The European Community's policy of import surveillance entails the public announcement that the Commission will collect detailed statistics on imports of particular goods either prior to or immediately after their importation, which appears unlikely to have any direct impact on trade flows, but may have an indirect effect by altering exporters' subjective assessment of the likelihood of further trade restrictions.
In Discussion Paper No. 404, Programme Director L Alan Winters considers the effects of import surveillance on the volumes, prices and origins of EC imports and casts light on the existence of `strategic trade behaviour' on the part of firms and policy-makers. Winters moves on from the existing calibration and simulation studies to test directly for the presence of strategic effects by identifying the changes in trade behaviour that would be consistent with strategic responses to import surveillance by non-member countries.
If surveillance is perceived as the precursor of a successful anti-dumping action, then exporting firms have a direct and immediate incentive to raise prices and reduce trade. Such a perception is unlikely, however, since most anti-dumping actions have been taken without previous surveillance. If surveillance is viewed as the harbinger of inevitable trade restrictions, then each firm will seek to expand its imports, in order both to maximize its sales before restrictions are imposed and to better its position in any negotiations over quotas after their imposition.
If exporters believe that surveillance is being used to test their own reactions, as seems most likely, then they will have an incentive to reduce the overall level of their trade, but each individual firm will still wish to increase its own trade at the expense of its rivals. Winters argues that the effectiveness of surveillance as a means of restricting imports will depend on the degree of cooperation achieved by the exporters surveyed, and that it will be greater when the number of countries surveyed is small, the industry uncompetitive and closely regulated by the exporting country government, and the goods sophisticated. He tests these hypotheses by comparing actual data on trade flows with those predicted by two simple `anti-mondes', modelling import behaviour in the absence of surveillance, for a variety of industries, including instances of both universal and country-specific surveillance. These two counterfactual base cases are based on the extrapolation of past trends and on an assumption of `no change' since the last non-surveyed year.
The comparison with the extrapolation of past trends supports the hypothesis that surveillance restricts imports for almost all commodity groups, while the comparison with `no change' shows no clear tendency at the aggregate level, but the differences in results for the various goods are broadly consistent with the initial hypotheses concerning the factors that determine the effectiveness of cooperation among exporters. Winters also shows that there is a corresponding tendency for the import share of a surveyed country to rise following the removal of surveillance. Although these results are statistically weak, and the number of instances of surveillance analysed small, they broadly support the hypothesis that import surveillance has detectable and lasting protective effects.

Import Surveillance as a Strategic Trade Policy
L Alan Winters

Discussion Paper No. 404, March 1990 (IT)