Unofficial guide to the 6 December 2008 ‘revised draft modalities’ CORRECTED: 9 DECEMBER 2008

The main purpose of this note is to walk you through the revised draft text circulated by Ambassador Crawford Falconer, chairperson of the agriculture negotiations, on 6 December 2008.

This time: four texts. This note summarizes the main points of the texts and indicates where changes have been made compared with the previous draft circulated in July 2008.

Because this note is simplified you should consult the original for a more complete and precise picture.

> Revised draft modalites for agriculture (6 December 2008)
> Chairperson Crawford Falconer’s statements and other news

> Original mandate: Article 20
> The Doha mandate
> The Doha mandate explained


See also:
> Negotiations gateway
> 2004 agreed framework
> 2005 Hong Kong Ministerial Declaration
> More on the modalities phase

Need help on downloading?
> Find help here

For starters

  • No surprises. This is the draft’s fourth revision this year. These drafts are painstakingly built up from ideas discussed in the talks. They reflect the latest thinking among negotiators and the chairperson, drawing on members’ evolving positions (ie, a “bottom up” process), including roughly 240 hours of talks between September 2007 and July 2008, lengthy separate consultations among delegations, negotiations among a group of ministers in July 2008, and further consultations since September 2008. Unsurprisingly, there are no surprise.

  • What’s new this time? The objective of this draft is slightly different from previous versions. With the possibility that ministers could come to Geneva soon to try to reach agreement, Amb.Falconer says he is also looking at what might be in a final package. “... The changes made at this time now represent a best estimate of where there is additional good reason to believe there would prove to be consensus if everything was to come together as a modalities package”. But some issues remain difficult and the chairperson has issued three additional papers containing his suggestions for how to deal with these. The changes in the main paper include: single numbers instead of ranges for cutting tariffs (including in-quota tariffs), for selecting sensitive and special products, and for domestic support; revised provisions on the present special safeguard (SSG) and on tariff simplification; and further fine-tuning of flexibilities for different groups of countries. The three additional texts deal with selecting sensitive products (including the cases of Canada and Japan), creating new tariff quotas (which affects what products a country can select as sensitive), and the special safeguard mechanism (SSM) for developing countries.

  • A lot of progress. Arriving at consensus does not always require dramatic breakthroughs. A number of issues, including some of the big picture numbers, have gradually been “stabilized” over time. The objective has been to whittle down the outstanding issues to a manageable few and to a large extent this has been achieved. The remaining questions can then be discussed politically, and in comparison with other subjects, particularly non-agricultural market access (NAMA).
    In that sense, a tremendous amount of progress and significant compromises have been made since September 2007. This has included sorting out a number of flexibilities targeted at specific situations — for over one-third of WTO members, including around 45 small and vulnerable economies, and different groups of countries that recently joined the WTO (the “recently-acceded members” or RAMs). As it turns out, the role of the formulas has changed somewhat.

  • There’s more to it than formulas. We should not exaggerate their importance. Sorting out other issues has taken some pressure off the big numbers because some of the biggest political difficulties are about flexibilities to deal with various sensitivities. The main formulas are more or less agreed.

    1. Some negotiators say that the major issue in market access for them is the selection and treatment of sensitive products, where there has been considerable progress but some problems remain. It’s highly technical — but with real commercial impact involving important traded products.

    2. Flexibilities are also being given to developing countries, small and vulnerable economies and recent new members, but many of these have also been the focus of attention not least because special treatment for one group of developing countries can have an impact on another group.

  • That said, the formulas are still important for countries and products where the numbers will apply, and because many flexibilities take the form of deviations from the formulas.


  • The negotiations aim to reform agricultural trade principally in three areas (the “three pillars”): domestic support, market access, and export subsidies and related issues (“export competition”).

  • The “modalities” would spell out how to achieve this, including steps to be taken each year over a period.

  • After the “modalities” have been agreed, each country would use them to cut subsidies, support, and tariff ceilings on thousands of products. These and new related rules would be part of the final deal.

  • Formulas in the “modalities” would describe the basic cuts in tariffs, support and subsidies. For domestic support and tariffs, “tiered” formulas are used: if support or a tariff is high (ie, in a higher tier) it will be cut more steeply. Export subsidies would be eliminated.

  • Not one-size-fits-all: the basic formulas for developing countries prescribe gentler cuts over a longer period. On top of that, a range of flexibilities would allow countries to deviate from the basic formulas, either totally or for some products, particularly in market access. This is designed to take account of countries’ different vulnerabilities, the liberalization already undertaken by new members, and a range of special circumstances for some products in different countries.
  • New or revised rules and disciplines would also be in the “modalities”: these are as important as the formulas and are part of the deal. They include reducing the potential that permitted domestic support could distort trade, ensuring the methods of administering quotas do not themselves impede trade, and disciplining export finance, exporting state trading enterprises and international food aid so that they do not provide loopholes for export subsidies.

What does this mean for …?

Market access: tariffs, tariff quotas and safeguards

What would this mean for wheat, rice, beef, sugar, milk, cheese, potatoes, pineapples, etc? How deep the cuts in their tariffs would be depends on:

  • how high the current tariff is: higher tariffs have higher cuts, ranging from 50% to 70% subject to a 54% minimum average for developed countries, 33.3% to 46.7% for developing or less if they meet a 36% average cut
  • whether the product is “sensitive” (all countries) or “special” (developing): sensitive products would have cuts of only 1/3, 1/2 or 2/3 of the normal cut but with a quantity allowed in at a lower quota; special products would also have smaller cuts, and some might be exempt completely
  • whether the applied tariffs are lower than the bound tariffs. Cuts are made from legally bound rates. Tariffs actually charged can be lower. If a developing country has a bound tariff of 100% but only charges 25%, the bound tariff would be cut by 42.7% ie, cut to 57.3%. That means no change in the 25% tariff actually charged, with room to more than double the tariff
  • the country’s status: least-developed countries would make no cuts on any products, developing countries in general would make smaller cuts and have more flexibilities than developed, small and vulnerable economies would make even smaller cuts with even more flexibilities, and countries that recently joined the WTO and some individual countries would also have special terms
  • … including if imports increase sharply or their prices fall a lot. Although the tariff will be cut, developing countries will be able to use a “special safeguard mechanism”, allowing them to increase the duty temporarily.

Support for farmers and for agriculture

Support that “distorts” markets would be cut but not eliminated. This is the type of support that depresses world prices and discourages production in poorer countries because it encourages farmers to produce more in the richer subsidizing countries than elsewhere. In times of plenty it has even created wasteful surpluses described as “wine lakes” or “beef mountains”. Examples of this type of support include price guarantees or support that is based on how much is produced. Countries providing large amounts of support would cut these the most; many are already reforming their programmes. They and the rest would still be allowed a small or “de minimis” amount limited to 2.5% of the value of production for developed countries, 6.7% for developing. The amount of support a country can give to individual products would also be limited.

But a wide range of support for agriculture as a whole would be allowed without limit under the “Green Box”, ie, for development, infrastructure, research, agricultural extension, structural adjustment, etc. Conditions would be tightened to prevent direct income supports, etc, from stimulating production.

Export subsidies

These would be eliminated by 2013, including subsidies hidden in credit, non-emergency food aid and the activities of exporting state trading enterprises.


What is this paper? This is NOT a “proposal” from the New Zealand ambassador (or from “the WTO”) in the sense that we would normally understand the word “proposal”. In other words, it is NOT his opinion of what would be “good” for world agricultural trade.

Rather, it is an assessment drawn from WTO member governments’ positions. It is the negotiations’ chairperson’s judgement of what they might be able to agree — based on what they have proposed and debated in over seven years of negotiations and their responses to his previous papers. He has stressed that this is not final. It puts the possible areas of agreement on paper so that members can react and further revise the draft. So this paper kicks off another intensive series of meetings and comment.



How are these issues being negotiated?

In this phase of the negotiations, the hard talking on agriculture has taken place in meetings of 36–37 representative delegations, a more manageable size than sessions of the full membership. The process is controlled by meetings of the full membership and is chaired by the talks’ chairperson, Ambassador Crawford Falconer of New Zealand. The 36–37 meet in Room E at the WTO and the sessions are sometimes called “Room E” meetings. All coalitions are represented to ensure the talks are inclusive and transparent.

In January 2008 there were 37 delegations in Room E:
Argentina (Cairns Group, G-20), Australia (Cairns Group coordinator), Benin (Cotton-4, African Group, least-developed, Africa-Caribbean-Pacific), Brazil (G-20 coordinator, also Cairns), Canada (Cairns), Chad (Cotton-4 coordinator, also African Group, least-developed, ACP), China (G-33, G-20, recent new member), Colombia (Cairns, tropical products group), Costa Rica (tropical products coordinator, also Cairns), Côte d’Ivoire (African Group coordinator, also ACP), Cuba (G-33, small and vulnerable economies), Dominican Republic (small-vulnerable economies coordinator, also G-33), Ecuador (tropical products, recent new member), Egypt (G-20, African Group), EU, India (G-33, G-20), Indonesia (G-33 coordinator, also G-20, Cairns), Jamaica (ACP coordinator, also G-33, small-vulnerable), Japan (G-10), Kenya (G-33, African, ACP), Rep. Korea (G-33, G-10), Lesotho (least-developed countries coordinator, also African Group, ACP), Mauritius (G-33, ACP, African), Malaysia (Cairns), Mexico (G-20), New Zealand (Cairns), Norway (G-10), Pakistan (Cairns, G-20, G-33), Paraguay (Cairns, G-20, tropical products, small-vulnerable), Philippines (G-33, G-20, Cairns), Switzerland (G-10 coordinator), Chinese Taipei (recent new members coordinator, also G–10), Thailand (Cairns, G-20), Turkey (G-33), Uruguay (Cairns, G-20), US, Venezuela (G-33, G-20)

(Previously, during 2007: Panama as recent new members coordinator; Uganda as African Group coordinator.)

> More on coalitions


Numbers in the draft tend to be in square brackets (indicating they are still to be negotiated) and in some cases the text offers ranges (e.g. tariffs) or alternatives (e.g. domestic support). Terms used in this box are explained in the longer summary.

Domestic support
Explanation of the “boxes”

  • Overall trade distorting domestic support (Amber + de minimis + Blue). EU to cut by 80%; US/Japan to cut by 70%; the rest to cut by 55%. “Downpayment” (immediate cut) of 33% for US, EU, Japan, 25% for the rest. Bigger cuts from some other developed countries, such as Japan, whose overall support is a larger % of production value. Cuts made over 5 years (developed countries) or 8 years (developing). (New: single figures from mid-points of ranges)

  • Amber Box (AMS). Overall, EU to cut by 70%; US/Japan to cut by 60%; the rest to cut by 45%. Bigger cuts from some other developed countries, such as Japan, Norway and Switzerland, whose AMS is larger % of production value. Also has downpayment. (Unchanged)

  • Per product Amber Box support: capped at average for notified support in 1995-2000 with some variation for the US and others. (Unchanged)

  • De minimis. Developed countries cut to 2.5% of production immediately. Developing countries to make two-thirds of the cut over three years to 6.7% of production (no cuts if mainly for subsistence/resource-poor farmers, etc). (Applies to product-specific and non-product specific de minimis support) (Minor change)

  • Blue Box (including “new” type). Limited to 2.5% of production (developed), 5% (developing) with caps per product. (Modified flexibilities for more vulnerable countries)

  • Green Box. Revisions — particularly on income support, to ensure it really is “decoupled” (ie, separated) from production levels, and on developing countries’ food stockpiling — and tighter monitoring and surveillance

Market access

  • Tariffs would mainly be cut according to a formula, which prescribes steeper cuts on higher tariffs. Ranges of cuts are now all single numbers. For developed countries the cuts would rise from 50% for tariffs below 20%, to 70% for tariffs above 75%, subject to a 54% minimum average, with some constraints on tariffs above 100%. (For developing countries the cuts in each tier would be two thirds of the equivalent tier for developed countries, subject to a maximum average of 36%.)
  • Some products would have smaller cuts via a number of flexibilities designed to take into account various concerns. These include: sensitive products (available to all countries), the smaller cuts offset by tariff quotas allowing more access at lower tariffs; special products (for developing countries, for specific vulnerabilities), with single figures proposed instead of ranges.
  • Contingencies. Developed countries will scrap the old “special safeguard” (available for “tariffied” products). The option for them to keep some has been removed. More proposed details of the new “special safeguard mechanism” for developing countries are in an additional paper.

Export competition

  • Export subsidies to be eliminated by end of 2013. Half of this by end of 2010.

  • Revised provisions on export credit, guarantees and insurance, international food aid (with a “safe box” for emergencies), and exporting state trading enterprises.

Details …

Domestic support

Explanation of the “boxes”

Background explanation: Cutting trade-distorting domestic support would operate simultaneously through several layers of constraints. Each category of supports would be cut or limited:

  • Amber Box (the most distorting, with direct links to prices and production, officially aggregate measurement of support or AMS)

  • De minimis (Amber Box but in relatively smaller or minimal permitted amounts defined as 5% of production for developed countries, 10% for developing countries)

  • Blue Box (less distorting because of conditions attached to the support)

Second, for each of these, there would also be some constraints on support for individual products (“product-specific”).

Third, on top of that would be cuts in the permitted amounts of all three combined:

  • “Overall trade-distorting domestic support” (OTDS)

(News reports of some countries being asked to cut their supports to certain amounts of dollars or euros are referring only to that last “overall” discipline.)

In these “modalities”: The cuts would be achieved by two methods (these are cuts in permitted ceilings, which may or may not bite into actual spending):

1. Tiered formulas. Like the tariff formula, the formulas for the Amber Box and overall distorting support are also expressed as “tiers” with support in the highest tier having the steepest percentage cuts. Countries with larger support go into higher tiers.

2. Limits (or cuts resulting in limits). For de minimis, Blue Box and support for each product.

Overall trade-distorting domestic support
(Amber + de minimis + Blue)

Most of this is essentially unchanged except that ranges in the formulas have been replaced by single numbers and additional flexibilities have been provided for more vulnerable countries. Cuts are to be made from figures for a base period of 1995–2000 (paragraph 1)


  • Highest tier (above $60bn, i.e. EU), cut by 80%.
    (EU’s starting point or “base level” — a combination of the current ceilings in Amber Box and “de minimis” support plus a limit on Blue Box support that applies when the concept of “overall trade-distorting domestic support” kicks in — for 15 members is estimated at €110.3bn. The cut would bring the ceiling down to €22.06bn)

  • Middle tier ($10bn–$60bn, i.e. US, Japan), cut by 70%
    (US’s starting point is estimated at $48.2bn. The cut would bring the ceiling down to $14.46bn) (Japan would make a bigger effort because its overall support ceiling is more than 40% of the value of its agricultural production — its cut would be 75% — Par.4)

  • Lower tier (below $10bn. i.e. all other developed countries), cut by 55%

Downpayment: 33.3% is cut from the start of the implementation period (a “downpayment”) for the top three subsidizers (ie, EU, US and Japan); 25% for other developed countries (Par.5)

Implementation: 6 equal annual steps over 5 years for developed countries, 9 steps over 8 years for developing (Pars.5, 8).

Base level: the starting point for the percentage cuts. This is needed because the concept of “overall trade-distorting domestic support” is new, because there is a new type of Blue Box programme, and because previously there were no limits on Blue Box payments. When countries make no cuts, they have to stay within the base-level amounts (except least-developed countries, and those net food importing developing countries that do not use the Blue Box) (Par.10) (Modified for net food importers).

The base level for developed countries = Amber Box commitment ceiling + 15% of the value of production (comprising 5% for current “de minimis” support for agriculture in general, another 5% for “de minimis” support targeted at specific products, and 5% for Blue Box support) — the Blue Box component could be higher if the actual support in this category was more than 5%. (Par.1)

Developing countries. Those with Amber Box commitments (ie, with ceilings higher than the minimal “de minimis” level and therefore required to reduce the ceilings): cut by two-thirds of the formula cut. But net food-importing countries (Tunisia, Morocco, Jordan, Venezuela) among these would be exempt. (Par.7) Those without Amber Box reduction commitments, would not have to reduce overall distorting support, but would have to stay within the base amount of support. Net-food-importing developing countries would not have any limits on overall distorting support if they do not use the Blue Box (but they would have to observe Amber Box and de minimis constraints). (Pars.6, 10)

Recent new members. New members who joined very recently (Saudi Arabia, the former Yugoslav Republic of Macedonia, Viet Nam, Ukraine), and some others with low incomes (Albania, Armenia, Georgia, Kyrgyz Rep, Moldova, Mongolia) would make no cuts. Others would make two-thirds of the formula’s cut. (Par.9) (Mongolia added)

Transparency: Included in the text is a requirement for some countries to provide their data on the value of production (used to calculate the overall limits) annexed to the “modalities”. These are all developed countries and those developing countries that have to cut their overall distorting support, ie, all countries whose Amber Box support ceilings exceed the minimal (“de minimis”) levels and have to be reduced — net food importing developing countries, least developed countries and some recent new members would not be included. (Par.12)

Amber Box (i.e. final bound total AMS)

(Par.13) (Unchanged)

  • Highest tier (above $40bn, i.e. EU), cut by 70%. (EU’s current ceiling is €67.16bn. Cut would bring ceiling down to €20.1bn)

  • Middle tier ($15bn–$40bn, i.e. US, Japan), cut by 60% (US’s current ceiling is $19.1bn; down to $7.6bn after cut.)

  • Lower tier (below $15bn. i.e. all others), cut by 45%

Japan would make the top tier cut, effectively putting it in the top tier. Other developed countries whose Amber Box support is more than 40% of the value of their agricultural production would also make a bigger cut, i.e. a cut halfway between the cut of their tier and the tier above. (Par.14) (Also unchanged)

Downpayment. The top three subsidizers (ie, EU, US and Japan) to cut 25% from the start. All other cuts in equal annual steps over five years (eight for developing countries). (Par.15) (Unchanged)

Various developing countries would make two-thirds of the formula cut or be exempt cuts (if their present ceilings are below US$100m — new — or if they are net food importers), and would continue to be allowed some types of support. (Pars. 16–18) (Unchanged)

Recent new members. New members who joined very recently, and some others with low incomes (Saudi Arabia, the former Yugoslav Republic of Macedonia, Viet Nam, Ukraine; Albania, Armenia, Georgia, Kyrgyz Rep, Moldova, Mongolia) would make no cuts. Some would be allowed to exclude investment subsidies from Amber Box calculations. Some would make two-thirds of the formula cut. (Par.19) (Mongolia added)

Inflation can have an effect on calculations of support, which in turn could run foul of committed limits. The text says allowance for this under the Agriculture Agreement will continue in effect. A sentence adds this will include consideration for developing countries facing sharp rises in food prices. (Par.20) (Unchanged)

Amber Box support per product would be limited to no more than the amounts actually provided on average in 1995–2000 (with some variation for developing countries). The calculation for the US would be based on total Amber Box support for specific products per year for that period but shared among products according to the average share over the years 1995–2004. Some additional adjustments would be made for special situations. Developed countries would annex data on their limits-per-product to the modalities. (Pars.21–29) (Unchanged.)

De minimis

(Amber Box supports in relatively small or minimal amounts, currently limited to 5% of production in developed countries, 10% in developing)

  • Developed countries: cut by 50% from day one (i.e. cap at 2.5% of the value of production, from the current 5%) (Par.30) (Unchanged)
  • Developing countries with Amber Box commitments: cut two-thirds of the above cuts (from the current 10% of the value of production, ie, ending up with about 6.7% of the value of production). Exempt from cuts: if almost all is for “subsistence and resource-poor farmers” or the country is a net food importer. (Pars.31–32) (Unchanged)
  • Recent new members: no cuts for those who joined very recently and some with low incomes (Saudi Arabia, the former Yugoslav Republic of Macedonia, Viet Nam and Ukraine; Albania, Armenia, Georgia, Kyrgyz Rep, Moldova, Mongolia). Others make at least one-third of the standard cut. (Par.33) (Mongolia added)

Blue Box

New type. (The present Blue Box distorts trade but the distortion is limited; it’s for direct payments to farmers based on the number of animals they have or the area planted, but with production limits so that over-production is curbed.) The Agriculture Agreement would be amended to add a new type of Blue Box based on payments that do not require production but are based on a fixed amount of production in the past (eg, for US “countercyclical payments”). (Par.35)

A country would have to decide which type of Blue Box to use. It would normally only use one type for all products and this would not change. Any exceptions would have to be approved now (when “schedules” of commitments are agreed). In any case, any product can only receive one type of Blue Box support. (Par. 36–37)

Limit: 2.5% of the value of production during the base period (Par.38). More is allowed for some countries (such as Norway) that now use a lot of Blue Box support as they reform their support by shifting away from the more distorting Amber Box — if the Blue Box support is more than 40% of trade-distorting support, it is cut by the same percentage as the Amber Box cut over two years (Par 39)(square brackets removed on two years). Developing countries: 5% of the value of production, with flexibility for some special circumstances. (Pars.48–50) Recent new members: 5% of the value of production, with some flexibility over the base period. (Par.51) (Unchanged, except for net-food-importers)

Other criteria: The 2008 texts spell out in greater detail how limits would also be imposed on Blue Box support for each product. Generally the limits are the average spent in 1995–2000, with adjustments if there are gaps in spending in some years. For the US, the limits are 10% or 20% more than estimates of maximums under the 2002 Farm Bill (sometimes called “headroom”). US data are in Annex A. Various provisions deal with a range of situations, including the possibility of going above Blue Box limits per product if an equivalent reduction is made in the Amber Box limits for that product, and for enabling Blue Box payments on products that did not previously receive them. For developing countries the combined Blue Box limit on these “new” products is 30% of the overall Blue Box limit, with a maximum of 10% for any single product, and flexibility for least developed and net food importing developing countries. (Pars.40–50) (The last point modified)

Green Box

(Ie, support that does not distort production or prices or causes minimal distortion.) The Agriculture Agreement’s provisions (its Annex 2) would be amended to allow more development programmes by developing countries and to tighten criteria for developed countries (e.g. on decoupled income support). The July revision further refined provisions dealing with the question of “fixed and unchanging” base periods for income support, structural adjustment and regional assistance programmes (including the notion that farmers expectations or decisions must not be altered by any exceptional changes). The July and December texts refine conditions so that some government intervention in developing countries is counted in the Green Box and not the Amber Box. These deal with purchases for stockpiling or to fight hunger and rural poverty and involve governments buying from low-income farmers or those with few resources, including at prices that are higher than the market. (Annex B) (Some modification)

(Some members have argued that in order to ensure Green Box programmes are genuinely “green” (i.e. non-distorting), transparency, monitoring and surveillance should be enhanced. This would be part of a general revision of monitoring and surveillance — Annex M)


Trade-distorting domestic support for cotton would be cut by more than for the rest of the sector. The text includes a formula reflecting this, based on a formula proposed by the “Cotton Four” African countries in 2006. (Par.54) (Unchanged)

Mathematically, the formula says that if a country’s general Amber Box cut is “Rg”, then, the percentage cut for cotton = Rg + ((100-Rg)x100)/3xRg

Eg, if the US Amber Box reduction is 60%, as above, then its cut in Amber Box support for cotton would be 82.2% i.e. (60+(40x100/180))%. That is unchanged and remains unsettled.

Blue Box support for cotton would be capped at one-third of what would be the normal limit (Par.55). (Unchanged)

Developing countries with Amber and Blue Box commitments would make two-thirds of developed country cuts for cotton and over a longer time period (Pars.57 and 58). (Unchanged, although the chairperson’s introduction does refer to cotton and the need to make progress on this critical issue)


Market access

Tariff reduction formula: the bottom line

The tiered reduction formula is the main approach for cutting tariffs (from the maximum rates that each country has legally bound in the WTO — “bindings” or “bound rates”). Products are categorized by the height of the starting bound tariff (Year 0 in the charts below). Products in higher tiers have steeper cuts. The 2008 texts have gradually replaced ranges of possible cuts with single figures. The latest to go was the top tier range, a single number in the December text (for details see charts on next page).

For developing countries, the standard cuts in each tier would be two-thirds of the equivalent cut for developed countries. The time periods for making the cuts are five years for developed countries and — new — 10 years for developing. Cuts would be made in equal annual steps, starting from the first day of implementation.

However, the general tiered formula will not apply to all products. Some flexibility is spelt out for some products (details below), including those that are politically “sensitive” and those that are “special” because they affect food security, livelihood security and rural development in poorer countries.

Developing countries have more exceptions, particularly the smallest and most vulnerable among them — the text lists around 45 small and vulnerable economies, meaning that over half of developing countries that are not least-developed would be eligible for even smaller reductions (Annex I). Least-developed countries and some recent new members will not have to make any cuts (Par.67, 151).

The charts (next page) indicate the scale of cuts for the two groups of countries. The purpose is only to illustrate how the formula works and to allow developed and developing countries’ cuts to be compared. The solid lines compare developed and developing countries’ cuts from starting tariffs that are mid-points in the developed countries’ lower three tiers and arbitrarily 100% in their top tier. The dotted lines show cuts from mid-tier or 150% in the top tier, for developing countries.

Note that the special treatment for developing countries can sometimes work doubly. Not only are the cuts in each tier gentler, but many products (such as those with a 100% tariff) fall into a lower tier in the formula (top tier for developed, upper middle tier for developing), meaning that the cut is even gentler.

The only products that are in the same tier for both developed and developing countries are those with tariffs above 130% (top tier), those with tariffs of 30%–50% (lower middle tier), and those with tariffs below 20% (bottom tier).

However, the tariff formula is by no means the whole story …

Download this
> Word  > pdf


Developed countries

Developing countries


Latest: developed countries 5 years

Top tier: tariffs above 75% — cut by 70%
Upper middle tier: tariffs below 75%, above 50% — cut by 64%
Lower middle tier: tariffs below 50%, above 20% — cut by 57%
Bottom tier: tariffs below 20% — cut by 50%

Subject to a minimum average cut of 54%, taking into account deviations from the formula — both larger and smaller cuts than the formula. If the result is a smaller average, then additional reductions would be made. (Pars.61-62)

Latest: developing countries
Two thirds of developed countries’ cuts in each tier, 10 years

Top tier: tariffs above 130% — cut by 46.7%
Upper middle tier: tariffs below 130%, above 80% — cut by 42.7%
Lower middle tier: tariffs below 80%, above 30% — cut by 38%
Bottom tier: tariffs below 30% — cut by 33.3%

Plus a maximum average cut of 36%. If the average is more than that, the cut by the formula can be reduced. (Par.63–64)

Flexibilities in brief: deviations and exemptions from the bottom line

For developing countries these could be quite extensive, and in some cases the bottom-line formula could be the exception rather than the rule, or it could be discarded completely:

  • Sensitive products (available for all) would have smaller cuts than from the formula, but with new quotas allowing imports at lower tariffs (“tariff quotas”) to provide some access to the market. Deviations would be one-third, half or two-thirds of the cut, with the tariff quota adjusted in relation to the deviation. (More details below)

  • Maximum average cut (developing countries) — 36%. Developing countries could reduce the formula’s cuts in order to stay within that average maximum. The average would take account of all deviations from the formula, including the smaller cuts made on sensitive products. (Par. 64)

  • Smaller maximum average cut without using the formula at all (45 small and vulnerable economies) — 24% achieved by designating products as “special” (see below) if they deviate from the formula including exemption from cuts, and no need to use indicators. (Pars.65, 157 and Annex I)

  • … or smaller cuts by 10 percentage points (45 small and vulnerable economies, those with “ceiling binding”, those with “low homogeneous bindings”). Suriname would be able to fix its bound tariffs at the average for the rest of the CARICOM Caribbean states (Pars.65, 130 and Annex I) (New: Suriname)

  • Smaller than formula cuts (other recent new members) — cuts can be reduced by up to 10% in the two top bands and 5% in the two bottom bands, starting one year after their current membership deals have been implemented fully and perhaps with two additional years to implement the new agreement. (Pars.66–70)

  • Would not have to make any tariff cuts: least-developed countries, “very recent” new members (Saudi Arabia, the former Yugoslav Republic of Macedonia, Viet Nam, Tonga, Ukraine), small low-income recent new members (Albania, Armenia, Georgia, Kyrgyz Rep, Moldova, Mongolia), and exceptionally Bolivia. (Pars.65 (footnote 9), 67–70, 151) (New: Mongolia)

  • Special products (developing countries) — The revised text retains the two-tier structure, removes the option of no products escaping tariff cuts, and replaces ranges with single figures. Now, 12% of products could be declared “special” guided by indicators for food and livelihood security or rural development (indicators in Annex F). Up to 5% of products (6% in the previous draft) could be exempt completely from cuts. In any case, the tariff cuts on special products would have to average 11%. (See above for small and vulnerable economies. Recent new members have different conditions — 13% of products can be special with a 10% average cut.) (Pars.129–131) (Revised)

Tariff cap

This version retains the incentive for countries to restrict tariffs above 100% to no more than 4% of sensitive products (or products with quotas) with separate treatment for sensitive and non-sensitive products.

Sensitive products are allowed to have tariffs above 100% but those that do have to add 0.5% of domestic consumption more to the tariff quota.

In square brackets: for Iceland, Japan, Norway and Switzerland, some non-sensitive products’ tariffs could also exceed 100%. These could be limited to 1% of tariff lines. There would be a payment for this: 0.5% more of domestic consumption added to the tariff quotas on all sensitive products, or the tariff cut would made 2 years faster than normal, or the tariff cut would be 10 percentage points more than normal. (Par.76) (Revised)

Sensitive products (all countries)

What and how many? These are sensitive essentially for political reasons — smaller cuts than the formula, can be made by all members. For DEVELOPED countries 4% of products could be “sensitive” (or two percentage points more if more than 30% of products fall into the top tier of the formula). The text notes that Canada and Japan oppose this, an issue discussed in an additional paper, TN/AG/W/5 (Par.71)

The text has two options for selecting sensitive products. Either they have to be products that already have tariff quotas (before the Doha Round) — sometimes confusingly abbreviated to the slogan “no new tariff quotas” — or any product can be sensitive. (Par.83) The chairperson says in a separate paper (TN/AG//W/6) that in his view neither option will be agreed. He offers some suggestions for how creating new tariff quotas could be constrained and the compensation required for doing so, including in some specific situations.

What tariff cut? the tariff cut would deviate from the formula cut by one-third, half or two-thirds of the formula cut, without any variation within a broad product category. (Par.73) (Revised)

For developing countries, one-third more (5.3% or 8%) of products (Par.72). The deviation would be the same as for developed countries. (Par.73)

The payment — some more market access, via a “tariff quota” (where quantities inside the quota are charged lower or no duty. The out-of-quota tariff is the normal rate determined by the reduction formula).

In return for being allowed a smaller tariff cut, developed countries have to allow at least some quantities into their markets at a lower tariff (inside the tariff quota, which expands if a quota already exists). This new “access opportunity” would be 4% of domestic consumption if the full two-thirds deviation is applied, 3.5% if only half the cut is made, or 3% if the deviation is the smaller one-third. (Par.74) (Ranges replaced with single numbers)

The text allows countries more sensitive products (by 2 percentage points, ie, 6% of products) if they have a large number of products (more than 30% of products) in the top band of the tariff-cutting formula. For those additional products they would have to provide additional access of 0.5% of domestic consumption to their markets.

But they can provide less access if normal imports are comparatively large. The quota expansions have to be made available to all members on equal terms (“most-favoured-nation”). (Pars.74–75, 77, 82)

For developing countries the quota expansion is two-thirds of the amounts for developed countries, and domestic consumption (see below) does not include subsistence farmers’ consumption of their own produce. Developing countries could choose to specify sensitive products without providing tariff quota access. Three complex options are offered where some products take smaller cuts than the formula, the number of these and the time period for making the cuts varying according ot the deviation from the formula. The remaining products would either take the full cut over a 3-year longer period, or deviate from the formula with quota expansion. Or, developing countries could apply the full cut to all sensitive products but over a longer period. (Par.78–81) (Significantly revised)

Complexity — domestic consumption. Behind these broad principles lie some highly complex questions. A considerable amount of progress has been made since the February and May drafts in trying to resolve different positions on these.

A major question is the extent of disaggregation for identifying “sensitive products” and for the tariff quotas. Must a sensitive product be a broad category such as “cheese”? Or can it be “hard cheese”, or even more detailed such as “cheddar cheese”? (“Partial designation” is the term used when countries consider some but not all products in a category to be sensitive.)

When products identified as “sensitive” are defined at a detailed or disaggregated level, this creates problems for what happens to those products. The more detailed the products, the greater the problems. There are two reasons for this. First, domestic consumption is going to be the yardstick for new or expanded tariff quotas, but data are not usually available for narrowly defined products such as cheddar cheese or wheat flour. Therefore consumption has to be estimated using “proxies” based on trade figures for the more detailed products — a subject of divergent opinions. Second, subcategories of products can be substitutes (which means they can compete with each other), so the distinctions are not always clear-cut.

The latest texts reflect the result of intensive and highly technical consultations. They describe how domestic consumption should be estimated when sensitive products are identified at high levels of detail.

The method starts out by listing products that members have said could potentially be declared “sensitive” (template in Attachment A — modified). The list defines broad product categories by specifying the more detailed products each category includes (identified at the 6-digit coding level of the World Customs Organization’s harmonized system — HS6). All categories have at least some “core” products, ie, raw or basic traded goods. Non-core products are split between those that have seen a lower amount of processing, and those that are highly processed. (Included in this revision are a number of changes to the list. Meanwhile, discussions continue.)

For example the product category “wheat” comprises 28 products identified as 6-digit codes (HS6). These include two types of wheat in the form of basic grain as “core” products, several products that have undergone a stage of processing, such as wheat flour, and finally some highly processed products such as pasta and bread.

The method then spells out how to calculate domestic consumption for each broad category, using available data (template in Attachment B). Finally (template in Attachment D), it prescribes how to estimate the consumption of products identified at a more detailed level, first at the 6-digit coding level (HS6), and then at a more detailed level such as 8-digit (HS8). Each detailed product’s consumption is a percentage of the broad category’s consumption, the percentages based on the product’s share of trade in the broad category, but adjusted to ensure that normally the “core” products — which are usually the most heavily traded — have 90% or more of the category’s consumption.

(The HS6 products’ consumption figures are assumed to be the same percentages of the product categories’ consumption for all members, but for HS8 products, depend on the shares of imports in each country. Note that under the Harmonized System, the HS6 codes are the same for all countries, but beyond that for HS7, HS8, etc, the codes vary from country to country.)

These estimates would be used to determine quota sizes when the more detailed products are declared sensitive. Normally, the size of the tariff quota would depend on the estimated consumption of the sensitive products within the same broad product category. And normally, this would have to be a single tariff quota. In a few cases (no more than three product categories), a country could set two tariff quotas within a single category.

Other disciplines, together with some flexibilities, are included to prevent the estimates leading to quotas that are too small — including a minimum quota size (“floor”) to cover cases where trade figures used (as “proxies”) to estimate domestic consumption are exceptionally low. (A summary of how this works is in the diagram on the next page. See Annex C and Attachment Ai of paper and additional attachments for details.)

Flow chart: example of estimating domestic consumption for tariff-quotas. Download: Word, pdf
or click on image to open a full-size image (GIF format, 23KB)

Additional criteria and other issues

“Tariff escalation” (the problem of higher tariffs on processed products than on raw materials, which hinders processing for export in the country producing the raw materials). Where the escalated processed product has a tariff that is significantly above the unprocessed product (ie, by 5 percentage points or more), it would take the cut of the tier above or if it is already in the top tier, 6 percentage points added to the cut of the top tier. Sensitive products would be exempt, and the tropical products cut would override the escalation cut if it is bigger. (Pars.84–90 and Annex D)

Commodities: This aims to strengthen provisions on tariff escalation for developing countries depending on commodity exports. It includes possibilities for eliminating non-tariff barriers and for price stabilization. (Pars.91–102)

Simplifying tariffs. The text includes options for all tariffs to end up as simple ad valorem (percentages of the price) or for this to be delayed for 10% of products subject to certain conditions. But in any case no tariffs would be made more complex than they are already. For the EU, 85% could be ad valorem after 5 years, with 5% kept as compound or mixed tariffs. And, in any case, the most complex tariffs (“complex composite matrices”) have to be simplified, either as ad valorem or specific duties (dollars, euros etc, per tonne, litre, etc). The text includes more technical issues such as the method of converting tariffs to their ad valorem equivalents. (Par.103–108, Revised, Annex N, new)

Tariff quotas (where a higher tariff is charged on quantities outside the quota, and a lower or zero duty for quantities inside. The out-of-quota tariff is the normal rate determined by the reduction formula). The latest revision includes provisions on bound in-quota tariffs, how much they should be cut, and whether new quotas should have zero in-quota duties. Under a simplified formulation, in-quota tariffs would be cut by 50% or to 10%, whichever gives a lower result (the 10% acting as a ceiling on the tariffs), while tariffs of 5% or less would be eliminated within a year. Developing countries would make a 15% cut, and small and vulnerable economies 7.5%, without a cap or elimination. Recent new members make smaller cuts and no cuts on low tariffs and those of the very recent new members and Venezuela would not be cut at all. There would be no cuts on in-quota tariffs of special products (Par.109–114) (Revised)

Provisions on tariff quota administration refer to the WTO Import Licensing Agreement with additional criteria. (Pars.115–119) The text includes the proposed treatment of cases where quotas are not filled (Par.120–125) and includes a new proposed compromise on monitoring tariff quotas and improving access to the market if imports are persistently less than the quota (“underfill”). (Annex E)

Tropical and diversification products and long-standing preferences: the provisions are designed to accelerate liberalization of tropical products — alternative proposals suggest imports could be duty-free if the present tariff is no more than 25% or 10%, otherwise having a range of cuts, depending on the proposal. Slower liberalization for products with long-standing preferences — alternative proposals suggest a 10-year delay in starting tariff cuts or simply two years longer to make the cuts. Where the two overlap, the tropical products (and tariff escalation) provisions could override those of preferences, except for some products (still to be identified). Recent work has focused on negotiating the lists of products in each category, but the discussion continues and therefore the lists remain unchanged. (Pars.147–150, products listed in Annexes G and H)


1. Special safeguard (SSG). Cut immediately to 1% of products and eliminate the current “special safeguard” after seven years. While still in use, the SSG could not raise a tariff above its pre-Doha Round bound rate. The new text includes provisions on additional tariff quota expansion. Developing countries would cut the number to 2.5% of products immediately, small and vulnerable economies to 5% over 12 years. (This safeguard can be used on products whose variable duties, discretionary import licensing, quotas or import bans were converted to tariffs in the Uruguay Round, and many developing countries gave up their right to use it because they chose to set ceiling bindings instead of to “tariffy”.) (Pars.126–128) (Some modifications)

2. (the new) special safeguard mechanism (SSM). This remains a difficult subject and the chairperson has issued an additional paper. Developing countries would be able to temporarily protect their producers by applying the new special safeguard mechanism. The main text proposes options for formulas for the mechanism, and includes possible disciplines to avoid the safeguard being triggered frequently and frivolously, and suggests when (if at all), and by how much, the increase in tariffs can exceed present bound ceilings (or “Pre-Doha Round bindings”), with more leniency proposed for small and vulnerable economies than other developing countries. (Pars.132–146)

The chairperson’s additional paper (TN/AG//W/7) offers draft text (with options) for when the SSM raises tariffs above pre-Doha bound rates: when it would be triggered, how high the tariff would go, how long it would last, when it could be triggered again, whether it could be triggered when prices are not falling. (Note that while the issue here is the case where the tariff goes above pre-Doha Round bindings, the increase — the “remedy” — is based on the “current” or post-Doha bound rate.) He observes that members remain wide apart on some additional issues related to seasonal perishable products and ensuring “normal” trade is not disrupted. In general, questions also remain on least-developed countries, small and vulnerable economies, and provisions for when tariffs do not go above pre-Doha rates. (New)
See also: Unofficial guide to agricultural safeguards

Least-developed countries

Least-developed countries would not have to reduce tariffs. The latest text also deals with duty-free and quota-free market access for at least 97% of products from these countries and says these products have preferential rules of origin (which determine whether a product qualifies as coming from a least-developed country) (Pars.152–154) (Revised)

Export competition

Export subsidies

Eliminate by the end of 2013 (developed countries), with half cut by the end of 2010, and revised details for cutting the subsidized quantities in the period. The elimination date for developing countries would be 2016. (Pars.162-163) (Unchanged). The text ensures commitments on net food-importing and least-developed countries are unaffected. (Par.164)

Export credits, export credit guarantees or insurance programmes

These would be disciplined to avoid hidden subsidies and ensure the programmes operate on commercial terms. Proposed conditions include limiting the repayment period to 180 days, ensuring programmes are self-financing (ie, not making losses over a period), etc. An earlier revision greatly simplified the text on self-financing: instead of listing criteria it just refers to recovering costs “to a commercially viable standard”, over a “rolling” period of four or five years. (Annex J) (Unchanged)

For developing countries providing credit, the 180-day maximum repayment term would be reached in three steps over a period, probably four years (or by 2013 if that’s earlier). Least-developed and net food-importing developing countries would be normally be allowed 360–540 days to repay (previously 360 days). Some additional flexibility in special cases would be allowed, monitored by the WTO Agriculture Committee. (Annex J) (Minor modification)

Agricultural exporting state trading enterprises

Their activities would be disciplined. A key question remains whether monopoly power would be outlawed or just disciplined. The definition of exporting state trading enterprises was simplified in the February text by referring to the relevant provisions in the General Agreement on Tariffs and Trade (Art.17). (Annex K) (Minor modification)

International food aid

Emergency food aid would be in a “Safe Box” with more lenient disciplines. Emergencies would be declared or appealed by relevant international organizations such as the UN, World Food Programme, Red Cross, etc.

Other food aid (ie, not emergency aid) would be disciplined to prevent the aid from displacing commercial trade, and with needs assessment, which would be under the responsibility of a UN agency.

The text gives the recipient government priority over all food aid operations, emphasizes needs assessment, and gives the UN the final say when NGOs assess needs. Members’ continuing differences over monetization (ie, selling donated products to raise funds for aid) no longer have options, offering revised disciplines the practice. It could be permitted under certain conditions both in emergencies and in other situations. (Annex L) (Modified)


Export subsidies would be eliminated from the start of the implementation period. (Par.168–169) (Unchanged)

Export prohibitions and restrictions

Disciplines would be tightened for introducing new export restrictions, with increased transparency and monitoring and a new paragraph on consultation. (Pars.171–180) (Slightly modified)

Other issues

Monitoring and surveillance

The text includes proposals for a flexible institutional structure based on the WTO’s regular Agriculture Committee. It includes clearer obligations on member governments to keep each other informed (through “notification”) on what they do under the agreement, including on export restrictions. The surveillance mechanism would be reviewed every five years. (Annex M) (Slightly revised)

(The following remain in square brackets with no other text, indicating no narrowing of opposing views.)

[Sectoral initiatives] (Duty-free trade in a particular sector) (now deleted)

[Differential export taxes] (Higher export duties on raw materials than on processed products — the mirror image of tariff escalation)

[Geographical indications] (Names of products — often food — that are identified by their origin and characteristics)

The annexes

Annex A: United States — Product-Specific Blue Box Limits

Annex B: The Green Box (“Annex 2 of the Agreement on Agriculture shall be amended as follows”)

Annex C: Basis for the Calculation of Tariff Quota Expansion

Annex D: Tariff Escalation Provisional Potential List (now complete with cocoa and cereals added)

Annex E: Tariff Quota Underfill Mechanism

Annex F: Illustrative List of Indicators for the Designation of Special Products

Annex G: Proposed List of Tropical and Alternative Products and Indicative List of Tropical Products Used in the Uruguay Round)

Annex H: Proposed Indicative List of Preference Erosion Products

Annex I: Small, Vulnerable Economies

Annex J: Possible New Article to Replace the Current Article 10.2 of the Agreement on Agriculture — Export Credits, Export Credit Guarantees or Insurance Programmes

Annex K: Possible Article 10 bis of the Agreement on Agriculture — Agricultural Exporting State Trading Enterprises

Annex L: Possible New Article 10.4 to Replace the Current Article 10.4 of the Agreement on Agriculture — International Food Aid

Annex M: Monitoring and Surveillance

Annex N: Tariff Simplification Methodology

Attachment A–G: Lists of potential sensitive products, “Partial Designation Modalities for Sensitive Products” and a set of templates for calculating and estimating domestic consumption. (Revised)

Jargon buster 

boxes: categories of domestic support
• Amber Box: domestic support considered to distort production and trade, eg, by supporting prices or being directly related to production quantities, and therefore subject to reduction commitments. Officially, “aggregate measurement of support” (AMS)
de minimis: Amber Box supports in small, minimal or negligible permitted amounts (currently limited to 5% of the value of production in developed countries, 10% in developing). To simplify this guide to the “modalities”, de minimis is treated separately from the Amber box
• Blue Box: Amber Box types of support, but with constraints on production or other conditions designed to reduce the distortion. Currently not limited
Green Box: domestic supports considered not to support trade or to cause minimal distortion and therefore permitted with no limits
distortion: when prices are higher or lower than normal, and when quantities produced, bought, and sold are also higher or lower than normal — ie, than the levels that would usually exist in a competitive market
tiered formula: a formula where higher tariffs have steeper cuts than lower tariffs — products with higher tariffs are put in a higher category or tier, which has a steeper cut than lower tiers. Also used for cutting domestic support
• tariff quota: when quantities inside a quota are charged lower import duty rates, than those outside (which can be high). (The reductions from the formulas apply to out-of-quota tariffs.)
• Tariff line: a product as defined in lists of tariff rates. Products can be sub-divided, the level of detail reflected in the number of digits in the Harmonized System (HS) code use to identify the product.
• export competition: term used in these negotiations to cover export subsidies and the “parallel” issues, which could provide loopholes for governments’ export subsidies — export finance (credit, guarantees and insurance), exporting state trading enterprises, and international food aid

> More jargon: glossary