SUMMARY In this paper we raise questions regarding the appropriateness of the traditional gravity model for evaluating the existence of trade creation and trade diversion in the context of regional trade arrangements. We demonstrate that existing methods fail to separate multilateral and allocation effects, and are therefore unable to say anything useful about either. Further, we point out that omitted variables-perennially unobservables-further contribute to the contamination of the results. Bearing these criticisms in mind, we provide a new methodology for assessing the effect of preferential trading arrangements on the trade volumes, and hence, the welfare of member and non-member countries. We offer a test that separates these effects, and employ it in an assessment of the EC. We find that the allocation effects of the EC caused 5 of its 11 members to divert trade toward bloc members, and 2 of its members to divert trade away from the EC. We also find that the EC increased multilateral trade for both members and non-members.
Stored only when both papers are in the corpus — 89 citations exist upstream.