Research on early warning system for antidumping petition: based on panel data logit model

Purpose – Based on the American antidumping cases against China, the purpose of this paper is to construct an early warning model for Chinese exports. Design/methodology/approach – In order to overcome the drawbacks of the existing early warning models for antidumping, first, the authors screen out six most relevant indices that play a key role in US textile corporations' decision of antidumping petition against China from 2002 to 2006, then design a early warning system for antidumping petition based on panel data logit model. Findings – The regression result indicates that unemployment ratio and import‐penetration ratio significantly influence the antidumping filing decisions; when the other invariables keep the same, with the market share of China textile goods increasing by 1 per cent point, the odds ratio of antidumping petitions against China textiles increases by about 3.7 per cent. Originality/value – As far as the authors are aware there is no definite research yet about early warning system of antidumping events, and this paper aims to specifically address this issue.