What we do



Sign up to our newsletter.

Follow ODI

EPAs: Distinguishing what we know from what we don’t know

Written by Chris Stevens

We always knew that the final few days before the EU’s self-imposed deadlines for initialling Economic Partnership Agreement (EPA) texts to forestall the application of generalised system of preferences (GSP) tariffs on 1 January would be frenetic. And so it has proved to be. This blog has been written at 17.00 GMT on 30 November and aims to make sense of the information available at this moment. It concentrates on the two agreements believed to have been initialled: by some Southern African Development Community (SADC) states and by the Eastern African Community (EAC).

The big issue
The key task is to work out how far African, Caribbean and Pacific (ACP) signatories have completed the EPA negotiations (or limited their freedom of manoeuvre in any that remain to be completed), and how far the agreements are empty shells that serve the purpose only of allowing the EU gracefully to back down from its threat to re-impose tariffs.

This is important both for the signatories and for the others: if the media reports are to be believed, the ‘agreements’ that have been initialled are very extensive and lock countries into substantial, fast liberalisation; other ACP states may take fright. If the press releases present a true picture they may be right to take fright; if not, they may fail to take up a low-cost opportunity to sidestep January’s threatened tariffs.

What we know …
… is not a lot. We have seen a text of the SADC agreement but only headline figures for EAC. Any conclusion may need to be amended radically as new information comes in. But the picture so far is one of empty shell, face-saving texts, not tightly binding ones.

The SADC text
We have a copy of the text that was initialled on 25 November by three of the five members of the Southern African Customs Union (SACU) and by Mozambique. It contains neither a commitment on the share of trade that will be liberalised nor a timetable for liberalisation, and it has no detailed annexes identifying precisely which goods will be liberalised. It does commit the parties to produce such lists by 6 December – so that is a date to watch.
The really important thing about the SADC text is that it has not yet been initialled by South Africa or Namibia, despite their having been given a deadline of noon yesterday (29 November) to do so. Instead, the South African Government has written to the Commission expressing its refusal to follow the EU’s timetable.

The reason that this is important is that the terms of the SACU Agreement (Article 31.3) ban the three signatories (Botswana, Lesotho and Swaziland) from agreeing a preferential trade deal with the EU (or any other party) without the consent of the others. So a text initialled by just three of the five is unenforceable in law.

How the Commission responds will provide a crucial lesson for other ACP states. If it intends to offer duty-free quota-free (DFQF) treatment to the three signatories from January it will have to present a proposal to the Article 133 committee (as it has already done for EAC – see below). If it does so, knowing that the text has no legal validity in SACU, it is a clear signal that it is looking for face-saving texts. If not …

The EAC text
The latest text that we have seen is one dated 16 November which may well have been changed. For what it is worth, that text included no details of what would be liberalised. The evidence of substantial and rapid liberalisation comes from EU press statements and from a submission (dated 28 November) by the Commission to the Article 133 Committee (in support of the inclusion of EAC states in the list of countries to which it is empowered to grant DFQF treatment).

The latter states that EAC market access commitments entail liberalisation of 82% of the value of its imports (62% in 2 years, 80% in 15 years, the rest in 25 years) representing 74% of their tariff universe. The items excluded are: agricultural products (animal products as well as fruit and vegetables), processed foods, wines and spirits, chemicals, plastic, wood-based products, paper, textiles and clothing, footwear and glassware.

How much will be liberalised?
The first point to note is that these percentages mark an uncanny resemblance to the figures in a table with the name of an EAC government written on it that has been in circulation for some weeks. This table is a simple, single sheet – it includes no details of which products fall into which categories, just some figures of shares of trade to be liberalised and time periods.
The second point is to get the percentages right. EAC is said to be liberalising 82% of its imports over 25 years and is therefore excluding 18%. The figures in parentheses (italicised for ease of reference) must reflect the shares of the portion of imports that will be liberalised. Hence what is to be liberalised ‘in two years’ is 62% of 82% i.e. 51% of total imports will liberalised within this time period. A further 18% (of 82%) will be liberalised ‘in 15 years’ too, i.e. 15% of total imports from the EU. And the final 16% of total imports will be liberalised ‘in 25 years’ to bring the total to 82%.

There is no indication as to whether the second and third tranches will be spread over the years 2–15 and 16–25, or concentrated right at the end. Nor is it made clear whether these shares apply to each EAC state separately or to the group as a whole.

This is important because, at present, some 51% of total EAC imports from the EU are already duty-free (see Table 1 below). If the figures apply to the group as a whole (which is how they are presented by the EU), it will not be necessary for the region to remove any tariffs during the next two years.

Table 1: Share of EAC imports from EU duty free 2006


Imports (€000)

Duty-free (€000)


























What about the 18% of imports that will be excluded? Table 2 (below) shows that EAC imports from the EU of the product groups listed as being excluded account for almost exactly this share. Since it would be most unusual for a country to exclude in their entirety such large product groups (and not to exclude anything at all in any other chapter of the trade nomenclature), the following is what may have happened.

One interpretation
The EAC have identified their main areas of sensitivity and totted it up to 18% of the total. They have worked out how many goods already enter duty-free – 51%. So they have ‘agreed’ a deal that ensures: (a) that nothing need be liberalised for at least two and possibly up to 15 years and (b) all sensitive chapters (at least) can be excluded in perpetuity.

If a text is circulated that specifies product x is liberalised on date y, this analysis will be revealed as being too cynical by half. But until then, other ACP states shouldn’t believe everything they read in the newspapers!

Table 2: Share of EAC exclusions in imports from EU

HS chapters


Total trade1

HS 1-97


Excluded product groups:

Agricultural products

HS 1-24 excl. HS 3 (fish)



HS 28 & 29



HS 39


Wood-based products (incl. wood)

HS 44 & 47



HS 48


T & C

HS 50-63



HS 64



HS 70


Excluded groups’ total


Excluded groups’ share of total trade value



  1. This is total of straightforward, identifiable HS4 codes in HS 1-97. The total of all codes (including those covering confidential trade, corrections, various ‘goods not elsewhere classified’, etc.) is €2,292 million – so excluded-group share of that would be 16.9%.

Source: Eurostat COMEXT database (data on EU27 exports to EAC EPA countries), downloaded 30.11.07.