The dispute

In September 2004, a WTO Panel requested by Brazil, Australia, and Thailand concerning EU export subsidies for sugar found against the EU, a result upheld by the WTO Appellate Body in April 2005. In what follows, I will briefly discuss the main issues involved in this dispute and in the WTO ruling, before turning to some comments on the Hoekman–Howse paper. I will focus my attention solely on the economics part of the paper, since this is what was available when I was asked to write my comments.

The dispute concerned two main issues in the EU sugar regime: C sugar and ACP/India re-exports. Regarding the first issue, the EC sugar regime establishes production quotas for two categories of sugar, labeled ‘A sugar’ and ‘B sugar’.Footnote ¹ These are the maximum amounts of sugar that may be sold within the EC in a given year. Producers are required to export any surplus amounts, designated ‘C sugar’. Domestic prices for A and B sugar are supported by an array of government measures and also receive direct export subsidies. EC sugar producers receive no additional funds from the EC if they export a large amount or no C sugar. The complainants argued that C sugar effectively benefited from a cross-subsidization from A and B quota sugar.

Regarding the issue of ACP/India re-exports, under the Sugar Protocol of the Cotonou Agreement, the EU grants preferential access to 1.3 million tonnes of sugar from African, Caribbean, and Pacific (ACP) countries. EU producers obtain export subsidies to re-export amounts of sugar equivalent to the imports of ACP-origin sugar. The EC has a similar sugar arrangement with India. The complainants argued that the EU should have included ACP/India re-exports in its calculation of total subsidized exports.

The WTO ruling

The Panel and AB reports found that the EU subsidies on sugar exports were beyond the level formally notified to the WTO – the so-called ‘commitment schedule’ – and were thus in violation of the WTO Agreement on Agriculture. The Panel finding was that both C sugar exports and ACP/India re-exports should be counted against the EU's WTO export-subsidy limits.

On the issue of C sugar, the Panel found that C-sugar exports benefited from a cross-subsidization from A and B quota sugar, effectively receiving an export subsidy. The Panel concluded that ‘A, B, or C sugar are part of the same line of production and thus to the extent that the fixed costs of A, B, or C are largely paid for by the profits made on sales of A and B sugar, the EC sugar regime provides the advantage which allows EC sugar producers to produce and export C sugar at below total cost of production.’

On the ACP/India re-exports, the WTO Panel and AB reports found that the 1.6 MMT of sugar re-exports should be counted against EU's export-subsidy commitments. The EU argued that a footnote in its commitment schedule excluded this type of sugar from the scope of its subsidies-reduction requirements. The Panel dismissed this argument, holding that the footnote had no legal effect and could not enlarge or modify the EU's specified commitment levels.

Comments

The most controversial issue in this dispute, the one on which both the Hoekman–Howse paper and the other papers in the literature have mostly focused, involves the cross-subsidization of C sugar. This is also what I have decided to concentrate my comments on.

The general conclusion from the WTO Panel and AB reports is that below-cost exports of an agricultural product may, even in the absence of ‘direct’ export subsidies, represent proof of export subsidization if there is close linkage between these exports and domestic support programs.

To evaluate whether the argument of cross-subsidization is a valid one, we have to first be able to answer this question: ‘What drives production of C sugar?’ Hoekman discusses three possible answers: economies of scale, insurance motives, and the efficiency of some EU sugar producers.

The first answer is the one put forward by the WTO reports: given fixed costs, rents derived from quota production are used to cover below-cost C sales and enhance profits. As stressed by Gohin and Bureau (2006), this argument is not fully compelling. This is because ‘in the beet sector, there are many opportunities to share machinery, to buy second hand equipment and to purchase contract work. Contract harvesting or planting costs are only slightly decreasing with the size of operation. The fixed component in the cost of contract work is not large enough to provide a significant incentive for producing C beet so as to spread this fixed cost on a larger output’ (p. 230).

A second possible reason for the production of C sugar, which is considered by Hoekman, is that sugar producers may rationally overshoot production due to ex ante uncertainty about future harvests. The existence of rents associated with A and B quota sugar may strengthen the insurance motive, since producers may overshoot to make sure that they can capture the quota rents in the case of poor harvests. As stressed by Hoekman, one clear limitation with this argument is the fact that the EU sugar regime contains carry-forward provisions that allow for quota rights to be transferred to the next year, thus attenuating the incentive to overproduce.

The third possible reason for the production of non-subsidized sugar mentioned in the Hoekman–Howse paper is that some EU producers may actually be efficient and able to produce at world market prices. Indeed, some studies have argued that countries such as France, Germany, Austria, and the UK have a comparative advantage in sugar production and would be net sugar exporters even in the absence of any policy intervention (see, for example, the paper by Fransden, Jensen, Yu, and Walter-Jorgenson, Reference Fransden, Hans, Wusheng and Aage2003).

In his analysis of the possible reasons for the production of C sugar, Hoekman–Howse disregard a possible fourth explanation that has been put forward by the previous literature, i.e. expectations about future reform. The idea is that farmers may overproduce because they expect that historical reference levels would be used in future reforms. Indeed, in past reforms of various EU farm policies, quota allocations, premium rights, or compensation payments have been given on the basis of historical references. Under some expectations regarding future reforms, precautionary behavior involving the building up of potential reference levels would be rational (see Gohin and Bureau, Reference Gohin and Jean-Christophe2006).

I find that the Hoekman–Howse paper does an extremely good job at explaining the complexities of the EU policy regime for sugar and the issues at stake in this dispute. My only criticism, if I have one, is that the paper does not go far enough in trying to answer the question of whether economies of scale are indeed the main reason behind the production of C sugar. Without a clear answer to this question, we are left with some serious doubts about the cross-subsidization argument put forward in the WTO Panel and AB reports.

A more critical assessment of the general reasoning of the panel in this dispute – that policies creating rents on a portion of a firm's output may allow for the cross-subsidization of exports – is of great importance, since this reasoning may lead to an increase in the number of future disputes. Other agricultural programs (e.g., US rice, corn, and soybeans programs) are particularly vulnerable to future dispute challenges, since Article 10.3 of the Agreement on Agriculture establishes the so-called ‘reverse burden of proof’. According to this doctrine, a complainant must prove only that another country has exceeded its export commitments; the respondent must then prove that it did not grant an export subsidy on the excess exports.