Statistics, Anti-dumping and the Emperor’s New Clothes

Abstract

ECIPE Working Paper No. 02/2009 on ‘Ten years of anti-dumping policy in the European Union (EU)’ has raised a number of interesting issues. It identifies five tendencies based on an empirical examination of the anti-dumping statistics over the past ten years. One tendency identified is that the products targeted by EU anti-dumping actions ‘are concentrated in sectors where European comparative advantage is declining that is, raw materials, industrial input goods and textiles’. This particular tendency cannot be deduced from the paper itself. In addition, if this tendency is removed, the other tendencies do not seem to show much need for the concerns expressed. They are merely the usual consequences of normal anti-dumping procedures. Thus, the paper’s conclusion that there is a tendency of ‘targeting fair and legitimate market competitors from abroad through anti-dumping protectionism’ is not shown in the paper either. Like many conclusions based on statistics, the paper fails to exclude, or even examine, other possible causes of the supposed trends identified. Nor does it show why a normal aspect of anti-dumping, such as the fact that the duties imposed are higher than bound duties, is disquieting or even remarkable in any way. In addition, the statistics cited appear to contain factual errors. This note is a brief commentary on the ECIPE anti-dumping paper. Its purpose is to highlight some of the problems in the approach taken. It shows that the statistical examination of the last ten years of EU anti-dumping practice does not support the bullet points set out in the paper’s abstract section or the conclusions themselves.