2.2.4 Rules of Origin
Complicated and restrictive rules of origin are often pointed out as a major reason for the low utilization of preferences. Rules of origin are needed to prevent trade deflection, whereby products from non-beneficiary countries are redirected through beneficiary countries to exploit the preferences that are available. Still, there are at least two types of costs associated with restrictive and complicated rules of origin. First, there are administrative costs for providing the necessary documentation to prove that the rules have been complied with. Cadot et al. (2006) estimate that for the relevant EU rules of origin, these administrative costs represent 6.8 percent of the traded goods’ value.6 From the discussion on preference margins above, it follows that if margins are not wide enough to cover these costs, exporters will not have any incentive to request preferential treatment, and will (if at all) export under the MFN tariff.
A second type of cost related to restrictive rules of origin is concerned with the production process. If an exporter wants to obtain preferential treatment, there are strict limits to the possibilities of sourcing inputs from the lowest cost location in other countries. For producers in larger markets, this may lead to unnecessarily high production costs, which will have to be weighed against the potential extra gains from preferential tariff treatment. For producers in small markets, and many preference receiving countries will certainly fall into this category, local sourcing of inputs may not even be possible, leaving the firm with the option of not producing, or importing the necessary inputs and then exporting the good under the MFN tariff. Depending on how they are formulated, rules on cumulation of origin may ameliorate this problem, but in many cases, the countries from which inputs can be imported, without affecting the origin of the final product, are not low-cost locations. For more on the importance of rules of origin, see e.g. Brenton (2003), Augier et al. (2005), Cadot et al. (2006) or Brenton and Manchin (2003).
7