Trade Protection Report (GTP) and the statistics on anti-dumping provided by the WTO.  My experience of preparing these statistics over the past 8 years is that such differences often arise.

One reason for such differences is that the source of data used for the GTP is slightly different to that used by the WTO.  The latest statistics in the GTP report are strictly based on the twice yearly notifications made to the WTO by Members.   From speaking to the WTO Secretariat about their approach in the past, my understanding is that the WTO uses both the Member notifications as well as other information received by the Secretariat that is not necessarily publicly available.

Counting anti-dumping cases is never a precise science. For example, prior to the new statistics published by the WTO in November 2006, the tables on the WTO website showed the number of AD initiations for the EC and India as 24 and 25 respectively. The new statistics produced on 27 November 2006 not only provided the numbers for the first six months of 2006 but some of the earlier numbers are also revised. For example, the EC and India initiations for 2005 are now 25 and 28.

Such revisions are inevitable. Information on anti-dumping is never perfect. Notifications from Member States can contain mistakes, some of which come to light at a later stage and are corrected. This can result in the WTO statistics being revised to reflect such corrected information.

In this context, it is not surprising that differences can emerge between the GTP and WTO statistics. However, the strict methodology of the GTP report is initially to rely solely on the Member State notifications. The assumption is made that the notifications are correct. Even if changes are made later, they are usually small and do not affect the overall trends identified and consequent conclusions. 

Where there are small differences in the figures, they do not in any way affect the overall trends and conclusions. Given the imprecise science of counting AD cases, the actual figures may never be 100% correct. However, if they are different, such differences are small, allowing accurate analysis and conclusions to be determined from the data. 

Please do not hesitate to contact me (cs@antidumpingpublishing.com) if you have any queries about the data.

Cliff Stevenson
February 2007