52010DC0077

EN

Brussels, 9.3.2010

COM(2010)77 final

COMMISSION WORKING DOCUMENT

The future of 'Duty Drawback' in the rules of origin of EU's Free Trade Agreements

COMMISSION WORKING DOCUMENT

The future of 'Duty Drawback' in the rules of origin of EU's Free Trade Agreements

On Wednesday 7 October 2009, a College debate took place on the EU-Korea FTA. In this context, the President invited DG Trade and DG Taxud to draft a common reflection note on the future of 'Duty Drawback'.

I. Introduction

'Duty Drawback' (hereafter referred to as 'DD'), as defined in the protocols of origin contained in preferential trade agreements, is the refunding, remission or non-payment - partial or complete - of customs duties or equivalent charges on foreign inputs (raw, semi-manufactured materials or components) that are used in the production of a final product which is exported to a third country [1].

DD is used in a non-preferential environment by the EU and many (if not all) countries. It puts exporting industries in different countries on an equal footing despite differences in the level of import duties. As to the import side, countries maintain the possibility to protect their domestic market through the MFN duties applied. DD is allowed under the provisions of the WTO Subsidies Agreement.

In a preferential environment, the EU has traditionally followed a practise of prohibiting DD in many of its free trade agreements [2], although it has permitted it partly or fully vis-à-vis many developing countries [3]. However, some of the EU FTA negotiating partners object to such a prohibition, and – as most recently seen in the case of Korea - the treatment of DD is becoming one of the key issues in these negotiations. An examination of the EU policy on DD is therefore required, and this note intends to provide some basic considerations and suggestions for a policy line on the issue, for subsequent discussion with the Council, the EP and stakeholders.

II. The impact of prohibiting or allowing DD in free trade agreements

a) Negative effects of DD

Traditionally, the liberalisation of international trade concerns the reduction or complete removal of existing trade restrictions. General removal of existing trade barriers on an erga omnes basis is the most direct way to achieve free trade, eliminating and reducing customs duties along with investment protection, opening of services market and so on. Rather than going for full trade liberalisation, many countries have chosen to take intermediate steps such as allowing DD, aimed at correcting the impact of import duties on exports and at promoting them.

DD can indeed be regarded as a compromise to trade liberalisation as it has some negative effects on it. Although there is no single view in the economic literature, some economic studies have found for example that DD decreases the incentive for exporting producers to lobby against high tariffs on their inputs which has an opposite effect on free commerce. DD may reward the maintenance of protectionism: it can allow a country to keep on protecting its producers of intermediate materials whilst reducing such protection selectively in order to promote its exports of finished products. While DD can have a positive impact on the export competitiveness of a country, it could also lead to exports with low domestic value and relatively high foreign content.

DD is however even more questionable when it is maintained in a full liberalisation process of a Free Trade Agreement (FTA). The most notorious negative consequence of allowing DD then is probably the following. When DD is allowed in a free trade area, a producer of FTA partner country A can source on a duty free basis from third countries for exports to FTA partner country B, whilst their competitors in country B have to pay the applicable MFN duties when sourcing from third countries for products sold on their own domestic market.

Example: a company from an FTA partner exporting fabrics to the EU would benefit from duty drawback for the fibres it imports from a third country to produce the fabrics; in contrast, an EU fabric manufacturer selling in the EU would not benefit from the reimbursement of duties paid on the fibres it may import from third countries for the manufacturing of those fabrics. If the share of such imported fibres in the value of the fabric is, say, 25 % for the EU manufacturer, the potential competitive advantage of that exporter in the EU (expressed as a percentage of the EU manufacturer's total price of the finished fabric) would be the result of multiplying by 25 % the import duty rate in the EU of such fibres (4 %), so some 1 % of the value of the fabric [4].

DD can thus lead to unbalanced competition in the market of the importing country which can have negative effects on domestic industry with possible consequences on employment. Additionally, allowing DD would also enable third countries to participate in a certain way in the benefits of the FTA, as regards trade in intermediate products and materials. Whilst a prohibition of DD would promote greater use of intermediate products and materials from the FTA partner countries via the use of cumulation possibilities, allowing DD would allow the treatment of inputs from third countries on a par with inputs originating from the FTA: both inputs would in reality enter the FTA partner at 0 % duty when re-exported after incorporation into a final product. Therefore, although DD is legally not an "export subsidy" it should not be regarded as a kind of export incentive that a free trade area should promote.

b) Parameters that determine the impact of DD

DD is a complex economic tool. The impact of allowing or prohibiting DD on the exporters, on competition in the market of the importing country, and on trade in intermediate products, depends on many factors, which are illustrated below [5].

a) As regards the impact on the competitive position of exporters of finished products, the relevant factors are the MFN duties that the exporting country applies on intermediate materials combined with the degree of foreign sourcing in the same country: the higher the MFN duties and the greater the use made of foreign inputs (depending on RoO within the FTA) are, the greater the impact of DD could be on the final product exported. This is one of the main reasons why, for example, traditionally in most countries in Asia where MFN duties are relatively high, allowing DD plays a much more important role than in US or EU where MFN duties are comparatively relatively low.

Example: Ruritania applies a 14 % tariff on the imports of fibres from third countries; if imported fibres account for 25 % of the value of the fabrics exported by a Ruritanian company, the value of the duties paid on such imported fibres would be 3,5 % (25 % of 14 %) of the value of the Ruritanian exporter's fabric. As seen above, for an EU exporter the value of the duties imposed on the import of such fibres would be 1 % of his own fabric's price.

b) As regards the impact on competition in the finished products market in the importing country, the most relevant aspect is the MFN duty that the importing country applies on intermediate materials: the higher the MFN duties that a preferential country applies on the import of intermediate materials and the higher the use by domestic manufacturers of such imported materials from a third country, the bigger potentially the competitive disadvantage of domestic manufacturers of finished products as compared with exporters from an FTA partner country that benefits from DD. .

To follow the same example above, the fabric manufacturer in the EU importing fibres from third countries would have to pay import duties on such fibres corresponding to 1 % of its selling price, regardless of whether duty drawback is allowed or prohibited. Thus, in case duty drawback is allowed in an FTA, the competitive advantage of an exporter from an FTA partner country, say, Ruritania over a domestic EU producer would be 1 % of the value of the fabrics.

If we look at the situation in Ruritania, a Ruritanian fabrics manufacturer importing the same amount of fibres from third countries would have to pay duties on imported fibres worth 3,5 % of the value of the fabrics: this would constitute the advantage for an EU manufacturer exporting such fabrics to Ruritania in case duty drawback were allowed.

However, in case duty drawback were prohibited in an FTA between the EU and Ruritania, the Ruritanian exporter to the EU would be faced with a negative advantage of 2,5 % of the value of the fabrics (3,5 % in duties it would pay on the fibres imported from third countries minus 1 % of the duties paid on the same fibres by its EU competitor); conversely, the EU exporter to Ruritania would have a competitive advantage in Ruritania of 2,5 % (3,5 % of duties paid by Ruritanian manufacturers minus 1 % paid on such imports by EU exporters).

c) To switch from trade in finished products to trade in intermediate materials, the key factor that determines the impact on trade of intermediate materials and components between the EU and our FTA partners is the level of import duties applied by both sides on these materials and components that will be eliminated under a FTA: the higher such duties, the more trade between FTA partner countries in those materials will be encouraged, to the detriment of third country suppliers.

Continuing with the same example above: in case duty drawback were prohibited, a Ruritanian company importing fibres from third countries to manufacture fabrics for export to the EU would face an additional cost of 3,5 %. This cost would constitute an incentive to source fibres either domestically or from the EU rather from other third countries. If duty drawback is permitted, such an incentive would disappear.

Another factor that plays a role on the impact of DD is the degree of restrictiveness or leniency of the rules of origin under the FTA: the product specific rules of origin, which establish the 'sufficient working or processing' that needs to be carried out on non-originating materials to obtain an 'originating' product that can benefit from preferential tariff treatment, determine the level of permissible foreign sourcing and therefore the maximum amounts of duty that may have to be paid or that can be drawn back if the final product is exported.

Example: for fabrics, the standard EU preferential rule or origin is manufacture from fibres, which implies two transformations – spinning and weaving/knitting – which typically account for about three quarters of the value added of the fabric, which have then to be provided domestically. This limits foreign sourcing and therefore the scope and impact of allowing or prohibiting duty drawback. The impact of allowing duty drawback would, however, be greater in sectors where the rule is more relaxed – for instance for some chemicals, metals or machinery, where materials from third countries can account for 40 %-50 % of the value of the product and the product is still considered as domestically produced. In those cases, prohibiting duty drawback could impose a significant additional constraint on the sourcing possibilities theoretically allowed by the rule of origin.

The importance of DD is also depending on the economy of the country. Indeed, as regards trade in finished products, the impact for the EU of allowing DD is relatively limited, as the EU generally applies very low tariffs on the import of intermediate products/materials and, because of the EU's very varied industry base, the need for manufacturers to rely on foreign sourcing could sometimes be limited, although globalisation is increasing the dependence of EU manufacturers on external suppliers. Such an impact may be greater in the EU trading partners, which often apply higher import duties and which have, because of their size, to rely more on imported materials. In other words, in closed, protectionist markets with high MFN duties on parts and in markets where foreign sourcing is important, allowing DD has more impact than in a more liberal customs territory with an integrated economy.

All in all, the potential negative impact of allowing DD constitutes a good reason in favour of seeking, as a matter of general policy, to prohibit DD in free trade areas.

c) Problems related to the prohibition of DD

Even if the prohibition of DD still seems the preferred option in an FTA context, this policy can create problems in its application.

Before a preferential arrangement is made, international commerce between the future partner countries of the free trade area is based on a non-preferential environment. In such trade relations, DD is always used or at least available. If this situation evolves towards a prohibition of DD in a free trade area, the maximum benefits one may expect from the tariff reductions agreed within the FTA could be significantly reduced.

Whilst a prohibition of DD would imply less benefits as compared with allowing DD, DD does not constitute per se a benefit for an FTA partner additional to the elimination of import duties. For instance, under MFN treatment the duty to be paid for the import into the EU of a car is 10 %; if the import price of the car is 10.000 €, the duty saved under an FTA would be 1.000 € as compared with the current MFN situation. Thus, assuming that a car manufacturer in, say, Ruritania contains 20 % of parts and components imported from non-preferential third countries, and as the import duties in Ruritania of such components is on average 16 %, the duty paid on such components would be 320 €. In case an FTA prohibits DD, the net benefit for the exporter would then be 680 € (1.000 € - 320 €) instead of 1.000 €, thus a reduction of 32 % of the benefit of duty elimination. Should the share of imported parts used by the Ruritanian car maker be the maximum permitted by the EU standard rules of origin (40 %), then the net benefit would be reduced to a bit more than a third of the EU import duties (1.000 € - 640 € = 360 €) [6].

For some products, the producers of a partner country could even prefer to use MFN tariff duty, because the cost of renouncing to duty drawback could be higher than the applicable MFN duty.

To which degree these problems arise and their relative importance as compared with the impact of allowing DD depend, as indicated above, on different economic parameters like respective MFN tariffs of both countries, level of reduction/abolition of tariffs within the FTA, and the level of 'leniency' of the 'rules of origin' and thus maximum foreign content allowed and the economies of both partners.

Whereas a prohibition of DD for an FTA would avoid certain adverse effects of DD on competition in the home market and promote bilateral trade in intermediate products/materials, in preferential trade arrangements with a developing country the additional development objective may bring an additional consideration. Indeed, a prohibition of DD in respect of developing countries might not always be desirable. It would for example be difficult to conceive that a developed country would give an LDC, on the one hand, duty free access to its market in order to assist and support the economy of the LDC while, on the other hand, it would limit the economic benefits resulting from such a free access by a prohibition of DD for foreign inputs used by the producers of the LDC, should both partners enter into a FTA.

This could be the case, for instance, if the EU enters into an FTA with a group of countries which includes LDCs. LDCs being EBA beneficiary countries, they can export to the EU duty free whilst benefiting from duty drawback. If under such an FTA duty drawback were prohibited, exporters from the LDCs concerned would have to pay duties for the inputs imported from third countries that are incorporated in their products exported duty free to the EU, something that is not required under EBA.

Similarly, since the share of exports of developing countries to the EU subject to EU import duties is relatively limited [7], a prohibition of duty drawback, that would imply a limitation of the benefits of import duty elimination, would thus reduce further their incentives to enter into a FTA with the EU.

Furthermore, developing countries are currently entitled to apply DD for all their exports to the EU under unilateral preferences (GSP); in fact, a large share of the exports of many developing countries to the EU are already duty free or subject to very low duty rates (sometimes called ‘nuisance duties’) under the GSP whilst benefiting from DD. In those cases, the resulting situation under an FTA would be worse than currently for those products benefiting from zero duty under the GSP [8].

To give an example: the EU MFN duty for motor vehicle engines is 4,2 %, but after the duty reduction under the GSP Regulation this duty is brought to zero for all developing countries. Thus, a developing country can export those engines under the GSP to the EU duty free and with duty drawback. If under an FTA with a developing country duty drawback were prohibited, the FTA treatment for the engines exporter from that developing country would be worse than the current one: assuming that the value of the engine is 2.000 €, and that the share of parts imported for the manufacture of the engine was 20 %, if the average import duties on such parts is, say, 16 %, the engines exported under the FTA from that developing country would be facing an additional cost of 64 € as compared with its exports under the current rules (20 % of 2.000 € x 16 % = 64 €, which would be equivalent to a duty of 3,2 % on the engine). Should the share of imported parts used reach the maximum allowed under the EU standard rules of origin (40 %), then the additional cost would be double, 128 €, which equals 6,4 % of the value of the export and which is higher than the EU MFN import duty.

III. Conclusions

Allowing DD is problematic in the context of a free trade area, as it introduces the possibility of competitive distortions between the participating countries, whereas goods and services should be traded on their respective markets on the basis of comparative advantages. There are thus good reasons in favour of seeking, as a matter of general policy, to prohibit DD in free trade areas.

However, since the prohibition of DD may also create problems in its application for our partner countries, some limited concessions to this general policy line may be considered in exchange of adequate concessions from the other party, and on the condition that 'rules of origin' (RoO) would fulfil the needs of the EU industry. Thus, a limited possibility for exceptions can be envisaged, based on a comprehensive assessment of the following criteria:

a) The extent to which the RoO of the FTA are satisfactory for the EU including for its industry. Indeed, having the right product-specific RoO may be as significant or more in economic terms than prohibiting DD. On the one hand, the RoO in our FTAs should require an adequate level of transformation and/or value added, in order to promote that the benefits of the FTA accrue primarily to the FTA partners. On the other hand, it would be highly desirable that the FTAs have the same or similar RoO, as it is not practicable for EU industry to adjust foreign sourcing to rules different depending on the market of destination. Thus, the acceptance of rules of origin as close as possible to the EU standard rules of origin is an important and relevant factor to take into consideration. Once acceptable RoO are agreed that overall fulfils the needs of EU industry, it could thus be acceptable to show flexibility on DD, although preferably within certain limits, taking into account the other criteria below.

b) The likely impact of allowing - or as appropriate prohibiting - DD, both in terms of competitive conditions in the EU market as well as on EU exporters, should be assessed and taken into account in evaluating the overall balance of the agreement. Such an analysis, including a quantitative assessment that should start early in the negotiating process and will in any case precede the decision on the conclusion of the FTA, will look into the impact on trade, production, investment and employment, on the use of the cumulation possibilities under the FTA, and on affected developing countries.

c) The ambition of the FTA in terms of market access and the extent to which it meets the interests of EU industry. Elements that may be considered, on a case by case basis, to assess such flexibility are the extent to which more advanced countries assume ambitious trade liberalisation commitments as a result of the FTA that constitute an overall satisfactory outcome for EU industry. Similarly, adequate market access conditions should be provided for EU exporters of intermediate products to whom it would not otherwise accrue some FTA benefits.

d) Development considerations, including the extent to which a FTA negotiating partner is already exporting duty free (or under so-called "nuisance duties") to the EU with DD, the impact of prohibiting DD on the benefits and on the incentives for them to enter into an FTA, and the effects of allowing DD on the use of domestically produced intermediate materials. The degree of development of the third country concerned should also be taken into consideration.

As to the possible concessions on DD, an example of limited flexibility could be to restrict DD to the difference between the average MFN rates applicable to intermediate inputs in the partner country and the EU respectively when the MFN duty rates of a future partner country are relatively high and relatively low in the EU. This could be attempted generally or for some sectors (or even products as necessary) where the impact of DD due to the differential MFN duties may be more significant. This would aim at maintaining a level playing field for the industries of both countries within the free trade zone. In other cases, time limitations or other limits could be considered. In the case of developing countries, a less demanding position would be justified in the light of a development-friendly policy (the degree of which may depend on whether it concerns an LDC or GSP+ country or not), and taking into account that they already benefit from DD for their exports to the EU under the GSP.

[1] To illustrate the mechanism, one could imagine that a Japanese car engine is charged with 10% duties when entering Korea. When this car engine is assembled into a Korean car which is then exported to the EU, the Korean car manufacturer could 'drawback' the 10% duties he paid on the foreign input/Japanese car engine.

[2] The EU is not the only WTO Member which has prohibitions of DD in free trade agreements. Other countries prohibit it as well in some of their FTAs, although it can depend sometimes on their partners: e.g. Mexico and Chile sometimes prohibit DD and sometimes allow it. Within Mercosur, DD is prohibited for certain cars. For the US, the most important FTA is NAFTA where DD is prohibited. Following a policy discussion in 2003, the last years the US has been deciding on a case-by-case basis, and DD has been allowed in the FTAs negotiated, except for the US-Chile FTA where DD was prohibited; it has to be noted though that a majority of these FTAs concerned developing countries (i.a. Morocco, Oman, Bahrain, KAFTA, Colombia, Jordan, Peru) - with the exception of Israel and Australia and the not-yet adopted US-Korea FTA – and the EU also applied flexibilities in the past for developing countries. In US-Singapore FTA, DD has no real economic consequences as Singaporean MFN is 0 for most products.

[3] DD is prohibited, subject to transitional arrangements, in the FTAs with developing countries like Mexico and Chile and all the Euro-Mediterranean FTAs. However, it has been permitted in trade preferences linked to development purposes: under the GSP rules applicable to all developing countries and for ACP countries under Cotonou, in the EPAs, and in the Trade, Development and Co-operation Agreement (TDCA) with South-Africa.

[4] To note that in most other industrial sectors and products the rules of origin may allow for a greater share of foreign content (up to 50 % in certain cases), and the figures and impact of allowing or prohibiting duty drawback could be higher depending on the actual use of the foreign sourcing possibilities.

[5] For the sake of simplicity, in the examples provided no account has been taken of additional factors such as whether the materials imported from third countries may be coming from preferential countries not subject to import duties or of the possibilities of cumulation. The calculation of the value of duty drawback has also been made for simplicity reasons on the price of import of the finished product rather than on the cost of manufacture, and therefore is somewhat overstated.

[6] One may however argue (see section II.a) above) that it is legitimate in a FTA to prevent exporters, through the prohibition of DD, from fully benefiting from the duty elimination since otherwise exporters would be placed in a better position than their competitors operating on the importing country's domestic market.

[7] Roughly around three quarters of the exports of the developing countries with which the EU is currently negotiating FTAs are subject to zero or very low duties in the EU under MFN and the GSP. Thus, in 2008 58 % of total imports from India came duty free into the EU and a further 7 % under nuisance duties (≤ 3%); 65 % of imports from ASEAN came duty free and a further 5,6 % under duties below 3%; 80 % of imports from Central America also came duty free; 77 % of Colombian exports came duty free, and 95 % from Peru. Thus, prohibiting DD would imply limiting the benefits of the FTA for about a quarter of the exports of these countries to the EU; however, these concern products subject to the highest tariffs in the EU and which are normally more sensitive.

[8] The shares of such exports subject to zero duty under the GSP were in 2008: 24 % for India, 10 % for ASEAN countries, 20 % for Central America, 13 % for Colombia and 23 % for Peru. In these cases, FTA treatment would be worse in case of a prohibition of DD.

--------------------------------------------------