Extending Special and Differential Treatment (SDT) in Agriculture
for Developing Countries

Christopher Stevens

Institute of Development Studies

24 April 2002

Introduction

The Doha Ministerial Declaration re-affirmed that ‘provisions for special and differential treatment are an integral part of the WTO agreements’, and it agreed that ‘all special and differential treatment provisions shall be reviewed with a view to strengthening them and making them more precise, effective and operational’ [para. 44]. The Ministerial Declaration also requires ‘modalities for the further commitments, including provisions for special and differential treatment, [to] be established no later than 31 March 2003’ [para. 14].

What might this mean in the context of the Agreement on Agriculture (AoA) negotiations? The argument put forward in this paper is that the instruments to give effect to special and differential treatment (SDT) are in a state of flux (because of changes in national and multilateral trade policy), but that the principles remain valid and justify the creation of new instruments. One feature of a new SDT regime is likely to be that there is greater differentiation of treatment between WTO members which, in turn, implies the establishment of objective criteria on which to determine the differentiation. Following WTO precedent, these criteria would probably be established by other international bodies to which reference would be made in the AoA (as with the UN-defined list of least developed states (LDCs)).

‘Old’ SDT

The history of SDT has been well covered [for example by Michalopoulos 2000, Whalley 1999 and Fukasaku 2000]. In essence, the argument is that:

♦ SDT had its origins in a view of trade and development that questioned the desirability of developing countries liberalising border measures at the same pace as industrialised countries;

♦ the popularity of this approach was (possibly temporarily) in decline in many developing country governments during the negotiation period for the Uruguay Round Agreement;

♦ consequently, many SDT provisions on border measures and subsidies envisage developing countries (other than the LDCs) following a similar path to that of the industrialised countries but at a slower pace;

other SDT provisions (particularly those covering positive support to developing countries via financial and technical assistance or technology transfer) were not agreed in a form that is enforceable within the WTO system.

In other words, the ‘actionable’ SDT provisions in the Uruguay Round Agreement are weaker than in its predecessors because support for the approach to trade and development that underpinned them was relatively weak during those negotiations. There are plenty of ‘non-actionable’ provisions, but many developing countries and observers are unhappy with implementation. The WTO Secretariat’s assessment of SDT implementation of the AoA [WTO 2000] is summarised in Table 1.

Could there be a ‘new’ SDT

The obvious question is whether new circumstances and new trade and development strategies would justify new, actionable, SDT measures. The default answer would seem to be positive. Unless there is now a view that all countries are equal, so that ‘one size fits all’, there exists a potential case for SDT. Since few industrialised countries fail to differentiate between geographical regions and/or social groups in their domestic economic policy it would be inconsistent for them to argue that globally – where objective differences between regions are much greater than is the case domestically – uniformity of policy should be the rule.

Table 1. Implementation of SDT under AoA: WTO Secretariat assessment

Provision	Usage
Superior market access for developing countries (Preamble)	Greater liberalisation on tropical products and some others of interest to developing countries.
Transition periods (Article 15.2)	Included in schedules.
Domestic support for agricultural and rural development subsidies (Article 6.2)	Included in schedules.
Higher de minimis for domestic subsidies (Article 6.4 (b))	Included in schedules.
Lower reductions for export subsidies (Article 9.2(b)(iv))	All ten developing countries with export subsidy reduction commitments have used this provision.
Export subsidies exempt from reduction (Article 9.4)	Included in schedules, and some developing countries have used the provision.
Exemption from reduction commitments on diversification of export prohibitions and restrictions (Article 12.2)	No developing country has notified the use of such measures.
Exemption from reduction commitments in respect of public stockholding for food security (Annex 2, para. 3, footnote 5)	Included in schedules, and several developing countries have used the provision.
Exemption from reduction commitments for food subsidies to the poor (Annex 2, para. 4, footnotes 5 & 6)	Included in schedules, and several developing countries have used the provision.
Exemption from reduction commitments for staple foods (Annex 5, Section B)	Schedules of Korea and the Philippines reflect recourse to this provision.
Developed country support for net food-importing developing countries and LDCs (Article 16.1 & 16.2)	Monitoring has been included in most meetings of the Committee on Agriculture.
Source: derived from WTO 2000: 23–7.

SDT is justified in cases where all three of the following criteria apply:

♦ a single, uniform policy is inappropriate for all WTO members;

♦ broad groups of states share similar characteristics;

it is possible to identify variations in the scope or implementation of WTO rules that are appropriate to the circumstances of each broad group.

If every country is different, the only place to introduce modulation into WTO commitments is in the national schedules. And shared characteristics among a group of states have no operational importance if it is not possible to identify instruments that answer to those shared differences from other WTO members. The task, therefore, is to identify the groups and the appropriate instruments.

One area in which there appears to be a consensus is that developing countries may require special treatment that they do not currently receive on WTO rules that are administratively or technically difficult or costly to implement [see Michalopoulos 2000, Whalley 1999, Finger and Schuler 2000, Henson and Loader 2001]. These relate in particular to some of the ‘new areas’ of trade policy. They touch agriculture most directly in relation to the Agreement on the Application of Sanitary and Phytosanitary Measures (SPS).

But they do not appear to be a special problem in the AoA. This is because the task of removing ‘old protectionism’ (i.e. border measures and direct subsidies) is less advanced for temperate agricultural products than for industrial goods and tropical agriculture. Since the AoA is doing for temperate agriculture what the GATT did for other goods decades ago, the ‘old issues’ still predominate.

But this does not mean that there is no scope for new SDT. One strong candidate is in the area of food security. The ‘special’ role of agriculture is a feature of several WTO members’ proposals to the Committee on Agriculture (not least that of the EU, with its concern for ‘multifunctionality’).

It is evident that:

♦ some agricultural exporting countries are highly suspicious of multifunctionality as a (none too) covert means of perpetuating industrialised country protectionism; but

for many developing countries agriculture does play a major role in the economy and in the livelihoods of vulnerable people that sets it apart from other sectors.

This would seem to be precisely the area where SDT could play a role: to distinguish between genuine and justifiable claims for special treatment (especially if those receiving it were unlikely to cause major distortions to world trade) and those that are not.

The rest of this paper begins an exploration of the characteristics of country groups in relation to food security, how they might be measured, and what SDT might be justifiable. It is not the only exercise in this direction [see, for example, Diaz-Bonilla et al. 2000], but it is in the nature of the task that these exercises are complementary.

What is meant by the term ‘justifiable’? The WTO has a dual mandate, and different answers could be given under each. One goal is to foster the liberalisation of world trade on the grounds that this is a common good. A foundation stone of the ‘old SDT’ was that this assumption was questioned for countries on the periphery. More recently, though, opinion has tended to swing towards the notion that liberalisation is generally good for development (and, therefore, that the burden of proof lies with those who wish to delay it).

But the WTO has another mandate which is, perhaps, a less controversial basis for SDT. This it to agree new rules for international trade. Since the WTO acts by consensus, rules can be adopted only if they are supported by all. And if some developing countries are reluctant to accept new rules, this will block progress. The justification for SDT is that it will provide an objective basis on which to modulate new rules and, hence, help consensus building.

Which States are Food Insecure?

The answer to the question ‘which states are food insecure’ is not obvious, since it is people not countries that are normally considered to be food secure or insecure. How can concepts and measures that have been developed in relation to individuals be applied to states? Is the prevalence of food insecurity best indicated by poverty indicators (like GDP per head), or are there countries that are especially food insecure even though they are not necessarily the poorest? And, having identified states that are food insecure, what practical modulation of current, or likely future, WTO commitments does this status justify?

What is ‘national’ food insecurity?

If food insecure states are to be given special support in the AoA negotiations (such as less onerous subsidy restrictions and priority food and financial assistance outside the WTO), it is important that they be identified correctly. By analogy to the analysis of individuals following Sen [Drèze and Sen 1990], the food security of a state can be said to depend upon:

♦ its production entitlements, which reflect the food that can be produced domestically;

♦ its trade entitlements, which reflect its ability to earn sufficient foreign exchange with exports to purchase imported food; and

its transfer entitlements, which cover food that can be obtained either directly through food aid or indirectly by (semi-)commercial imports financed through financial aid.

This suggests that the most food insecure states are those that combine inadequate domestic production with an export structure (not necessarily just for agriculture) that is unsatisfactory in terms of one or more of the following characteristics:

♦ low per capita value and poor growth prospects;

♦ heavy dependence upon a small number of commodities facing fluctuating supply or demand;

heavy reliance of exports on a single market with fluctuating demand.

It follows that neither low GDP nor dependence upon imported food are, by themselves, necessarily indicators of national food insecurity. Some modest importers would be more insecure than larger exporters – because they cannot afford greater imports! It is the combination of characteristics that is important.

Does existing WTO terminology adequately capture this combination? The answer is ‘No’. Within the WTO the term ‘food security’ is used in a much narrower sense and relates primarily to the adequate supply of imported food to member states. Some 20 states are recognised as net food-importing developing countries (NFIDCs), and the 49 LDCs also receive special attention. This usage reflects concern that the liberalisation of world agricultural trade could lead to a rise in world prices for commercial imports and a reduction in the volume of food aid. Yet only one NFIDC (Kenya) falls among the 30 countries with the lowest calorie availability (a fairly robust proxy for food insecurity), and three of those 30 are neither LDC nor NFIDC [UNDP 2000: Table 23 – see Appendix 1].

But if the existing definitions appear inadequate, how could they be developed? The entitlements analysis demonstrates that, in one sense, almost all aspects of the WTO may have food security effects. For example, the Multifibre Arrangement, by restricting the clothing industries of many developing countries, has certainly impacted adversely on trade entitlements. But it is hard to imagine a consensus emerging among WTO members for substantial and enforceable SDT treatment to all states satisfying such broad criteria unless it was restricted to the very poorest and smallest states.

The LDC group comes close to satisfying these broad criteria. All have a low level of economic diversification. But limiting differentiated treatment just to LDCs would represent a substantial retreat of SDT. A reasonable working assumption is that there exist some non- LDCs that are food insecure – but how are they to be defined in a way that commands respect? Some focusing will be necessary. An operationally effective definition is needed to allow modulation of those WTO rules with greatest food security implications.

A first step establishing the criteria for membership of this group is to identify the areas of WTO rule-making that might be problematic, and why. This is done in the next section. Then, an initial illustrative analysis is made of criteria that are relevant to such concerns and of the range of countries captured by various thresholds.

The scope of SDT

Current provisions

To the extent that the food security implications of multilateral reforms have been identified, the texts agreed in the WTO do not provide any direct operational measures to deal with the consequences except to the extent that slower and lower commitments on market access and domestic subsidies are adequate to deal with the problem. The assumption from the various declarations made in the final texts is that such modulation of domestic measures by food insecure states is not considered to be adequate.

The Marrakech Agreement includes, for example, a Decision on Measures Concerning the Possible Negative Effects of the Reform Programme on Least-Developed and Net Food-Importing Developing Countries which, inter alia, commits the signatory Ministers ‘to establish appropriate mechanisms to ensure that the implementation of the results of the Uruguay Round on trade in agriculture does not adversely affect the availability of food aid’ [WTO 1995]. The Ministers further agreed ‘to ensure that any agreement relating to agricultural export credits makes appropriate provision for differential treatment in favour of least-developed and net food-importing developing countries’ [ibid.]. But, as is clear from Table 1, little concrete action has been reported to the WTO.

There are two, not mutually exclusive, avenues for the future:

♦ to identify more clearly the types of domestic action that can be justified for food insecure states both in order to focus attention on the need to support them and to ensure that the freedom of manoeuvre of food insecure states is not limited by future rules;

to make the existing hortatory declarations enforceable, either by endowing the WTO with a financial facility or by making final agreement on the next AoA conditional upon appropriate decisions being made in the multilateral and bilateral bodies that have the relevant portfolio responsibilities and financial resources.

A precise identification of appropriate new SDT measures cannot be provided until there exists some greater understanding of the new rules likely to be adopted in the AoA negotiations. But there is an expectation that the new Round will cover all three of the main elements of the AoA ‘architecture’: market access, export subsidies and domestic subsidies. Any tightening of rules in these three areas will tend to cause concern in different groups of states. These are set out analytically in Table 2.

Table 2. Potential areas of concern over new agricultural trade rules

Rules on:	Potential legitimate concerns in:
Lowering import controls	Countries aiming to increase domestic agricultural production Food importing states
Reducing export subsidies	Food importing states
Reducing domestic subsidies	Countries aiming to increase domestic production

Market access

The interests of food insecure developing countries concerning market access will be largely focused on any obligations they accept in relation to their own barriers against imports. Many insecure countries either export primary commodities that face low barriers in OECD markets or, if they do not, they have preferential access to protected markets. For those exporting non-sensitive products, OECD liberalisation is unlikely to result in any significant change (and, arguably, developing country liberalisation is likely to be less important in stimulating world demand for their exports than is developing country growth). For countries with preferential access for sensitive products, preference erosion is likely to affect trade entitlements, but the main arena in which the pace of erosion is set is that of the importing countries rather than the WTO. Although multilateral liberalisation will erode (and eventually remove) preferences, it is unlikely that the current WTO Round will take more than one, modest step in this direction.

There are strong views about the desirability or otherwise of developing countries liberalising their import regimes. But this brings us back to the central question posed earlier, which is whether the WTO’s rule-making or liberalising mission should take precedence when the two cannot easily be advanced simultaneously. Even among those who favour a broadly liberal trade policy for developing countries, there is a recognition that the agricultural sector of some developing countries has been artificially depressed by previous policies (including, above all, neglect) and that increasing agricultural production is a high priority. There could be a case, therefore, for allowing poor countries with large agricultural sectors some relief on liberalisation in order to allow for greater incentives to domestic production.

Countries in which food imports represent a significant element in total supply but with fragile trade entitlements will also have an indirect interest in liberalisation of market access. To the extent that it contributes to a decline in production in the most heavily protected markets (mainly the OECD), it will tend to increase world prices.

Export subsidies

An increase in world prices following cut-backs in OECD production will tend to result in an adverse movement in the terms of trade of food importing developing countries which may have an impact on food security for those with limited opportunities to boost exports or to increase domestic production. Such countries will be affected more directly by curbs on the export subsidies currently provided by a small number of OECD states. Such curbs are likely to have an immediate impact on world prices, at least until the exports of non-subsidisers bounce back from their current, artificially depressed, levels to take advantage of the new opportunities.

Moreover, there may be longer-term effects. To the extent that exports from subsidisers are replaced by exports from non-subsidisers, there could be a reduction in the availability of imports to very poor countries even if world prices do not rise. This could occur if poor countries currently receive food aid or so-called ‘grey’ imports (that do not qualify as food aid, but are sold at below market prices) from a subsidising state. There is no reason to suppose that the increase in exports from non-subsidisers will be made available to the same poor countries and on the same terms as the concessional exports that they replace.

Domestic subsidies

Countries in which there is an objective need to boost domestic agricultural production will also be concerned by any new WTO rules that limit the scope for domestic subsidies. It has been well remarked that few developing countries are able (or willing) to provide subsidies for agriculture that come anywhere close to the current limits. But a further tightening of such limits might cause difficulties for some states.

Identifying and applying indicators

There are thus two categories of (possibly overlapping) countries that might be affected in different ways by change to the three principal elements of the current AoA architecture. These are:

♦ countries aiming to boost domestic agricultural production that may wish to increase incentives to farmers by keeping import prices high and increasing domestic subsidies;

food importing states with weak trade entitlements that may be concerned about their capacity to import sufficient food in future.

Relevant indicators

What indicators exist to identify the countries that would be most vulnerable to such changes? At present there exist the LDC group, which may equate to the first category of states, and the NFIDC group, which is focused on the concerns of the second category. But are these sufficient?

An analysis of the types of indicator that might be relevant is provided in Table 3. This takes the two categories of countries identified in Table 2 and lists for each some illustrative indicators. One group consists of those countries in which agriculture is an important source of livelihoods but production is low (where a legitimate emphasis of policy is to boost agricultural production – a task that might be made more difficult by curbs on import controls or domestic subsidies). The other consists of those countries that are dependent on imports for a significant part of domestic consumption but have weak trade entitlements (and which would be vulnerable, therefore, to sudden increases in world prices).

Table 3. Relevant indicators for SDT

Characteristics of country	Indicator
Agriculture is important source of livelihoods but production is low	• High share of agriculture in GDP • Low per capita calorie supply
Import dependence with weak trade entitlements	• High food imports as share of GDP • High vulnerability • Low per capita calorie supply

In both cases, it is a combination of characteristics that indicates particular vulnerability. For the first group of countries a necessary condition is that agriculture should represent a relatively high proportion of gross domestic product (GDP). But this would include wealthy countries or those with sufficient non-agricultural production that they can easily assure the food security of their populations. An additional criterion, therefore, is that average per capita calorie supply should be low.

Similarly a high share of food imports in GDP is a necessary criterion for establishing import dependency but not a sufficient one. Low per capita calorie supply will indicate which among such countries have substantial vulnerable populations. On top of these, some indicator is required of a country’s weak trade entitlements. The indicator suggested in Table 3 is the composite vulnerability index compiled under the auspices of the Commonwealth Secretariat [Commonwealth Secretariat/World Bank 1999].

Calorie supply: the basic indicator

To what extent do these criteria overlap, either with each other or with the existing LDC and NFIDC groups? Tables 4 and 5 show what happens when an attempt is made to identify a coherent group of countries which have the optimum combination of characteristics. Table 4 starts with per capita calorie supply. The FAO/WHO-recommended minimum level is 2,300 calories per day. Since there will be substantial variations between consumption levels within a country it is unrealistic to characterise as low calorie availability only those countries with an average per capita supply of less than this level. On the other hand, it would be inappropriate for WTO rules to give special consideration to countries just because they have highly unequal consumption patterns. A threshold of an average per capita calorie supply of 2,500 has been taken as an initial indicator to illustrate the range of countries that would be brought in by such a threshold. It allows for a limited degree of unequal calorie availability within a country.

Table 4 presents the 72 countries for which data are available that have an average per capita daily calorie supply of less than 2,500 in ascending order of calories. It also indicates whether or not the countries are classified as LDC or NFIDC.

It is evident that the LDC and NFIDC categories cover some but not all states. Twenty-seven of the countries are neither LDC nor NFIDC. Moreover, a further four states are classified as LDC but have a per capita calorie supply in excess of 2,500 (and range from Mauritania with 2,622 to Cape Verde with 3,015). Seven NFIDC states have calorie supply in excess of 2,500, ranging up to 3,287 (Egypt). Hence the LDC and NFIDC categories combined cannot be used as an adequate indicator of food insecurity.

Table 4. Average per capita calorie supply

Country	Daily per capita calorie supply a 1997	LDC	NFIDC b	Country	Daily per capita calorie supply a 1997	LDC	NFIDC b
Eritrea	1,622	Yes	 	Nicaragua	2,186	 	 
Burundi	1,685	Yes	 	Papua New Guinea	2,224	 	 
Congo Dem. Rep.	1,755	Yes	 	Guinea	2,231	Yes	 
Mozambique	1,832	Yes	 	Azerbaijan	2,236	 	 
Comoros	1,858	Yes	 	Lesotho	2,243	Yes	 
Ethiopia	1,858	Yes	 	Dominican Rep.	2,288	 	 
Haiti	1,869	Yes	 	Peru	2,302	 	Yes
Angola	1,903	Yes	 	Sri Lanka	2,302	 	Yes
Mongolia	1,917	 	 	Turkmenistan	2,306	 	 
Zambia	1,970	Yes	 	Venezuela	2,321	 	Yes
Kenya	1,976	 	Yes	Guatemala	2,339	 	 
Tanzania	1,995	Yes	 	Gambia	2,350	Yes	 
Tajikistan	2,001	 	 	Thailand	2,360	 	 
Central African Rep.	2,016	Yes	 	Antigua/Barbuda	2,365	 	 
Madagascar	2,021	Yes	 	Philippines	2,366	 	 
Mali	2,029	Yes	 	Nepal	2,366	Yes	 
Chad	2,032	Yes	 	Armenia	2,371	 	 
Sierra Leone	2,035	Yes	 	Sudan	2,395	Yes	 
Malawi	2,043	Yes	 	Honduras	2,403	 	Yes
Cambodia	2,048	Yes	 	Senegal	2,418	Yes	 
Yemen	2,051	Yes	 	Guinea-Bissau	2,430	Yes	 
Rwanda	2,056	Yes	 	Panama	2,430	 	 
Djibouti	2,084	Yes	 	Uzbekistan	2,433	 	 
Bangladesh	2,085	Yes	 	Bahamas	2,443	 	 
Uganda	2,085	Yes	 	Croatia	2,445	 	 
Niger	2,097	Yes	 	Kyrgyzstan	2,447	 	 
Lao PDR	2,108	Yes	 	Togo	2,469	Yes	 
Cameroon	2,111	 	 	St Vincent	2,472	 	 
Burkina Faso	2,121	Yes	 	Pakistan	2,476	 	Yes
Solomon Islands	2,122	Yes	 	Cuba	2,480	 	Yes
S.Tome/Principe	2,138	Yes	 	Swaziland	2,483	 	 
Congo Rep.	2,143	 	 	Vietnam	2,484	 	 
Zimbabwe	2,145	 	 	Maldives	2,485	Yes	 
Bolivia	2,174	 	 	Benin	2,487	Yes	 
Namibia	2,183	 	 	Seychelles	2,487	 	 
Botswana	2,183	 	Yes	India	2,496	 	 
Notes: (a) Amount available for human consumption. Per capita supply represents the average supply available for the population as a whole and does not necessarily indicate what is actually consumed by individuals. Source: UNDP, Human Development Report 2000: Table 23. (b) According to the FAO NFIDC group definition [http://apps.fao.org/lim500/showareas.pl?area=401&ItemType=Trade. CropsLivestockProducts&Language=].

Indicators of agricultural dependence

To identify those states in which the food insecurity may be agriculture related, Table 5 brings in information on agricultural value added as a share of GDP. It shows the share of agriculture in GDP for all of the states with a per capita calorie supply of under 2,500 (excluding six for which data are unavailable) and also any other developing country where agriculture accounts for more than 20% of GDP. Those countries in the table in which average calorie supply exceeds 2,500 are indicated by shaded lines. Once again, an indication is also given of the status of each country as LDC or NFIDC.

The LDC and NFIDC categories appear to overlap only partially with these other criteria of vulnerability. No fewer than 33 of the 76 countries in the table are neither LDC nor NFIDC, and 19 of these have an agricultural sector that accounts for over 20% of GDP. The states that would be excluded if only the criteria of calorie supply or LDC membership were taken into account include Albania, Côte d’Ivoire, Guyana, Nigeria, Moldova, Georgia, Paraguay and Dominica (using 20% as the agricultural cut-off).

Table 5. Agricultural dependence and low-calorie status

Country	Daily per capita calorie supply a 1997	Agric. value added share of GDP c 1998	LDC	NFIDCb	Country	Daily per capita calorie supply a 1997	Agric. value added share of GDP c 1998	LDC	NFIDCb
Guinea-Bissau	2,430	62.4%	Yes	 	Georgia	2,614	26.0%	 	 
Albania	2,961	54.4%	 	 	Côte d'Ivoire	2,610	26.0%	 	Yes
Burundi	1,685	54.2%	Yes	 	Vietnam	2,484	25.8%	 	 
Myanmar	2,862	53.2%	Yes	 	Paraguay	2,566	24.9%	 	 
Lao PDR	2,108	52.6%	Yes	 	Mauritania	2,622	24.8%	Yes	 
Central African Rep.	2,016	52.6%	Yes	 	Turkmenistan	2,306	24.6%	 	 
Cambodia	2,048	50.6%	Yes	 	Papua New Guinea	2,224	24.4%	 	 
Ethiopia	1,858	49.8%	Yes	 	Guatemala	2,339	23.3%	 	 
Rwanda	2,056	47.4%	Yes	 	Guinea	2,231	22.4%	Yes	 
Mali	2,029	47.0%	Yes	 	Bangladesh	2,085	22.2%	Yes	 
Kyrgyzstan	2,447	46.0%	 	 	S.Tome/Principe	2,138	21.3%	Yes	 
Tanzania	1,995	45.7%	Yes	 	Sri Lanka	2,302	21.1%	 	Yes
Uganda	2,085	44.6%	Yes	 	Azerbaijan	2,236	20.3%	 	 
Sierra Leone	2,035	44.2%	Yes	 	Honduras	2,403	20.3%	 	Yes
Cameroon	2,111	42.4%	 	 	Dominica	3,059	20.3%	 	 
Togo	2,469	42.1%	Yes	 	Zimbabwe	2,145	19.5%	 	 
Niger	2,097	41.4%	Yes	 	Yemen	2,051	17.6%	Yes	 
Nepal	2,366	40.5%	Yes	 	Senegal	2,418	17.4%	Yes	 
Chad	2,032	39.8%	Yes	 	Zambia	1,970	17.3%	Yes	 
Sudan	2,395	39.3%	Yes	 	Philippines	2,366	16.9%	 	 
Comoros	1,858	38.7%	Yes	 	Maldives	2,485	16.4%	Yes	 
Benin	2,487	38.6%	Yes	 	Swaziland	2,483	16.0%	 	 
Malawi	2,043	35.9%	Yes	 	Bolivia	2,174	15.4%	 	 
Guyana	2,530	34.7%	 	 	Angola	1,903	12.3%	Yes	 
Mozambique	1,832	34.3%	Yes	 	Dominican Rep.	2,288	11.7%	 	 
Nicaragua	2,186	34.1%	 	 	Congo Rep.	2,143	11.5%	 	 
Burkina Faso	2,121	33.3%	Yes	 	Lesotho	2,243	11.5%	Yes	 
Armenia	2,371	32.9%	 	 	Thailand	2,360	11.2%	 	 
Mongolia	1,917	32.8%	 	 	St Vincent	2,472	10.9%	 	 
Nigeria	2,735	31.7%	 	 	Namibia	2,183	10.0%	 	 
Uzbekistan	2,433	31.2%	 	 	Croatia	2,445	8.9%	 	 
Madagascar	2,021	30.6%	Yes	 	Panama	2,430	7.9%	 	 
Haiti	1,869	30.4%	Yes	 	Peru	2,302	7.1%	 	Yes
India	2,496	29.3%	 	 	Tajikistan	2,001	5.7%	 	 
Moldova	2,567	28.9%	 	 	Venezuela	2,321	5.0%	 	Yes
Gambia	2,350	27.4%	Yes	 	Seychelles	2,487	4.1%	 	 
Pakistan	2,476	26.4%	 	Yes	Antigua/Barbuda	2,365	4.0%	 	 
Kenya	1,976	26.1%	 	Yes	Botswana	2,183	3.6%	 	Yes
Notes: (a) Amount available for human consumption. Per capita supply represents the average supply available for the population as a whole and does not necessarily indicate what is actually consumed by individuals. Source: UNDP, Human Development Report 2000: Table 23. (b) According to the FAO NFIDC group definition [http://apps.fao.org/lim500/showareas.pl?area=401&ItemType=Trade. CropsLivestockProducts&Language=]. (c) Source: World Bank, World Development Indicators database website.

Calorie availability and size of agricultural sector have been taken as indicators of countries that might have objective concerns about lowering their agricultural import barriers and accepting lower ceilings for domestic subsidies. What legitimate concerns might other WTO members hold over requests from these states for SDT on market access and domestic subsidies? One set would include questioning of the developmental appropriateness of import controls and/or domestic subsidies by these states. But, as indicated above, this paper takes as paramount the WTO’s rule-making rather than its liberalising mandate. With this self-imposed limitation, the principal legitimate concern of other WTO members is whether agreeing to SDT for this group of countries would destabilise world markets.

How ‘dangerous’ would it be to the international trade system if such countries were relieved of obligations in respect of import controls and domestic subsidies for agriculture? How likely is it that such relief would result in disruption to world trade? Two indicators are provided in Table 6. This takes the countries listed in Table 5¹ and shows for each the share of agricultural exports in GDP and the country’s share of world agricultural trade. These indicators are used on the assumption that the principal ‘danger’ for other WTO members is that, sheltering behind high import barriers and benefiting from substantial subsidies, some of these states might boost substantially their agricultural exports, in competition with those of other WTO members. (Arguably, another concern is that SDT will result in lower imports by these states and, hence, lower exports by other WTO members. But, given that all the countries covered by Table 5 are ones with low calorie availability, it can be inferred reasonably that any effect on global demand will be minimal.)

The countries in Table 6 are listed in declining order of their agricultural exports as a share of world exports. Only two countries – Thailand and India – account for over 1% of world agricultural exports, and only seven account for over 0.25%. Of these, only one (Côte d’Ivoire) has a per capita calorie supply exceeding the 2,500 threshold (although a further three – India, Vietnam and Pakistan – come close).

Table 6. Trade share of vulnerable states

Country	Agric. exports as share of GDP 1999	Agric. exports as share of world agric. exports 2000	Country	Agric. exports as share of GDP 1999	Agric. exports as share of world agric. exports 2000
Thailand	5.80%	1.76%	Tajikistan	n/a	0.03%
India	1.00%	1.20%	Togo	8.30%	0.03%
Vietnam	8.50%	0.53%	Turkmenistan	n/a	0.03%
Côte d'Ivoire	21.20%	0.46%	Albania	n/a	0.02%
Guatemala	8.20%	0.38%	Azerbaijan	2.00%	0.02%
Philippines	1.80%	0.37%	Botswana	1.70%	0.02%
Pakistan	2.00%	0.26%	Chad	5.70%	0.02%
Kenya	9.70%	0.25%	Georgia	1.30%	0.02%
Uzbekistan	n/a	0.24%	Guinea-Bissau	23.30%	0.02%
Sri Lanka	n/a	0.23%	Madagascar	2.20%	0.02%
Zimbabwe	14.50%	0.20%	Mongolia	10.10%	0.02%
Peru	1.30%	0.17%	Nepal	1.60%	0.02%
Cuba	n/a	0.16%	Yemen	0.90%	0.02%
Paraguay	7.50%	0.15%	Zambia	1.90%	0.02%
Honduras	8.30%	0.14%	Bahamas	n/a	0.01%
Cameroon	5.20%	0.12%	Burundi	7.70%	0.01%
Nigeria	1.30%	0.12%	Cambodia	1.50%	0.01%
Uganda	6.90%	0.11%	Central African Rep.	2.70%	0.01%
Bolivia	4.40%	0.10%	Congo Dem. Rep.	n/a	0.01%
Ethiopia	6.30%	0.10%	Djibouti	n/a	0.01%
Papua New Guinea	12.00%	0.10%	Dominica	n/a	0.01%
Croatia	n/a	0.09%	Gambia	9.30%	0.01%
Malawi	20.00%	0.09%	Guinea	1.00%	0.01%
Sudan	n/a	0.09%	Haiti	n/a	0.01%
Dominican Rep.	1.90%	0.08%	Lao PDR	2.20%	0.01%
Nicaragua	13.60%	0.08%	Mauritania	3.60%	0.01%
Panama	3.20%	0.08%	Mozambique	0.80%	0.01%
Swaziland	28.50%	0.08%	Rwanda	2.40%	0.01%
Tanzania	3.50%	0.08%	Solomon Islands	n/a	0.01%
Moldova	27.70%	0.07%	St Vincent	n/a	0.01%
Venezuela	0.40%	0.07%	Congo Rep.	0.80%	0.01%
Mali	9.00%	0.06%	Armenia	1.00%	0.004%
Myanmar	n/a	0.06%	Comoros	3.40%	0.002%
Guyana	32.00%	0.05%	Lesotho	0.80%	0.002%
Kyrgyzstan	n/a	0.04%	Sierra Leone	1.20%	0.002%
Senegal	2.30%	0.04%	Angola	0.10%	0.001%
Bangladesh	0.30%	0.03%	Eritrea	0.50%	0.001%
Benin	5.70%	0.03%	S.Tome/Principe	10.60%	0.001%
Burkina Faso	5.10%	0.03%	Seychelles	0.30%	0.001%
Namibia	4.00%	0.03%	Antigua/Barbuda	n/a	0.0001%
Niger	5.50%	0.03%	Maldives	n/a	0.00002%
Sources: World Bank, World Development Indicators database website (GDP); FAO Statistical Databases website (agricultural export values).

It goes beyond the scope of this initial essay to assess whether or not WTO members would consider countries supplying such low shares of world exports to be a ‘threat’ and, if so, whether one could identify additional parameters for SDT (such as limiting it in such cases to staple foods) that would overcome the problem. Suffice it to note that a combination of the following criteria would include a larger number of low-calorie-supply countries (66) than does the LDC criterion alone (37):

♦ daily per capita calorie supply of under 2,500; and/or

♦ agriculture accounting for over 20% of GDP; and

0.25% or less of world agricultural exports.

Import dependency

It was suggested in Table 2 that poor countries, dependent upon agricultural imports and with weak trade entitlements, might legitimately be concerned about industrialised country actions that would tend to increase the price, or otherwise reduce the availability, of food imports. The appropriate SDT in such cases would not be relief from tougher rules governing their own trade and production policies, but compensatory action (either within the WTO or by prior agreement via other institutions) to help them to adjust to such change.

One key issue for the Doha Round is how such compensating action is to be enforced; this falls outside the scope of the present study. Another is whether a better categorisation of such countries can be obtained. The aim of the LDC and NFIDC categories combined is to identify such countries, but one may question how well this is achieved.

Table 7 provides an illustration of the inadequacy of the existing categories. The table shows calorie availability, the vulnerability index, agricultural imports as a share of GDP, and LDC/NFIDC status for:

♦ all the states with per capita calorie supply of under 2,500 (except 14 for which vulnerability data are not available) and

any state with greater vulnerability (higher index number) than the mean for states registering under 2,500 calories (shaded in the table).

The states are presented in declining order of agricultural imports as a share of GDP.

Table 7. Trade vulnerability and low-calorie status

Country	Daily per capita calorie supply a 1997	Composite vulnerability index b	Agric. imports as share of GDP c 1999	LDC	NFIDC
Gambia	2,350	9.331	27.87%	Yes	 
S.Tome/Principe	2,138	7.69	19.98%	Yes	 
Sierra Leone	2,035	5.06	19.51%	Yes	 
Mauritania	2,622	6.068	19.40%	Yes	 
Lesotho	2,243	5.985	18.68%	Yes	 
Swaziland	2,483	9.633	17.76%	 	 
Comoros	1,858	5.425	14.26%	Yes	 
Nicaragua	2,186	4.92	13.50%	 	 
Yemen	2,051	5.259	11.46%	Yes	 
St Lucia	2,734	7.449	10.68%	 	 
Senegal	2,418	5.026	10.40%	Yes	 
Seychelles	2,487	6.375	9.01%	 	 
Guyana	2,530	7.953	8.31%	 	 
Honduras	2,403	5.373	8.07%	 	Yes
Mauritius	2,917	6.51	7.37%	 	Yes
St Kitts/Nevis	2,771	6.362	7.01%	 	 
Angola	1,903	6.282	6.80%	Yes	 
Niger	2,097	4.957	6.70%	Yes	 
Fiji	2,865	8.888	6.55%	 	 
Jamaica	2,553	7.484	6.46%	 	Yes
Belize	2,907	6.652	6.39%	 	 
Botswana	2,183	10.158	6.15%	 	Yes
Benin	2,487	5.06	6.00%	Yes	 
Papua New Guinea	2,224	6.308	5.93%	 	 
Togo	2,469	5.248	5.47%	Yes	 
Nepal	2,366	5.173	5.20%	Yes	 
Malaysia	2,977	5.903	5.00%	 	 
Burkina Faso	2,121	4.923	4.65%	Yes	 
Mozambique	1,832	4.907	4.55%	Yes	 
Bangladesh	2,085	4.744	4.51%	Yes	 
Congo Rep.	2,143	5.961	4.49%	 	 
Panama	2,430	4.995	4.16%	 	 
Pakistan	2,476	4.795	3.95%	 	Yes
Uganda	2,085	4.876	3.77%	Yes	 
Guinea	2,231	5.282	3.71%	Yes	 
Rwanda	2,056	4.797	3.63%	Yes	 
Philippines	2,366	4.595	3.51%	 	 
Mali	2,029	5.083	3.48%	Yes	 
Central African Rep.	2,016	4.802	3.37%	Yes	 
Guatemala	2,339	4.431	3.33%	 	 
Namibia	2,183	6.527	3.31%	 	 
Dominican Rep.	2,288	4.858	3.21%	 	 
Kenya	1,976	4.935	3.12%	 	Yes
Zimbabwe	2,145	4.969	2.97%	 	 
Tanzania	1,995	5.035	2.96%	Yes	 
Zambia	1,970	5.549	2.80%	Yes	 
Ethiopia	1,858	4.786	2.55%	Yes	 
Malawi	2,043	5.2	2.44%	Yes	 
Bolivia	2,174	4.691	2.37%	 	 
Chad	2,032	5.12	2.24%	Yes	 
Burundi	1,685	4.929	2.11%	Yes	 
Madagascar	2,021	4.785	1.90%	Yes	 
Peru	2,302	4.461	1.88%	 	Yes
Thailand	2,360	4.264	1.87%	 	 
Cameroon	2,111	4.952	1.60%	 	 
Venezuela	2,321	4.887	1.37%	 	Yes
India	2,496	3.782	0.86%	 	 
Antigua/Barbuda	2,365	11.246	n/a	 	 
Bahamas	2,443	10.433	n/a	 	 
Congo Dem. Rep.	1,755	5.186	n/a	Yes	 
Djibouti	2,084	7.932	n/a	Yes	 
Dominica	3,059	8.122	n/a	 	 
Gabon	2,556	6.229	n/a	 	 
Grenada	2,768	7.848	n/a	 	 
Haiti	1,869	4.474	n/a	Yes	 
Maldives	2,485	8.654	n/a	Yes	 
Solomon Islands	2,122	8.398	n/a	Yes	 
Sri Lanka	2,302	5.076	n/a	 	Yes
St Vincent/Grenadines	2,472	6.563	n/a	 	 
Sudan	2,395	4.655	n/a	Yes	 
Vanuatu	2,700	13.295	n/a	Yes	 
Notes: (a) Amount available for human consumption. Per capita supply represents the average supply available for the population as a whole and does not necessarily indicate what is actually consumed by individuals. Source: UNDP, Human Development Report 2000: Table 23. (b) Commonwealth Secretariat/Word Bank 1999. (c) World Bank, World Development Indicators database website (GDP); FAO Statistical Databases website (agricultural import values).

The table suggests that the LDC and NFIDC categories miss some states that ought to be included. For example, if one takes agricultural imports of over 5% of GDP as a threshold, there are three states that do not fall into either group and for which agricultural imports exceed this level and calorie supply is below 2,500 per capita and trade vulnerability is high. This number increases to six if the states at the foot of the table (for which import and/or GDP data are not available) are assessed on the other two criteria.

At the same time, 11 of the 20 NFIDCs and 5 LDCs² do not appear in the table at all, indicating that they have over 2,500 calories per capita plus either a lower trade vulnerability and/or agricultural import dependence. Since the source of most compensatory assistance will probably be the main aid donors (regardless of the institutional route through which it is provided), and given that they recognise the LDC group as one for special aid attention, the absence of some LDCs from Table 7 is probably not serious. If one identified as the countries requiring special concern all LDCs plus non-LDCs that have low calorie availability, high trade vulnerability, and significant proportionate food imports, then coverage would be reasonably good.

Next Steps

The principal focus of action for food insecure states is to remove the causes of the insecurity, i.e. to address low domestic production and weak exports. Many of the more important measures fall well outside the ambit of the WTO. But there is none the less scope for ample action within the WTO.

The minimum requirement is that nothing agreed in the AoA should make matters worse. This includes avoiding rule changes that might make it even more difficult than is currently the case for food insecure states to pursue policies that have proved to be useful in supporting food security in some states. These are described in Table 8, which identifies the area of WTO rules that might constrain government action, the AoA provisions available to developing countries and most widely used by them, the types of change that might be proposed in the current negotiations, and some potential alternative instruments.

Table 8. Provisions relating to food security in the AoA

Food security policies	Related WTO trade policy area	AoA provisions available to developing countries and widely used	Types of change that could affect food security	Potential alternative instruments
Input credits and subsidies; capital expenditure in agriculture; food price stabilisation and subsidies	Domestic subsidies	SDT exemptions from cuts in agricultural investment and input subsidies for poor farmers 10% de minimis exemption LLDCs exempt from any cuts	Erosion of SDT provisions on investment and input subsidies Cut in de minimis provisions	Recalculation of AMS Redesign of Green Box
All policies involving government expenditure; export development; protection to domestic farmers	Tariffs	High bound rates containing ‘water’ Lower (for developing) and zero (for least developed) tariff cutting obligations	Removal of ‘water’ from tariffs before end of developed country subsidised production	Provision of SSGs Countervailing duties
Labour-intensive public works and targeted feeding programmes; food stamps	Export subsidies	Food aid exempt from developed country reduction commitments	Reduction of subsidised imports available to vulnerable countries and groups that is more rapid than feasible adjustment	Targeting of concessional food (possibly outside WTO framework)
Source: adapted from Stevens et al. 2000.

But action can go further than this. For example, many WTO rule changes could have an effect on export diversification by food insecure states – and this effect will not always be favourable. Such potential effects need to be identified, and the right recognised of food insecure states to oppose them unless appropriate, actionable supporting measures are agreed.

Lobbying for such positive treatment is likely to cause friction if it is done on behalf of states that have a significant effect on world trade and are not especially food insecure (even though they may import a lot of food). There is a major technical role to play, therefore, to produce objective indicators of relative national food insecurity that would form a basis for group formation.

Appendix 1. Daily per capita supply of calories, 1997

Rank	Country a	Daily per capita supply of calories b 1997		LDC		NFIDC c
1	Eritrea	1,622		Yes		 
2	Burundi	1,685		Yes		 
3	Congo, Dem. Rep.	1,755		Yes		 
4	Mozambique	1,832		Yes		 
5	Comoros	1,858		Yes		 
6	Ethiopia	1,858		Yes		 
7	Haiti	1,869		Yes		 
8	Angola	1,903		Yes		 
9	Mongolia	1,917		 		 
10	Zambia	1,970		Yes		 
11	Kenya	1,976		 		Yes
12	Tanzania, U. Rep. of	1,995		Yes		 
13	Tajikistan	2,001		 		 
14	Central African Republic	2,016		Yes		 
15	Madagascar	2,021		Yes		 
16	Mali	2,029		Yes		 
17	Chad	2,032		Yes		 
18	Sierra Leone	2,035		Yes		 
19	Malawi	2,043		Yes		 
20	Cambodia	2,048		Yes		 
21	Yemen	2,051		Yes		 
22	Rwanda	2,056		Yes		 
23	Djibouti	2,084		Yes		 
24	Bangladesh	2,085		Yes		 
25	Uganda	2,085		Yes		 
26	Niger	2,097		Yes		 
27	Lao People's Dem. Rep.	2,108		Yes		 
28	Cameroon	2,111		 		 
29	Burkina Faso	2,121		Yes		 
30	Solomon Islands	2,122		Yes		 
31	São Tomé and Principe	2,138		Yes		 
32	Congo	2,143		 		 
33	Zimbabwe	2,145		 		 
34	Bolivia	2,174		 		 
35	Botswana	2,183		 		Yes
36	Namibia	2,183		 		 
37	Nicaragua	2,186		 		 
38	Papua New Guinea	2,224		 		 
39	Guinea	2,231		Yes		 
40	Azerbaijan	2,236		 		 
41	Lesotho	2,243		Yes		 
42	Dominican Republic	2,288		 		Yes
43	Peru	2,302		 		Yes
44	Sri Lanka	2,302		 		Yes
45	Turkmenistan	2,306		 		 
46	Venezuela	2,321		 		Yes
47	Guatemala	2,339		 		 
48	Gambia	2,350		Yes		 
49	Thailand	2,360		 		 
50	Antigua and Barbuda	2,365		 		 
51	Philippines	2,366		 		 
52	Nepal	2,366		Yes		 
53	Armenia	2,371		 		 
54	Sudan	2,395		Yes		 
55	Honduras	2,403		 		Yes
56	Senegal	2,418		Yes		 
57	Guinea-Bissau	2,430		Yes		 
58	Panama	2,430		 		 
59	Uzbekistan	2,433		 		 
60	Bahamas	2,443		 		 
61	Croatia	2,445		 		 
62	Kyrgyzstan	2,447		 		 
63	Togo	2,469		Yes		 
64	Saint Vincent and the Grenadines	2,472		 		Yes
65	Pakistan	2,476		 		Yes
66	Cuba	2,480		 		Yes
67	Swaziland	2,483		 		 
68	Viet Nam	2,484		 		 
69	Maldives	2,485		Yes		 
70	Benin	2,487		Yes		 
71	Seychelles	2,487		 		 
72	India	2,496		 		 
73	Guyana	2,530		 		 
74	Jamaica	2,553		 		Yes
75	Gabon	2,556		 		 
76	El Salvador	2,562		 		 
77	Paraguay	2,566		 		 
78	Moldova, Rep. of	2,567		 		 
79	Colombia	2,597		 		 
80	Côte d'Ivoire	2,610		 		Yes
81	Ghana	2,611		 		 
82	Georgia	2,614		 		 
83	Iraq	2,619		 		 
84	Mauritania	2,622		Yes		 
85	Costa Rica	2,649		 		 
86	Trinidad and Tobago	2,661		 		Yes
87	Macedonia, TFYR	2,664		 		 
88	Suriname	2,665		 		 
89	Ecuador	2,679		 		 
90	Bulgaria	2,686		 		 
91	Vanuatu	2,700		Yes		 
92	Saint Lucia	2,734		 		Yes
93	Nigeria	2,735		 		 
94	Grenada	2,768		 		 
95	Saint Kitts and Nevis	2,771		 		Yes
96	Saudi Arabia	2,783		 		 
97	Ukraine	2,795		 		 
98	Chile	2,796		 		 
99	Uruguay	2,816		 		 
100	Iran, Islamic Rep. of	2,836		 		 
101	Estonia	2,849		 		 
102	Algeria	2,853		 		 
103	Brunei Darussalam	2,857		 		 
104	Myanmar	2,862		Yes		 
105	Latvia	2,864		 		 
106	Fiji	2,865		 		 
107	Indonesia	2,886		 		 
108	China	2,897		 		 
109	Russian Federation	2,904		 		 
110	Belize	2,907		 		 
111	Mauritius	2,917		 		Yes
112	Japan	2,932		 		 
113	Albania	2,961		 		 
114	Brazil	2,974		 		 
115	Malaysia	2,977		 		 
116	Slovakia	2,984		 		 
117	South Africa	2,990		 		 
118	Jordan	3,014		 		 
119	Cape Verde	3,015		Yes		 
120	Dominica	3,059		 		 
121	Morocco	3,078		 		Yes
122	Kazakhstan	3,085		 		 
123	Argentina	3,093		 		 
124	Kuwait	3,096		 		 
125	Mexico	3,097		 		 
126	Finland	3,100		 		 
127	Slovenia	3,101		 		 
128	Iceland	3,117		 		 
129	Canada	3,119		 		 
130	Korea, Rep. of	3,155		 		 
131	Barbados	3,176		 		Yes
132	Sweden	3,194		 		 
133	Hong Kong, China (SAR)	3,206		 		 
134	Switzerland	3,223		 		 
135	Australia	3,224		 		 
136	Belarus	3,225		 		 
137	Czech Republic	3,244		 		 
138	Romania	3,253		 		 
139	Lithuania	3,261		 		 
140	United Kingdom	3,276		 		 
141	Lebanon	3,277		 		 
142	Israel	3,278		 		 
143	Tunisia	3,283		 		Yes
144	Netherlands	3,284		 		 
145	Egypt	3,287		 		Yes
Notes (a) In ascending order of number of calories available for human consumption. Table includes all LDCs (apart from those for which data are not available – Afghanistan, Bhutan, Equatorial Guinea, Kiribati, Liberia, Samoa, Somalia and Tuvalu) and all NFIDCs, and all other countries lower ranked than the highest of these. (b) Amount available for human consumption. Per capita supply represents the average supply available for the population as a whole and does not necessarily indicate what is actually consumed by individuals. Source: UNDP 2000: Table 23. (c) According to the FAO NFIDC group definition [http://apps.fao.org/lim500/showareas.pl?area=401&ItemType=Trade. CropsLivestockProducts&Language=].

References

Commonwealth Secretariat/World Bank 1999. ‘Small States: A Composite Vulnerability Index. First Draft’. Paper prepared by the Joint Commonwealth Secretariat/World Bank Task Force for the ‘Conference on the Small States’, St Lucia, West Indies, February 17–19 1999.

Diaz-Bonilla, E.; Thomas, M. and Robinson, S. 2000. ‘Food Security and Trade Negotiations in the World Trade Organization: A Cluster Analysis of Country Groups’, TMD Discussion Paper No. 59. Washington DC: Trade and Macroeconomics Division, International Food Policy Research Institute (December).

Drèze, J. and Sen, A. (eds) 1990. The Political Economy of Hunger. Oxford: Clarendon Press.

FAO Statistical Databases (http://apps.fao.org/default.htm).

FAO 1996. State of Food and Agriculture 1996 (FAOSTAT TS software by the United States Department of Agriculture). Rome: Food and Agriculture Organization of the United Nations.

Finger, J.M. and Schuler, P. 2000. ‘Implementation of Uruguay Round Commitments’, The World Economy 23(4):511–25.

Fukasaku, K. 2000. ‘Special and Differential Treatment for Developing Countries: Does It Help Those Who Help Themselves?’, Working Paper No. 197. Helsinki: UNU World Institute for Development Economics Research.

Henson, S. and Loader, R. 2000. ‘Barriers to Agricultural Exports from Developing Countries: The Role of Sanitary and Phytosanitary Requirements’, World Development. 29(1):85–102.

Michalopoulos, C. 2000. ‘The Role of Special and Differential Treatment for Developing Countries in GATT and the World Trade Organization’, Policy Research Working Paper 2388. Washington DC: The World Bank, Development Research Group, Trade.

Stevens, C.; Greenhill, R.; Kennan, J. and Devereux, S. 2000. ‘The WTO Agreement on Agriculture and Food Security’, Economic Paper 42. London: Commonwealth Secretariat.

UNDP 2000. Human Development Report 2000. Geneva: United Nations Development Programme.

Whalley, J. 1999. ‘Special and Differential Treatment in the Millennium Round’, Working Paper No. 30/99. Coventry: University of Warwick, Centre for the Study of Globalisation and Regionalisation.

World Bank. World Development Indicators database (http://www.worldbank.org/data/ dataquery.html)

WTO 1995. The Results of the Uruguay Round of Multilateral Trade Negotiations: The Legal Texts. Geneva: the World Trade Organization.

WTO 2000. ‘Implementation of special and differential treatment provisions in WTO Agreements and Decisions: Note by the Secretariat’, WT/COMTD/W/77. Geneva: World Trade Organization, Committee on Trade and Development (25 October).

¹ Plus the six countries excluded from Table 5 because of lack of the relevant data.

² Plus nine for which calorie supply data are not available.