PART II: AGRICULTURAL POLICY AND AGRICULTURAL MARKETS
2 Theoretical foundations of the model
EPACIS1 is a static partial equilibrium model for the agricultural sector of the CIS. A similar model has been applied to estimate the effects of EU Eastern enlargement on EU households (W E-BER, WAHL & MEINLSCHMIDT, 2000). This model includes the supply of agricultural products, the corresponding input and labour use, and the demand for food.
The model is used for comparative-static analysis. Producers are modelled as maximising profit and consumers as maximising utility, subject to technical and budget constraints, respectively.
They are assumed to have perfect knowledge about market and technical conditions. Markets are assumed to be competitive, i.e. producers and consumers are price takers.
The international trade in agricultural commodities is shown as net trade, i.e. the difference between domestic supply and demand is shown as a trade with the trading partners – CIS members and the rest of the world (ROW). Consumers and producers differentiate between the country of origin or destination, respectively. In line with the so-called Armington-approach, domestically pro-duced goods and imports from different trading partners are imperfect substitutes in consumer de-mand. Analogously, products sold to the domestic market and exported to different destinations are imperfect transforms in production (TACIS, 2000). Intra-industry trade is modelled by constant elas-ticity of substitution on the consumer side, and by constant elaselas-ticity of transformation on the pro-ducer side. The former describes the substitution possibilities between domestically produced goods and imports, and the latter shows the transformation possibilities between sales to domestic market and exports. Countries are modelled as being price takers (small country assumption).
The EPACIS model includes three main blocks which are shortly described as follows:2 1. In the supply block, the supply of and input demand for a commodity are determined. The
general assumption of the model is that the producers maximise their profit, optimising pro-duction as prices change. The profit function3 is a mathematical representation of the solution to an enterprise's optimisation problem (CHAMBERS, 1988). A profit function meets the fol-lowing regularity conditions: 1) continuity in output and input prices; 2) non-decreasing in output prices and non-increasing input prices; 3) homogeneity of degree 1 in prices; and 4) convexity in prices (VARIAN, 1992). Technical progress is modelled as an autonomous proc-ess.
2. The demand block defines the volumes that are consumed domestically. It is assumed that utility-maximising consumers adjust their demand based on changes in prices and incomes (WEBER, WAHL & MEINLSCHMIDT, 2000). A consumer expenditure function meets the fol-lowing regularity conditions: 1) continuity in prices; 2) non-decreasing in prices; 3) homoge-neity of degree 1 in prices; and 4) concavity in prices (VARIAN, 1992). Changes in the size of the population and income are exogenous in the model.
3. The price transmission block of the model links the prices at different levels (world market, farm-gate, consumer, etc.). Based on the small-country assumption, producers and consum-ers have no influence on world market prices. Thus, border prices are exogenous in the model. Governments can intervene in agricultural trade via import tariffs or subsidies, as well as export subsidies or tariffs, for each product and trading partner. These tariffs and sub-sidies are foreseen in the price transmission block where they link world market prices with farm-gate prices. Consumer prices are a weighted average of average import and domestic
1 This model has been developed by Prof. K. Frohberg, G. Weber and O. Wahl, IAMO, and implemented by the authors in the framework of the TACIS project on the modelling of agricultural trade among the CIS in Moscow in 2000.
2 The economic theory underlying EPACIS can only be sketched very briefly here. Details are available from the author and the sources listed at the end of this chapter.
3 The profit function in the model is defined as: profit = revenue – input costs – labour costs.
market prices, plus a margin representing the costs of processing, wholesale and retail ser-vice.
EPACIS equilibrates the supply and demand systems, making it possible to study the effects of agricultural policy on agricultural production as well as the demand for the agricultural inputs and food. In addition to tariffs and subsidies, policy instruments include quotas. All of these instruments can be determined exogenously to create different scenarios for the model.
EPACIS also calculates the impact of policy alternatives on producer and consumer welfare, as well as on the budget:
1. Producer welfare is measured by net revenue, including market income and subsidies.
2. Consumer welfare calculations follow the concept of the money merit indirect utility func-tion. This indirect utility function measures the minimum income that would be necessary in the base scenario to make consumers as well off as they would be in the scenario in question (VARIAN, 1992, cited in TACIS, 2000).
3. Government net budgetary expenditures are computed. These are composed of export and import tariffs or subsidies, direct payments, input subsidies and general services. Export and import tariffs represent a budgetary income, whereas export and import subsidies lead to government expenditures. All the subsidies to producers are added to government expendi-tures.
Adding up producer welfare, consumer welfare and net budgetary expenditures provides a measure of total national welfare which can be used to evaluate proposed policy scenarios. Changes in producer and consumer welfare, as well as changes in net budgetary expenditure, can be used to study the distributional implications of proposed policies.
3 Data
The EPACIS model is based on actual data for the 1999 calendar year. The major sources of these data are the STATE STATISTICS COMMITTEE OF UKRAINE (Agricultural Yearbook in 2000 and Reports on Sales of Agricultural Commodities through Different Marketing Channels), the STATE
CUSTOMS COMMITTEE OF UKRAINE (Export – Import of Agricultural Products and Food in 1999), UKRAGROCONSULT (product balances), the MINISTRY OF AGRICULTURAL POLICY OF UKRAINE (An-nual Aggregated Accounting Reports of the Public Farms in 1999), the INSTITUTE FOR A GRICUL-TURAL ECONOMICS (Agroindustrial Complex of Ukraine: State, Tendencies, and Development Per-spectives, 2000), and IAMO (the Institute for Agricultural Development in Central and Eastern Europe), Halle/Saale, Germany (conversion coefficients and elasticities).
Output and input quantities: The agricultural goods are represented in raw product terms with the exception of oilseeds and sugar beet. Sugar beet output is expressed in terms of refined sugar. The following outputs are included in the model: wheat (WHEAT), other cereals (OCER), oil equivalents (OILS), sugar equivalent (SUGAR), vegetables (VEPO), fruits and berries (FRUI), pota-toes (POTAT), milk products (MILK), beef (RMEA), pork (PORK), poultry (POUL) and the rest of agriculture (RAO). Output is defined as gross production minus seed use and waste (net production).
For those products for which feed use is not explicitly modelled (e.g. milk), quantities fed are also deducted from gross production.
The model uses the following inputs: feed wheat (FWHEAT), feed cereals (FOCER), feed vegetables (FVEPO), fertilisers (FERT), labour (LAB), and rest of variable inputs (RVI).
Demand quantities: Demand is the final domestic consumption of agricultural and food products. The list of the consumed products is identical with the output list, but includes the addi-tional item 'rest of food expenditure' which is equal to total food expenditure minus expenditures on all of the food items explicitly included in the model.
Export and import: Export and import quantities and prices are taken from the actual data of the STATE CUSTOMS COMMITTEE OF UKRAINE on foreign trade of 1-24 groups of commodities for different countries in 1999.
Prices: Output prices are defined as 'at farm-gate'. These are the prices paid to agricultural producers when they sell their products. For sugar, the wholesale prices of sugar is adjusted using the conversion coefficient of sugar beet to sugar. The same procedure is applied for oils. Consumer prices reflect the retail level. For wheat, for example, retail flour prices are used to represent 'retail prices' for wheat. A similar procedure is adopted for other products such as sugar, oilseeds, milk, etc. World market prices are average weighted border prices (export price in a case of net exporter, and import price in a case of net importer). Export and import border prices are differentiated by country of destination and origin. These prices are taken from official foreign trade statistics.
Tariffs and subsidies: Tariffs and subsidies are set exogenously in the model. The actual ad valorem import tariffs applied at the end of 1999 are used to represent the import tariffs. As the model employs seven different trading partners, the import tariffs are differentiated among them.
The trading partners are the following: Belarus, Russia, Moldova, Kazakhstan, Uzbekistan, the rest of the CIS, and the rest of the world (ROW). Ukraine has signed free trade area agreements with five countries at the top of the above list; thus, import tariffs equal zero for these countries. For the rest of the CIS and ROW actual MFN (most favoured nation) import tariffs are applied. Thus, the rest of the CIS is added to the ROW in further discussions.
Export tariffs include the 23% oilseed export tax for all trading partners. The 25% Russian import duty against Ukrainian sugar is also included. Ukraine did not use any export and import subsidies in 1999.
The model also includes direct payments, input subsidies and general services for farms.
These subsidies are added to producer prices on a per t basis. The data on farm subsidies in Ukraine is taken from the Annual Aggregated Accounting Reports of the Public Farms. Direct payments in-clude the subsidies paid by processors to milk and meat producers. Input subsidies inin-clude financial aids to the development of production, while general services include government research and de-velopment (R&D) funding, as well as social guarantees for the rural population and payments from the Chernobyl Fund.
Elasticities: Elasticities measure consumers’ and producers’ response to changes in price or income.4 The model employs three sets of elasticities: supply, demand, and Armington. Raw data on supply and demand elasticities were taken from MÜNCH (2000), discussed with IAMO5 and then, adjusted to Ukrainian conditions. The resulting elasticities, in accordance with the calibrated restric-tions of microeconomic theory described in section 2, are used for policy simularestric-tions. The Arming-ton elasticities are the constant elasticities for substitution between import and domestically pro-duced goods and for transformation between export sales and sales on domestic markets. These elas-ticities are the same for all trading partners. This means that imports from Russia and from ROW, for example, are not differentiated. The Armington elasticities for the import of agricultural products and for processed products are roughly 1.5 and 3.5 respectively: the higher the level of the process-ing, the larger the preference differentiation. For export, similar elasticities are applied.
4 See chapter 8 on Promoting Food Security in Ukraine for a definition of elasticities.
5 The authors appreciate the assistance of Prof. Dr. K. Frohberg and Dr. G. Weber from IAMO in determining the elas-ticities for Ukraine. Elaselas-ticities were also discussed between the TACIS participants from Russia and Ukraine. Deter-mining consistent sets of elasticities is always difficult, especially in the case of transition economies where data is of-ten unavailable and consumer and producer behaviour is changing rapidly. Improving the elasticities that we use is an important task for future research.
4 Simulation results
4.1 The scenarios
To study the effects of different agricultural trade policy instruments on farms and consum-ers, the following simulations have been done:
Scenario 1: Free trade with all partners (export duties are zero).
Scenario 2: Imposition of the import tariffs that currently apply to ROW (including export duties for oils to all partners and sugar to Russia) to the CIS as well.
Scenario 3: Decrease of actual import tariffs by 36% for all countries (export duties are zero).
Scenario 4: Introduction of a uniform import tariff of 10% (export duties are zero).
Scenario 5: Introduction of a prohibitive uniform import tariff of 50% (export duties are zero).
The projections are made based on 1999 world prices, farm subsidies, population and in-come. These scenarios cover a wide range of the agricultural trade policy alternatives. Scenario 1 simulates free trade with all partners. Note that the base model already assumes free trade with Ukraine's main CIS trading partners. Scenario 2 simulates the introduction of the same trade regime for all trading partners using actual import and export duties for ROW. Scenario 3 simulates the ef-fects of the reduction of all import tariffs by 36% as required by the WTO's Uruguay Round Agri-cultural Agreement. Scenarios 4 and 5 show the effects of the introduction of moderate (10%) and prohibitive (50%) uniform import tariffs. Import tariffs for the different scenarios are shown in ta-ble 1.
Table 1: Import tariffs for different scenarios (%)
Base scenario Scenario 1 Scenario 2 Scenario 3 Scenario 4 Scenario 5 Products
CIS ROW CIS ROW CIS ROW CIS ROW CIS ROW CIS ROW
WHEAT 0 30 0 0 30 30 19.2 19.2 10 10 50 50
OCER 0 30 0 0 30 30 19.2 19.2 10 10 50 50
OILS 0 50 0 0 50 50 32.0 32.0 10 10 50 50
SUGAR 0 50 0 0 50 50 32.0 32.0 10 10 50 50
VEPO 0 35 0 0 35 35 22.4 22.4 10 10 50 50
FRUI 0 25 0 0 25 25 16.0 16.0 10 10 50 50
POTAT 0 50 0 0 50 50 32.0 32.0 10 10 50 50
MILK 0 35 0 0 35 35 22.4 22.4 10 10 50 50
RMEA 0 30 0 0 30 30 19.2 19.2 10 10 50 50
PORK 0 30 0 0 30 30 19.2 19.2 10 10 50 50
POUL 0 30 0 0 30 30 19.2 19.2 10 10 50 50
Source: Own calculations.
The results of the simulations are presented in the appendix. In the following, these results are briefly discussed.
4.2 Changes in prices and quantities
Scenario 1 assumes free trade with all trading partners. In the case of Ukraine, this means the abolishment of import tariffs vis-à-vis the rest of the world. Moreover, all export tariffs are abol-ished as well. As the import tariffs are abolabol-ished, import border prices fall. The magnitude of this fall for each product depends on the shares of the CIS and ROW in imports. Since wheat, for exam-ple, was imported mainly from the CIS in the base scenario, the overall import price decrease is not large. As a result of the decline in import prices, domestic farm-gate and consumer retail prices de-crease as well. Again, the magnitude of this dede-crease depends on the share of imports on domestic markets and the substitution elasticities between imports and domestically produced goods.
The simulation shows that the reduction in farm-gate prices induces a slight reduction in out-put (in some cases even a slight increase due to transformation opportunities), while lower consumer prices encourage more consumption. Since production decisions depend on the own and cross sup-ply elasticities, wheat production, for example, depends not only on wheat prices but also on prices for all other products. Moreover, the impact of variable inputs and labour should be taken into con-sideration as well. The same is true for domestic consumption.
The small changes in supply and demand are explained by the initially low share of imports on Ukrainian markets. Only for poultry was the import share 83% in 1999. For all other groups of products, the import share did not exceed 1-5%.
Finally, increased imports allow exports to grow. The changes in exports depend on the changes in domestic supply and demand, as well as the elasticities of transformation between ex-ports and sales on domestic markets. In scenario 1, fruits and milk products witness the highest in-crease in exports.
Scenario 2 assumes that the ROW tariffs are applied to all partners, including those in the CIS. This leads to an increase of average import prices, which translates into increased farm-gate and consumer prices. The largest import price increase is for wheat (14.1%) and the lowest is for poultry meat (0.03%). This is because in the base scenario wheat is imported mainly from the CIS (Russia and Kazakhstan) and, thus, increasing the import tariffs for CIS significantly increases bor-der prices for wheat. In contrast, poultry meat is mainly imported from the ROW, thus the import tariff increase for the CIS plays a minor role. Generally, as in the scenario 1, imports from the CIS decline, but imports from ROW increase.
Contrary to the first scenario, the output of most products increases as a result of higher farm-gate prices, while consumption decreases. Again, the percentage changes in the supply and demand are very small. As imports fall, exports of domestically produced goods decline as well.
This means that increases in domestic production do not suffice to compensate for reductions in im-ports minus reductions in domestic consumption.
Scenario 3 simulates the one of the requirements of the WTO Agreement on Agriculture – a 36% reduction of all import tariffs. In this case, the reduced import duties are applied to all trading partners and then reduced by 36%, because the WTO may not recognise Ukraine's preferential trade agreements with the CIS countries and may insist on maintaining the most-favoured-nation (MFN) regime with all trading partners. Moreover, all export duties are abolished as the WTO would hardly permit a new member to apply them.
The tariff reduction explicitly reduces average import prices, and thus, domestic prices as well. This results in an increase in imports from the ROW and a decrease in imports form the CIS.
The resulting net increase in imports makes it possible to increase export despite falling output and increasing demand. As WTO membership opens better export opportunities for Ukraine, the poten-tial for export absorption is higher.
Scenarios 4 and 5 assume the imposition of moderate and prohibitive uniform import tariffs of 10% and 50%, respectively. The low uniform tariff would allow Ukraine 1) to import more and use domestic resources more efficiently, and to increase the food assortment on the domestic mar-ket; 2) to maintain domestic prices slightly above the world market level, providing producers with some support, but without creating major distortions; 3) to avoid differentiating among different products, i.e. choosing 'sensitive' and 'non-sensitive' products for protection purposes; and 4) to in-crease government revenues through import duty payments. In contrast, the prohibitive import tariff would 1) encourage import substitution; 2) foster an inefficient allocation of domestic resources; 3) provide artificially high support for producers, slowing their adjustments to world market condi-tions; and 4) make domestic consumers pay considerably more for food.
The outcomes of scenarios 4 and 5 are, as might be expected, quite different. The former simulation results in a reduction in import prices (with a price increase of 10% for imports from the CIS and a larger decrease for imports from the ROW) as the initial average import tariff was higher than 10%. This decrease leads to a reduction in domestic prices and domestic supply and an increase in domestic consumption. Moreover, increased imports make it possible to export more.
In the latter scenario (5), the simulated average import tariff is higher than the actual average.
Thus, average import prices increase (by 50% for imports from CIS and by less for imports from the ROW). This induces an increase in domestic prices and, thus, an increase in domestic supply and a reduction in domestic demand. Lower imports result in lower exports.
4.3 Changes in welfare
Changes in welfare for producers, consumers and taxpayers (in the form of changes in net budget income) are presented in table 2 for the various scenarios.
Table 2: Changes in welfare indicators in the five scenarios (change vis-à-vis the base scenario in mill. UAH, % in brackets)
Scenario 1 Scenario 2 Scenario 3 Scenario 4 Scenario 5
Producer welfare -2.5 0.8 -6.3 -13.7 8.0
Consumer welfare -0.1 -0.1 0.5 1.2 -0.6
Net budget income -100.0 4.4 -25.4 -64.5 30.4
Source: Results of the model simulations.
Producer welfare is measured as the changes in net revenue (profit) due to the changes in farm-gate prices and supply plus subsidies received. The elimination of all import and export tariffs in scenario 1 results in a reduction in farm welfare of 2.5%, entirely due to changes in prices and supply, as government subsidies do not change. The largest reduction in producer welfare occurs in scenario 4, while the 50% uniform import tariff (scenario 5) benefits farmers the most.
Consumer welfare calculations follow the concept of the ‘money merit indirect utility func-tion’. This indirect utility function measures the minimum income that would be necessary in the base scenario to make consumers as well off as they would be in the scenario in question. The impo-sition of the 10% uniform import tariff brings the largest welfare benefit for consumers (1.2%), while the 50% import tariff hurts consumers the most.
Net budget income is a difference between the income received from import duties and all subsidies paid to producers directly from the budget or revenues not received by the budget (tax privileges). In our simulations, only one of these elements changes – the income from import tariffs.
The largest budget losses result from the introduction of the free trade regime, and the largest budget revenues result from the introduction of the 50% uniform import tariff.
Note that all of the welfare effects reported here are static: They measure changes that occur yearly vis-à-vis the base scenario. The dynamic effects of policy distortions are likely to be consid-erably larger. These include the costs of rent seeking by interest groups to secure or increase policy induced benefits. They also include the costs that accumulate as welfare losses lead to less invest-ment, reducing future technical change and growth. Hence, the welfare changes discussed above likely represent underestimates, especially when a medium to long-term view is taken.
5 Conclusions
In the modern world the policy makers need information. The success of any policy measure depends on the quantity and especially quality of the information available to policy makers. Models can be an important instrument for the creation of policy information systems. Agricultural models
help policy makers and businessmen estimate the effects of both changes in agricultural policies and exogenous shocks.
This chapter presents an agricultural policy simulation model for Ukraine (EPACIS) – actu-ally the first such model built for Ukraine. EPACIS covers the most important agricultural com-modities and food products while taking international trade flows into consideration. The simulation results show the complexity of the interactions within agriculture, as well as consumers' decisions and trade opportunities. Changing import (export) tariffs by a certain amount does not mean that domestic prices and supply (demand) response by the same magnitude. Rather a long chain of own- and cross-price effects determines the changes in producers' and consumers' welfare.
The advantage of EPACIS is that it can be adjusted to different tasks and scenarios. For illus-tration purposes, several such scenarios are presented in this chapter. The lists of products and trad-ing partners can easily be changed, while policy simulations can cover both domestic policy meas-ures and exogenous shocks, such as changes in world market prices. The result is a useful policy instrument for policy makers, researchers and businessmen in Ukraine.
Of course, this instrument is still quite primitive (recall the caveat mentioned in the introduc-tion of this paper). Much work remains to be done to increase the model’s sophisticaintroduc-tion and adapt it to specific Ukrainian conditions. Ukrainian policy makers should be interested in supporting this work, because in the end, having a functioning and dependable set of modelling tools at their dis-posal will help them choose the right policies for Ukraine.
6 References
CHAMBERS, R. (1988): Applied Production Analysis: A Dual Approach. Cambridge. University Press.
HECKELEI, T.; P. WITZKE & W. HENRICHSMEYER (eds.) (2001): Agricultural Sector Modelling and Policy Information Systems. European Association of Agricultural Economists.
Wissenschaftsverlag Vauk, Kiel.
MÜNCH, W. (2000): Effects of CEC-EU Accession on Agricultural Markets: A Partial Equilibrium Analysis. A Doctoral Dissertation, Georg-August-University Göttingen.
TACIS (2000): Agricultural Policy Simulation with the Partial Equilibrium Model (EPACIS):
Training Manual. FDNIS 9801, Action 3, Moscow.
VARIAN, H. (1992): Microeconomic Analysis. Third Edition, New York and London.
WEBER, G.; O. WAHL & E. MEINLSCHMIDT (2000): Auswirkungen einer EU-Osterweiterung im Bereich der Agrarpolitik auf den EU-Haushalt. IAMO Discussion Paper No. 26, Halle/Saale.
WESTHOFF, P. & R. YOUNG II (2001): The Status of FAPRI's EU Modelling Effort. In: Heckelei, T.;
P. Witzke & W. Henrichsmeyer (eds.): Agricultural Sector Modelling and Policy Information Systems. European Association of Agricultural Economists. Wissenschaftsverlag Vauk, Kiel.