Annex I - Methodology

Contents - Previous

Introduction .To undertake the assessment of the impact of the agreements reached in the Uruguay Round of Multilateral Trade Negotiations, the existing projections to the year 2000 for the main agricultural commodities prepared in 1992/93¹ have been adjusted to take into account both the commitments to reduce tariffs and export subsidies and to introduce minimum levels agreed in the Final Act of the Uruguay Round as well as the effects on income as estimated by the OECD/World Bank due to the impact of the Round as a whole.

The methodology used in the previous round of projections - the "baseline" projections in this publication is described in the aforementioned FAO study. In order to accomplish this task the FAO World Food Model (WFM), used in the last round of projections for the cereal/livestock/fats and oil complex, has been extended and improved to better reflect the impacts of specific policy changes while simple single commodity models have been developed to cover some of the commodities not included in the WFM.

The "central" or "baseline" scenario based on the assumptions regarding the economic and demographic conditions expected to prevail by 2000, the most likely development of technology, no change in agricultural policies (including however known policy changes at the time of writing, e.g. including reform of the EC's CAP) and normal weather conditions - has been used as a point of reference to evaluate the impact of the policy changes reflected in the "Uruguay Round scenario".

A number of additional policy scenarios were simulated. These include: a) a "higher income growth scenario" reflecting a more optimistic estimate of the impact on world income deriving from the Uruguay Round; b) a "crop failure scenario" simulating the influence of bad weather on the output of cereals during the projection period; c) a "bumper crop scenario" following exceptionally favourable weather conditions affecting the output of cereals.

Assumptions .The projections made under the alternative scenarios rest on a set of assumptions on population and income growth. The study uses a single population growth assumption for each country, based on demographic projections prepared by the United Nations Population Division, corresponding to the UN "medium variant" calculated in 1990.² Detailed information on the methods and assumptions used in making the projections can be found in the documentation of the UN Population Division (op.cit.).

As regards income growth, the baseline scenario uses a single growth rate of Gross Domestic Product (GDP) for each country. The GDP growth assumptions for the 1990s are mainly based on the long-term economic forecasts prepared by the International Economics Department of the World Bank. These were supplemented, for countries where such forecasts are not available, with estimates derived from other sources.

Estimates of net income gains resulting from the Uruguay Round has been made by various authors. According to the GATT Secretariat, estimates of the increase in world income from the liberalization of trade in goods range from a low of US$ 109 billion to a high of US$5 10 billion in 2005 depending on the assumptions used³. These estimates are based on a general equilibrium model of the world economy, elaborated and applied by the GATT Secretariat. Three versions of the model have been used, with different assumptions about the nature of competition in domestic markets, economy of scale, the degree of product differentiation and the extent to which income gains in turn stimulate savings and investment. The World Bank/OECD has estimated gains of around US$213 billion in 2002⁴. For the purpose of the present study the World Bank/OECD estimates were retained, with appropriate adjustments, for the main "Uruguay Round scenario" while double this amount was taken for tile higher income growth scenario. Tables A1 to A3 summarize the assumptions of population, total and per caput GDP which were used in the three scenarios covered by tit is study, namely tile "baseline" the "Uruguay Round central scenario" and "Uruguay Round higher income growth scenario".

Table A1 - Total Population, 1980 to 1990 Actual and 1990 to 2000- Projected

Source: UN World Demographic Projections as Assessed in 1990

Note: For country derails see: FAO "Compendiums of Demographic and Macro-Economic Indicators", ESC/M/93/2/Rome 1993.

Table A2 - Total GDP at 1980 Prices, 1980 to 1990 Actual and 1990 to 2000 Projected

Table A3 - Per Caput GDP at 1980 Prices, 1980 to 1990 Actual and 1990 to 2000 Projected

Simulating the impact of tire Uruguay Round

The present assessment relates primarily to the market and policy changes likely to arise up to the year 2000 when the bulk of the Uruguay Round commitments will have been completed. It should be clearly understood, however, that the study covers only those measurable policy changes in trade and agricultural policies explicitly defined in the GATT commitments and likely to occur in practice over the implementation period.

The simulation runs cover principally the effect of the Uruguay Round Agreement for the agricultural sector related to the tariffication of trade distortion measurements and of the agreed reductions. The country specific commitments concerning cuts in subsidized exports and increases in imports under the minimum access provisions are also included in the simulation runs.

The main source of data on tariff reductions and bindings is the Schedules of Market Access Concessions which cover 89 participant countries (the 12 Member States of the European Union coupling as one). Agricultural product categories are defined in terms of the six-digit Harmonized System code. Tariff lines for individual commodities are provided for the "Base Rate of Duty" relating to the year 1986-88 and for the "Bound Rate of Duty" relating to the final year of implementation (generally the year 2000 for the developed and transition economies and 2004 for the developing countries), expressed in "specific" and "ad-valorem" tariffs.

For the purpose of the simulation, all tariffs, including those for derived products, were converted and aggregated into a weighed average primary product equivalent, the weights being the total domestic availability of the products concerned in each country.

The "specific" tariffs, generally expressed in national currencies, were converted into US dollars and divided by the respective producer prices in order to include them into the WFM normalized price transmission equations. Annex tables 1-14 contain the summary aggregated tariffs for the basic foodstuffs for all countries for which data were available.

Some observations are in order for a proper use of the Schedules. The Uruguay Round on Agriculture requires the conversion of all non-tariff barriers into ordinary customs duties, and the binding of the resulting tariffs. All tariff equivalents created by the tariffication process, are bound and all tariffs including pre-existing ones are to be reduced, on average, 36 percent over the period 1995 to 2000, with a minimum reduction of 15 percent per tariff line⁵. For developing countries the figures were 24 and 10 percent respectively.

Tariffication of non-tariff barriers to agricultural trade is one of the most significant achievement of the Uruguay Round⁶. The resulting tariff equivalents are transparent and, moreover, they provide a common basis for reductions in future negotiations. However, the particular way in which the conversion of non-tariff barriers into tariffs has been achieved in the Agreement has sometimes been called "dirty" tariffication, because the tariffs are often very high when compared to the actual level of protection as measured by the PSEs. The reason for this is that the years 1986-88 were chosen as the base years for tariffication when world prices for many agricultural commodities were the lowest in decades. This means that, when the world price in the base period is compared to the protected and supported internal prices, the gap was unusually wide. In addition many countries used prices to calculate their tariff equivalents which resulted in higher initial tariffs than more objective calculations might have produced. Many developing countries opted to offer ceiling bindings to their tariffs instead of tariffying. Finally, the simple (unweighted) averaging that country could use when allocating tariff reduction to individual products has allowed much scope for continued protection of the "sensitive" products.

In practice, it is possible that in some cases countries will not apply the full tariff recorded in the Schedules, so that the new tariffs can be interpreted as representing maxima.

Despite the above reservations, the simulation of the "Uruguay Round" scenario assumes that countries will undertake reforms according to their offers on the reductions in tariffs, as presented in the Schedules of reduction commitments. In addition, the simulation assumes that developing countries will not raise their tariffs, even though in many cases the new tariff bindings are above the previously observed rates, implying a presumption that countries will not exercise the option to increase protection as a result of the Agreement.

The minimum access provisions have been introduced into the WFM on an "ad-hoc" basis in all cases where the model did not generate a sufficient volume of imports to meet national commitments. The additional import quantities have been modelled by decreasing the production and/or increasing the demand, depending on the particular circumstances. Figures on the current and minimum access commitments by commodity and country are provided in Table A4.

Regarding export subsidy commitments (Table A5), the approach followed in the WFM differs according to the type of export subsidies used in the past. In particular, for countries that have used export subsidies for all their exports, a maximum has been introduced exogenously on the volume of their exports according to the commitments for specific commodity and year. In order to accommodate this reduction, it was necessary to increase the demand, i.e in the feed sector, or to adjust the production by lowering yields or restricting the projected cultivated area, taking into account current and expected country adjustment policies and plans. For those exporters which only subsidize a part of their exports targeted to some countries only, no such constraint has been modelled but it is still assumed that this will erode part of their competitiveness and hence influence the volume of their exports. Net exporters subsidizing their exports create an "artificial" price incentive on the internal marker and contribute to depress world market prices especially when the quantities exported are drawn from government stocks. The approach followed in the WFM to simulate the impact of reduced targeted export subsidies was to include an additional clement in the supply and demand price transmission equations reflecting the price reduction expected to prevail in the domestic market of the subsidizing exporters and the corresponding increase in the price of the importing countries.

Table A4 - Market access under minimum access opportunity commitments by commodity1/

1/ As products are expressed at different stages of processing in the schedules, the totals are only indicative.
2/ The implementation period generally begins in 1995 and ends in 2000 for the developed countries and in 2004 for the developing countries and for the economies in transition.
3/ Including other fats

Table A5 - Subsidized export reduction commitments by commodity

Alternative Scenarios

Evaluating the impact of policy changes is done by comparing the results of the simulation run which include these policy alterations with those of the base run. The latter provides a neutral point of comparison for the assessment and, hence, depicts the evolution of agricultural commodity markers, given past developments of all major determinants including policies.

No feedbacks effects due to liberalization are included in the "Uruguay Round" scenario with the exception of its likely impact on income growth. The income assumptions used in the "baseline" scenario have been modified to include the income gains to be expected from trade liberalization. Since the WFM covers only a limited number of agricultural products it was not possible to estimate directly the overall welfare gain relative to the level of income assumed for the "baseline" projections. The present simulations, therefore, utilize estimates derived from other models with a more comprehensive coverage. In particular, the World Bank/OECD "conservative" estimates of US$213 billion were retained for the base "Uruguay Round" scenario⁷. These were included as percentage increases in income relative to the benchmark levels for individual countries and regions. In addition, in order to see the effect of an acceleration of income growth on the pattern and value of world agricultural trade, a Higher Income Growth scenario was generated doubling the percentage increases in income gains.

One of the most important benefits likely to arise from the Uruguay Round Agreement is the expected reduction in price instability. The idea behind this view is that there will be increasing the number of countries open to world market forces which could eventually absorb shocks generated by production fluctuations, such as those due to unexpected production shortfalls or bumper crops as a consequence, for example, of large weather fluctuations. Since the WFM is basically a determinist model, i.e. does not contain stochastic elements, the approach followed was essentially to examine the impact of production shocks on world price stability in order to verify if tariffication and reduction of tariffs had the expected effect. The production shocks chosen for examination were a generalized 5 percent decline (and symmetrically a 5 percent increase) in cereals output in the year 1999 with respect to that projected for the same year in the "baseline" scenario. Due to the lag nature of the cereal supply equation their effect are reflected on world prices in the year 2000.

Methodology for other commodities The "baseline" projections for commodities not included in the WFM, namely coffee, cocoa, tea, sugar, bananas, rubber and bovine hides, were made using methodologies ranging from detailed econometric commodity models (e.g. sugar) to demand and supply projections based on analyses and adjusted extrapolations of past treads supplemented by expert judgments of the FAO commodity specialists⁸. In a number of cases, the projections were prepared jointly or in cooperation with other international organizations.

In particular, the demand projections for tropical beverages were made using demand functions which links per caput consumption to the projected growth rates of per caput GDP by means of income elasticities estimated mainly from cross-section studies. It was assumed that real prices would remain constant during the projection period, i.e. the effect of price changes was implicitly excluded as influencing future demand. On the supply side, projections were made separately for area and yield. The area projections are mainly based on trends, which were adjusted by incorporating relevant information on policies and technological changes. An important property of most of these crops is their long production period and at the same time short term (usually biennial) yield fluctuations. Given both the long term production profile and the short term output fluctuations, the development of yields is to be considered the crucial determinant for the future evolution of production of these commodity groups. The yield projections were made by the FAO commodity specialists taking into account (i) the composition of the total area of the respective commodity according to the year of planting, e.g. the number of bearing and nonbearing trees etc.; (ii) the percentage of trees uprooted or replanted because of age, disease or damage; (iii) the average yield profile during its life; (iv) technical progress and (v) the impact of other exogenous factors influencing future production (e.g. frost in the past). Net trade was calculated as the difference between quantities supplied, defined as production minus stock changes, and quantities demanded, defined as the sum of the various form of utilization. At the country level this difference represents export availability (if positive) or import requirements (if negative).

The projections for sugar were prepared in cooperation with the International Sugar Organization using an econometric model where standard production, demand and price linkage equations were estimated using data from 1970 to 1990. The "baseline" projections were generated on the basis of a series of assumptions about income growth, agricultural policies, weather and technological changes. Agricultural policies relating to sugar production and consumption were assumed to be continued in all countries. For the production projections average weather conditions and historical rates of technological change were assumed to prevail during the projection period. However, for several countries production constraints were introduced on an ad hoc basis, details of which are given in the document⁹.

Projections of import Requirement and export availability for bananas were made on the general assumption of a continuation of present national policies. Although it was possible to make a simplifying assumption that price relations would remain unchanged during the projection period this has not been done. Unchanged price relations would lead to a large disequilibrium between supply and demand at the world level that is unlikely to occur in practice. Therefore, some change in price relations has been allowed when preparing the projections. In addition, recent trends in actual imports and exports of bananas have been given greater weight to obtain a more realistic balance between projected export availabilities and import requirements. In interpreting the results, several factors need to be considered. These include the varying degree of accuracy of historical data on which past trends are assessed and the unsettled situation in several areas that theoretically have strong prospects for future growth of consumption.

Projections of bovine hides were directly linked to the WFM projections of cattle population, using historically-based conversion factors to derive hides production from the projected number of animals slaughtered. For countries where hides and skins are produced predominantly from "fallen" animals, a time trend was added. Per caput consumption was projected for each country or group of countries using a regression equation incorporating income projections and a time trend. the resulting projections were then adjusted so that the aggregate world demand equalled total supply. Net trade was calculated as the difference between domestic production and consumption.

Simulating the impact of tariff reductions For the non-WFM commodities a "simple" price-equilibrium model was used to calculate the effects of the tariff reductions following the Uruguay Round Agreement. This is a single-commodity model which uses multiple spreadsheets. The model, therefore, does not presently allow a multi-commodity approach similar to that of the WFM, although substitution and complementary relations to other products could theoretically be included. However, in order not to add complexity to the model and to facilitate interpretation, it was considered unnecessary to expand it to add these features also because its use was limited to commodities which compete only marginally. This "simple" model only takes one primary commodity into account. All derived products from the commodity concerned are expressed in primary commodity equivalent including tariff rates following a methodology similar to that utilized in the WFM.

The model does not itself generate projections but utilizes the "baseline" scenario as a starting point for the simulation. Supply and demand respond to changes in domestic producer and retail prices, respectively, through a semi-logarithmic function and selected price elasticities of supply and demand. Domestic prices are linked to border prices (i.e import unit values and export unit values) by margins representing internal marketing, processing, distribution costs and excise taxes. Border prices are, in turn, linked to world market prices (i.e the clearing price) by price wedges reflecting both transport cost and trade policies. In a situation in which a country imports the primary commodity, the cif price is equal to the world market price plus transport costs from the world market to the country. The landed price includes the rate of import tariff, or rather the nominal rate of protection. Should the country, however, turn out, in a given model run, to be an exporter of the primary commodity, the f.o.b. price is relevant. Moreover an export tax may apply or export subsidies can be granted.

To simulate the impact of the Uruguay Round, model assumes that all other margins remains proportionally constant in relation to the base period with the exception of the tariff component which is reduced in line with the country's reduction commitment.

The equilibrium solution of the model is found through an iterative process. The first iteration starts with an arbitrary level of world market price (i.e. generally the price projected under the "baseline" scenario). Domestic prices are modified as a consequence of the reduction in the effective rate of protection. For each country new supply and demand levels are calculated and net trade position derived. If net exports do not sum up to zero, the world market price is changed fur the second iteration and this procedure is repeated until world supply and demand are in equilibrium (i.e. world net trade equal to zero or close to zero with any predetermined precision). The final solution determines the new levels of supply and demand and the market clearing price.

Annex Table 1 - Wheat - Aggregated Tariff Equivalents

Annex Table 2 - Rice - Aggregated Tariff Equivalents

Annex Table 3 - Maize - Aggregated Tariff Equivalents

Annex Table 4 - Millet and Sorghum - Aggregated Tariff Equivalents

Annex Table 5 - Coarse Grains - Aggregated Tariff Equivalents

Annex Table 6 - Fats and Oils - Aggregated Tariff Equivalents

Annex Table 7 - Butter - Aggregated Tariff Equivalents

Annex Table 8 -Oilmeals -Aggregated Tariff Equivalents

Annex Table 9 - Beef - Aggregated Tariff Equivalents

Annex Table 10 - Milk - Aggregated Tariff Equivalents

Annex Table 11 - Sheep and Goat Meat - Aggregated Tariff Equivalents

Annex Table 12 - Pigmeat - Aggregated Tariff Equivalents

Annex Table 13 -Poultry Meat Aggregated Tariff Equivalents

Annex Table 14 - Tariff reductions in selected sugar importing and exporting countries

a Countries committed to a maximum rate. b Tariff rate on above-quota imports. Source: International Sugar Organisation calculation based on WTO schedules.

REFERENCES

1. "Medium-term prospects for agricultural commodities: Projections to the year 2000", FAO Economic and Social Development Paper 120, Rome 1994. "Impact of the Uruguay Round on Agriculture", FAO CCP: 95/13, Rome, January 1995.

2. UN - Department of International Economic and Social Affairs: "World Population Prospects 1990", New York, 1991.

3. " The Results of the Uruguay Round of Multilateral Trade Negotiations", GATT, Geneva, November 1994

4. "Trade Liberalization: Global Economic Implications" by Ian Goldin, Odin Knudsen and Dominique van der Mensbrugghe, OECD and World Bank, 1993. See also "The Uruguay Round: An Assessment of Economy wide and Agricultural Reforms" by Ian Golding and Dominique van der Mensbrugghe, presented at The Uruguay Round and the Developing Economies, a World Bank Conference, January 1995.

5. For developing countries the reduction is 24 percent, on average, over the period 1995-2004, with a minimum of 5 percent per tariff line.

6. See Tangermann S. "An Assessment of the Uruguay Round Agreement on Agriculture" paper prepared for the Directorate for Food, Agriculture and Fisheries and the Trade Directorate of OECD, Stanford, 1994.

7. According to the World Bank/OECD simulations, the most important source of income gains from the liberalization of trade in goods is the elimination of quotas on industrial products, particularly MFA quotas, and the reduction in industrial tariffs. However the different estimates of income gains produced by the World Bank and the OECD depend on the version of the model used, i.e. with static or dynamic specifications.

8. The FAO "baseline" projections to the year 2000 extend the commodity coverage to other agricultural products, such as roots and tubers, pulses, citrus fruit and textile fibres. The Secretariat is considering the possibility ( resource permitting ) of conducting further studies of the impact of the Uruguay Round on selected agricultural commodities not included in the present study.

9. FAO-ISO "The World Sugar Market: Prospects for the Nineties", Rome 1992.

Contents - Previous