1. The European Commissioner for trade relations and the Head of the US Trade Department have just made reduction proposals of their trade-distorting agricultural domestic supports apparently impressive in order to unlock the Doha Round negotiations and to prompt developing countries to make parallel offers on the opening of their markets to the EU and US exports of services and non agricultural products.
2. Peter Mandelson has proposed to reduce the EU total AMS (“Aggregate Measurement of Support”) by 70%, the “de minimis” supports by at least 65%, and has told that the EU could even contemplate to reduce its blue box subsidies below what is prescribed by the Framework Agreement of 31 July 2004, that is below 5% of the value of EU agricultural production. We will show that these proposals would not challenge the common agricultural policy (CAP) at least as it is viewed from Brussels and they would even leave space to increase subsidies. On the other hand, since the EU has cheated massively in the notification of its agricultural supports in the various WTO boxes, the condemnation of these cheatings would make the CAP collapse. Another analysis will deal with the US offers on the reduction of its domestic trade distorting agricultural supports.
Viewed from Brussels, Peter Mandelson’s proposals are compatible with the CAP
Preliminary definitions of the components of domestic trade distorting supports
3. The total AMS corresponds to the trade distorting (or ’coupled’, implied to the price or production level of the current year), in other words to the ’amber box’. The total AMS is the sum of all product-specific AMSs and of the non product-specific AMS. Besides we have to distinguish between the authorized or bound total AMS and the applied or notified total AMS: the first has remained fixed at €67.2 billion since 2000 for the EU-15 whereas the second has been lowered to €39.3 billion in 2001-02, the last marketing year for which the EU has notified its agricultural domestic supports to the WTO. The Agreement on agriculture (AoA) has prescribed to developed countries to reduce their authorised total AMS by 20% from 1995 to 2000. As the total AMS of the base period 1986-88 in relation to which the reductions of the bound total AMS was due was of €83.949 billion (after conversion of the 80.975 billion ecus in euros), which explains that the authorized total AMS has fallen since 2000 at €67.159 billion.
4. One first remark concerns the €27.9 billion gap between the authorized and applied total AMS in 2001-02. It is actually a legacy of the gap between these two indicators which was already of 31.146 billion ecus (respectively 78.672 and 47.526 billion ecus) in the first year of implementation of the AoA in 1995-96. Therefore, from the beginning, the applied total AMS had already been reduced by 39.6%, almost twice the reduction (20%) required for the whole implementation period 1995-2000!
5. This first EU’s swindle (the US one is even larger) can be explained mainly by the fact that, during the base period 1986-88, the farmers’ incomes were essentially based on market prices supports, among which high intervention prices, whereas the 1992 CAP reform has greatly reduced the intervention prices of cereals and bovine and ovine meats, replacing them by 18.8 billion in direct payments put in the blue box, then taken out of the total AMS. Besides, the total applied AMS was particularly high in 1986-88 for two reasons: 1) world agricultural prices were very low in that period so that the product-specific AMSs were particularly high; 2) The EU and USA have benefited from a “credit for AMS reduction” having consisted in an increase of the base period total AMS to reward countries which did not wait the end of the Uruguay Round to reduce their coupled domestic supports.
6. The de minimis supports are the same coupled supports but which are not included in the product-specific AMSs as long as they remain lower than 5% of the production value of each product, nor in the non product-specific AMS as long as they remain lower than 5% of the value of the total agricultural production of the WTO Member. This value being €246.4 billion for the EU- 15 in 2001-02, the authorized non product-specific de minimis support therefore was €12.3 billion.
7. Finally the blue box is composed of partially decoupled subsidies - because based on fixed production factors (acreage of the 1989-91 period, yields of the 1986-91 period, and cattle heads of 1992) - granted to farmers to compensate the reduction of guaranteed prices (’intervention prices’). The EU blue box (the USA do not have one) is made of the direct payments created since the CAP reform of 1992 for cereals, oilseeds, pulses, bovine and ovine meats, the amount of which has been increased by hectare or cattle head by the CAP reform of 1999. The CAP reform of 2003-04 has introduce new direct payments to compensate the reduction of skimmed milk powder, butter, rice, cotton, olive oil, tobacco and hemps. Sugar would follow within few months.
8. Above all this last CAP reform has decided to transfer (from January 2005 for most EU Member States and from January 2006 for France) the bulk of direct payments from the blue box to the green box - which includes supports that are not subjected to reductions -, by ’decoupling’ them totally, i.e. by granting them without any obligation to produce or with the possibility to produce products which were not previously subsidized: it is the ’single farm payment’ (SFP) generating ’single payment rights’ (SPR).
The reduction of the authorized total AMS
9. In other words, from 1992 on the EU has progressively operated a double transfer of its agricultural domestic supports from the amber box to the blue box and then to the green box. So much so that, when the last CAP reforms started in 2003 will have been implemented at about the same time when the new Agreement on agriculture, which would result form the Doha Round, will itself been implemented by the 1 st January 2008, the applied total AMS would have been reduced from €39.3 billion in 2001-02 to about €18.8 billion, a €20.5 billion reduction imputable to product-specific AMSs. These reductions would be approximately the following: €9.709 billion for bovine meat (following the elimination of its intervention price the 1 st July 2002), €378 million for rice, €1.898 million for dairy products, €1.242 billion for olive oil, €517 million for cotton, €964 million for tobacco and €5.809 billion programmed for sugar.
10. All this would reduce by €48.4 billion the applied total AMS relatively to the authorised total AMS of €67.2 billion, i.e. a reduction of 72%. We now understand why Peter Mandelson has proposed to reduce the EU authorised total AMS by 70% , without having to reduce the foreseeable applied total AMS at all which could even increase by some €1.6 billion!
11. Indeed the EU could even lower its applied total AMS to €10.4 billion by replacing the fruit and vegetable AMS, which has been of €8.4 billion in 2001-02, by direct payments included in the ’single farm payment’ of the green box, in which case the possible reduction of the authorized total AMS would reach 84.5%!
The product-specific AMSs linked to administered prices are economically meaningless but allow the EU to look like reducing greatly its coupled domestic supports
12. At this stage, a more technical explanation is needed on the way to calculate the product-specific AMSs. In the EU, about 95% of them are market price supports linked to an intervention price, the remaining 6% being subsidies linked to the production or price levels .
13. The writers of the Agreement on agriculture did not have a high intelligence quotient because the AMS linked to an intervention price is economically meaningless since it is defined and worked out as the gap between the intervention price of the current notified year and the world reference price of the base period 1986-88 , gap multiplied by the production volume which could possibly benefit from it, in practice the whole production. Therefore this AMS is independent from the actual intervention operations and from the current world price. For example, as the intervention price of wheat is €101.31 per tonne (€/t) since July 2001 and as the 1986-88 world price was 86.5 €/t, the per tonne AMS has remained at 14.8 €/t ever since although the average world price of wheat has increased from 103 $/t in 2000/01 to 115 $/t in 2001/02, 142 $/t in 2002/03 and 153 $/t in 2003-04. However the actual support by tonne, measured by the gap between the intervention price and the world price of the current year and concretised by the public stocks level, has decreased greatly!
14. A second evidence of the low intelligence quotient of the writers of the Agreement on agriculture has been to consider the intervention price as a measure of support of the domestic market price sufficient by itself. Actually it would not have any impact on the domestic price without coexisting with much more determinant market prices supports : import protection first but also export subsidies, production quotas, set aside and domestic and foreign food aid. Therefore we are dealing with a false market price support, the elimination of which has no significant consequence on the income level of farmers or on the agricultural price level, as long as are present a high import protection and the other above measures. The more so as the elimination of the intervention prices is generally replaced by possible subsidies to private storage which, being not triggered automatically, do not lead to the calculus of an AMS, which allows to lower the applied AMS.
15. This absurdity of the AMS linked to intervention prices has thus led the EU to eliminate the intervention price of bovine meat the 1 st July 2002, a true sleight of hands which has allowed to reduce from one day to the other its applied total AMS by 24.7%, i.e. by €9.7 billion. And this without any negative impact on the market price level or the ranchers’ income since the CAP reform of 1999 had planned a large increase in direct payments to compensate the elimination of the intervention price . The EU is about to renew this sleight of hands with the proposed elimination of the intervention price of sugar, which will reduce its applied total AMS by a new €5.8 billion.
16. Besides, the way the AMS linked to an intervention price (or to an equivalent measurement of support) is computed corresponds to a theoretical market price support, not to an actual price support, the more so as the world reference prices of 1986-88 were very low. And this AMS is much larger than the cost of the compensating subsidies, so that the WTO Members such as the EU have a large interest in replacing the product-specific AMSs linked to administered prices by subsidies since this lowers its total support, the more so when they can be put in the blue or green boxes. For instance the elimination of the intervention price of sugar will replace the €5.8 billion AMS by a direct payment to sugar beet growers of €1.5 billion.
However the trick does not stop here: it extends to the de minimis support and the blue box.
The de minimis supports: increasing the supports while giving the impression they will be reduced
17. As long as the intervention price of bovine meat existed, the bovine meat AMS represented a proportion of the production value of bovine meat much above the 5% de minimis level: its €9.709 billion AMS in 2001-02 represented 47% of the €20.671 billion value of bovine meat production. Because there is no longer an intervention price since July 2002, then no bovine meat AMS, the EU avails of an authorizedde minimis support for bovine meat of 5% of its production value, i.e. of €1.033 billion. The same will occur for sugar when its intervention price will be eliminated if the reform presented by the European Commission to the Council and the European Parliament is adopted.
18. Therefore as soon as the EU eliminates intervention prices or lowers them enough so that the applied product-specific AMSs become lower that 5% of the production value of the corresponding products, the value of the agricultural production of products without AMS increases and therefore the overall authorized product specific de minimis. So that the elimination of the intervention price, which is presented as a reduction of a trade distorting support, is not only converting the false market price support in coin of the realm of the blue or green boxes but allows also to reintroduce a significant amount of authorized trade distorting supports.
19. As Australian researchers have underlined it, the calculus of the authorized de minimis of the product-specific AMSs has to avoid the double counting which would occur if we took into account 5% of the value of the whole agricultural production, because we should not take into account the production value of the products for which a de minimis exemption has not and could not be notified since their product-specific AMS is higher than the 5% exemption level . It is possible and even likely that the European Commission did not enter into such subtle technicalities and is considering that the de minimis support of the product-specific AMS is worth 5% of the total agricultural production value.
20. In fact the EU has not used the de minimis supports much, both for the product specific AMS and the non product specific AMS, contrary to the USA which has a large applied non product specific AMS. In 2001-02 the de minimis support linked to product specific AMS has been of €468 million whereas the total production value of the products with an AMS higher than the 5% de minimis level has been of €113.1 billion, on a total agricultural production value of €246.4 billion. Which means that the agricultural production value of the products without product specific AMS has been of €133.3 billion, and therefore that the authorized product-specific de minimis was of 5% of that value, i.e. of €6.7 billion. And since the non product specific AMS has been of €574 million, the increase in the authorized non product specific de minimis support was €11.7 billion in relation to the total authorized €12.3 billion.
21. Finally the authorized total de minimis support - product-specific and non product-specific - was of €19.0 billion (6.7 + 12.3) whereas the applied (notified) total de minimis support was only of €1.042 billion. Therefore we understand quite easily why Peter Mandelson has declared that the EU could reduce at least 65% of its de minimis supports since it could even reduce them by 94.5%! Reducing them by 65%, i.e. by €12.4 billion, would let the EU increase them to €7.5 billion, an increase margin of €6.6 billion in relation to the applied level of 2001-02.
Capping the blue box at 5% of the agricultural production value is not restrictive
22. The EU has notified €23.726 billion in blue box subsidies for 2001-02, representing 9.6% of the agricultural production value. Since the Framework agreement of 31 July 2004 has capped the blue box at 5% of the agricultural production value, i.e. at €12.321 billion, this would imply a reduction of €11.4 billion. In practice this would not be necessary since the CAP reforms of 2003-04 are transferring the bulk of blue box subsidies in the green box of the ’single farm payment’ (SFP), so that the residual blue box would presumably not go beyond €7 billion the 1 st January 2008, when the new Agreement on agriculture would begin to be implemented. Which would leave an increase margin of €5.3 billion. We understand once more why Peter Mandelson has declared that the EU could agree to lower the blue box below the 5% ceiling agreed to in the Framework agreement.
Provisory conclusion: the claim of huge cuts in agricultural supports turns into an increase of €13.3 billion
23. A partial conclusion is that Peter Mandelson’s offers are actually compatible with the CAP reforms of 2003-04: by the magic box-shifting and the absurd existence of the product specific AMS linked to administered prices, not only the EU would not be obliged to reduce its applied total domestic trade distorting agricultural supports but would be able to increase them by €13.4 billion (1.5 for the total AMS + 6.6 for de minimis supports + 5.3 for the blue box). And this after having reduced by 70% its authorized total AMS, by 65% its authorized de minimis supports and after having reduced its applied blue box to 5% of the agricultural production value, although its applied blue box subsidies represented still 9.6% of that value in 2001-02!
24. When you think about it, what we have qualified of low intelligence quotient of the writers of the Agreement on agriculture is rather reflecting the EU and US treachery which, having negotiated practically in a face to face its rules, had realized all the advantage they could draw from this apparently absurd concept of product-specific AMS linked to an administered price, without speaking of the de minimis concept and of the blue and green boxes. But there’s always someone cleverer than you and those treacheries will turn against their authors which have cheated massively.
However all these calculations presume that we ignore the EU’s massive cheatings
25. The self confidence of the European Commission that its offers of reduction of its agricultural domestic supports do not put the present CAP into question is however extremely fragile since the EU has cheated at three levels: 1) it does not abide by an important provision of the Framework Agreement of 31 July 2004; 2) the massive cheatings in its past notifications to the WTO; 3) the present CAP does not abide by the AoA rules.
The EU does not take into account the ceiling in the product-specific AMSs foreseen by the Framework Agreement
26. Paragraph 9 of the agricultural annex of the Framework Agreement of the 31 July 2004 states: “To prevent circumvention of the objective of the Agreement through transfers of unchanged domestic support between different support categories, product-specific AMSs will be cappedat their respective average levels according to a methodology to be agreed”.
27. Since, according to article 6 of the Agreement on Agriculture, “The commitments are expressed in terms of Total Aggregate Measurement of Support and”Annual and Final Bound Commitment Levels“”, there is no bound ceiling for each product-specific AMS and the “average levels” referred to in article 9 of the Framework Agreement above can only refer to the applied, i.e. notified, product-specific AMSs. Besides two interpretations are possible depending on the fact that the cap is applied to each product-specific AMS or to all product-specific AMSs taken together. It seems that the WTO Members did not discuss yet of this issue of article 9 in the Committee on agriculture, the more so of the “methodology to be agreed”. This is quite understandable for developed countries as it was not in their interest but this is less understandable from the part of developing countries: presumably they did not realize the importance of the issue.
28. Confining ourselves to the less restrictive interpretation for the EU and US that it is the applied product-specific AMSs taken as a whole which should be capped, this would imply the same ceiling as the €39.3 billion of the applied total AMS for 2001-02 since the non-product-specific AMS was nil, being much below the de minimis cap. In other words €39.3 billion will become the 1 st January 2008 the cap of the authorized total AMS and the reduction of the applied total AMS at €18.8 billion by this time would only allow to concede a reduction of 52.2%, much lower than the 70% proposed by Peter Mandelson.
29. Above all, the margin of reduction of the EU authorized total AMS will reduce to almost zero once reintroduced in the amber box the EU massive cheatings in the notifications, or lack of notification at all, of its domestic subsidies.
The EU massive cheatings in the notification of its coupled domestic supports
30. Since the implementation of the Agreement on agriculture in 1995 the EU and USA have cheated massively by notifying in the blue and green boxes, or by not notifying at all, the bulk of their subsidies which should have been put in the amber box and subjected to reductions, or which should have been notified as export subsidies. These massive cheatings can be explained partially by the existence of the ’peace clause’ which protected largely domestic supports from prosecutions at the WTO. But the main reason is that the WTO does not check the veracity of its Members’ notifications: it falls on the Member convinced of the cheating of another Member to prosecute it at the WTO. Even then the Member must have access at the information data of the cheating Members, which is very difficult for developing countries, and above all that they have the political will to prosecute the giant EU and and US players, which control also indirectly the IMF and World Bank taps. And, since both the EU and US are cheating and know it, no one moves.
31. Having asked the WTO on this cheating issue, Gabrielle Marceau, of the Dispute Settlement Body, replied the 27 February 2001 in an internet forum: “The WTO has neither the resources nor the skills to act like”a regulator“of these notifications. It is up to each Member to do these verifications... This is the very spirit of the whole disputes settlement system of the WTO: every Member country acts as a guard-dog of the system”. This WTO’s casualness about its Members’ notifications is all the more unexplainable that its second fundamental mission, after the dispute settlement, is the trade policy review of its Members (every other year for the EU and USA). Although this review is mobilizing large resources, it is only based on the information that each Member is willing to communicate to the WTO so that their conclusions are always laudatory for the Members, and the more so for the most powerful of them.
32. The largest cheating is related to feedstuffs. The Agreement on agriculture states clearly (article 6.2) that input subsidies are coupled and subject to reductions for developed countries, before dealing with the blue box subsidies (article 6.5). As about 60% of the EU and US production of cereals, oilseeds and pulses (COP) are fed to animals, are inputs for animal products, 60% of direct payments going to COP are clearly coupled and subject to reductions. But the EU has notified all its direct payments to COP in the blue box and the USA has notified them in the green box. So that the EU has cheated on an amount of about €63 billion in its notifications on COP subsidies from 1995-96 to 2001-02 (of which about €9 billion for 2001-02), and these amounts should have been notified in the amber box, i.e. in the COP specific AMSs. The CAP reform of June 2003, which is supposed to have transferred about 80% of direct payments from the blue box to the green box of the ’single farm payment’ (SFP), will not change their statute of input subsidies to be put in the specific AMS of the amber box as long as the farmers getting the SFP will go on growing COP eventually fed to EU animals.
33. The EU has then cheated in three ways:
1) It has “forgotten” to notify, even in the green box, some subsidies to be put in the non product-specific AMS: irrigation subsidies for at least €300 million, and tax rebates on agricultural fuel for at least €2 billion (€980 million in France alone).
2) It has under notified in the non product-specific AMS interest subsidies on agricultural loans for at least €700 million and subsidies on agricultural insurances for at least €100 million.
3) Above all it has cheated by putting in the green box subsidies which should have been in the non product-specific AMS: this is the case first of subsidies to investments of farmers and agri-food industries for an average of €5.6 billion since 1995-96, of which €5.4 billion in 2001-02.
34. Therefore, with only these three items, and adding the €574 million notified in the non product-specific AMS in 2001-02, this one reaches €9.1 billion and approaches the authorized de minimis ceiling of €12.3 billion, leaving only a possible increase of € 3.2 billion for the applied non product specific de minimis. However the right notification in the non product-specific AMS of practically all green box subsidies (other than the direct payments that the EU intends to transfer from the blue box to the green box) will make explode this AMS much beyond the de minimis ceiling so that this non product-specific de minimis will disappear.
35. As for the authorized de minimis support of the product-specific AMSs, it is much lower than the €6.7 billion computed above (paragraph 20) since the production value of agricultural products without a product-specific AMS falls from €133.3 billion to €53.0 billion. Indeed taking into account the feedstuffs subsidies gives now a product-specific AMS to the production of pigmeat (production value of €25.625 billion), poultry and eggs (€17.277 billion) and milk (€40.134 billion). The actual authorizedde minimis support of the product-specific AMSs is therefore of €2.7 billion only, and it is the same amount for the authorized total de minimis support since the non-product specific de minimis support has vanished. Therefore reducing it by 65% as Peter Mandelson has proposed would take it back to a mere €945 million. There is nothing to brag about.
36. Without speaking of cheatings properly, but rather of the inconsistency of the criteria of Annex 2 of the Agreement on agriculture, which has been exploited fully by the EU and USA, we have shown that all blue box subsidies and practically all green box subsidies (the domestic food aid excepted) are coupled (see paragraph 42 below for the SFP) are coupled, which, for 2001-02, corresponds to €44.161 billion, of which €23.726 billion for the blue box and €20.435 billion for the green box.
37. All in all, adding to the €44.2 billion the €3.1 billion of subsidies not notified at all or under notified, then the €18.8 billion of the likely total applied AMS on the 1 st January 2008 (paragraph 9 above), the actual applied total AMS would be of €66.1 billion and the possibility to reduce it in regard of the authorized total AMS of €67.2 billion would shrink to a minuscule 1.6% not to say zero given the approximation of these calculations. We are very far from the claim of a 70% reduction by Peter Mandelson. The only tiny margin remaining would be to eliminate the last product-specific AMSs, such as on fruit and vegetables (€9,7 billion in 2001-02), replacing them by lower direct payments, which should themselves remain in the product-specific AMSs however.
38. Finally the notification in the product-specific AMSs of all the €23.726 billion subsidies notified improperly in the blue box (the majority of which the European Commission thinks have been transferred in the green box since 2005) will make the blue box vanish totally so that there is clearly no possible reduction.
39. Although this paper is focusing on domestic supports, let us underline that the EU and USA are also cheating massively on their export subsidies since they forgot to include in them the domestic subsidies benefiting also to exported products, which are to-day much larger than the proper export subsidies. These have been reduced for cereals from 2.16 billion ecus in 1992 to €121 million in 2002. But, taking into account the direct payments going also to exported cereals, which have increased from 117 million ecus in 1992 to €1.28 billion in 2002, and cereals exports having dropped from 36.4 to 18.4 million tonnes, the subsidy per exported tonne has increased by 20% (from 62.5 ecus to €75.1).
40. Similarly, the EU and USA forget to count as export subsidies the direct payments on EU feedstuffs (cereals, oilseeds and pulses) consumed by the animals whose products are exported (meats, eggs and dairy products). As the issue in this paper is domestic supports, we will not go farther here.
41. Happily enough three judgments of the WTO Appellate Body - “Dairy products of Canada” of the 3 December 2001, “US cotton exports” of the 3 March 2005 and “EU sugar exports” of the 9 April 2005 - have created precedents underlining that US green box subsidies are coupled and have dumping effects when the products are exported. They have also shown that one can no longer consider there is no dumping as long as exports are made at the domestic price but the export price cannot be lower than the average full production cost of the country, taking into account crossed subsidies. Consequently all subsidies of the blue and green boxes going to exported products can from now on be prosecuted for dumping.
The single farm payment is not in the green box
42. The ’single farm payment’ which concentrates presently the bulk of EU direct payments is coupled since it does not comply with three of the five conditions to be fully decoupled:
1) It is based on the amount of direct payments received from 2000 to 2002, a criterion not provided for by paragraph 6 of the Annex 2 of the AoA.
2) The farmer cannot produce what he wants since many productions are either forbidden (fruits and vegetables, and milk and sugarbeet if he has no production quota) or capped (cotton, tobacco, olive oil and not beyond the milk or sugarbeet quotas). Now the only interdiction to grow fruits and vegetables has been enough for stating that the US direct payments were not decoupled.
3) The farmer must show each year that he has eligible hectares to receive the SFP.
43. Above all, as the SFP cannot be ascribed to a particular production, it is ascribable to all productions of which it contributes to reduce the production cost, then the price below the full production cost without this subsidy. Therefore all EU agri-food exports can be attacked at the WTO on dumping grounds, even those which did not received any export subsidy such as quality cheese or wines as long as their producers are receiving a SFP.
Conclusion: all subsidies should be allowed, provided they do not benefit exported products
44. To conclude, the question is not to condemn all agricultural subsidies by themselves. They are legitimate as long as a country does not harm other countries through exported subsidized products - taking into account upstream subsidies on inputs and investments -, i.e. at prices below the full average production cost of the country.
45. In the Doha Round negotiations the rules of the game are unfair: only rich countries can support their farmers through subsidies compensating prices which have been reduced below the production cost, generating a dumping effect and an import substitution effect. At the same time poor countries are forced to reduce their only available instrument: import protection. And if the WTO is not compelling the least developed countries to reduce their tariffs, they have been compelled to do so by the IMF and the World Bank. Therefore food sovereignty, i.e. an efficient import protection together with the ban of any export below the average full production cost, is paradoxically the least protectionist way of supporting farmers all over the world.
46. Besides, rebuilding the CAP and the AoA on food sovereignty without a dumping camouflaged below green box subsidies is clearly in the EU selfish interest since its agri-food exports as a percentage of its total production have only represented, from 2000 to 2003: 10.7% for cereals, 6.9% for all meats and 9.5% for dairy products. In the present game where EU agri-food products are a bargaining chip in its negotiations at the WTO and with Mercosur to open new markets for its exports of services and industrial products, the EU would lose much more than 11 million of agricultural employments, given the specific multifunctionality of agriculture.
Notes:
· Economist, author of “L’agriculture, talon d’Achille de la mondialisation, L’Harmattan 2001” ( berthelot ensat.fr).
FAO, Soutien interne: aspects liés aux échanges et indications empiriques, n°5 des Documents techniques de la FAO sur les politiques commerciales relatives aux négociations de l’OMC sur l’agriculture, 2005.
Ivan Roberts, WTO Agreement on agriculture. The blue box in the July framework agreement, ABARE, March 2005.
See the book “50 fiches pour refonder les échanges agricoles sur la souveraineté alimentaire” by Jacques Berthelot which will be published soon. An English version will follow.