F&V Trade Model to assess Euro-Med Agreements. An application to the fresh tomato market

Transcription

1 F&V Trade Mdel t assess Eur-Med Agreements. An applicatin t the fresh tmat market Jsé Mª García Álvarez-Cque, Víctr Martínez-Gómez y Miquel Villanueva mgarcia@upvnet.upv.es Paper prepared fr presentatin at the I Mediterranean Cnference f Agr-Fd Scial Scientists. 103 rd EAAE Seminar Adding Value t the Agr-Fd Supply Chain in the Future Eurmediterranean Space. Barcelna, Spain, April 23 rd - 25 th, 2007 Cpyright 2007 by [Jsé Mª García Álvarez-Cque, Víctr Martínez-Gómez y Miquel Villanueva]. All rights reserved. Readers may make verbatim cpies f this dcument fr nn-cmmercial purpses by any means, prvided that this cpyright ntice appears n all such cpies.

2 ABSTRACT OF: F&V Trade Mdel t assess Eur-Med Agreements. An applicatin t the fresh tmat market. Jse-Maria Garcia-Alvarez-Cque Victr Martínez-Gmez Miquel Villanueva (Grup de Ecnmía Internacinal Universidad Plitécnica de Valencia) The cmplexity derived frm the bilateral trade liberalisatin prcess in the Mediterranean regin is difficult t represent in a trade mdel, nt nly because f the range f instruments still cnstraining trade but als because f the special nature f the mst imprtant traded gds (prduct differentiatin and seasnality). Tariff-rate qutas (TRQ s) and the entry price system are clearly defined n a mnthly basis fr the fruits and vegetables trade flws twards the Eurpean Unin (EU). This pint makes effrts t mdel such a trade in yearly basis nt representative f reality. We prpse a static partial equilibrium mdel tailred t mdel trade impacts f specific plicy instruments which cnsiders imprts frm different surces as imperfect substitutes, fllwing the nn-linear Armingtn type mdel. Different plicy scenaris have been run using the mdel, cnsidering changes in TRQ s and Entry Price regimes, its tarifficatin and preference ersin. The results f mdel runs shw that, as regards t EU prducers, bilateral trade liberalisatin with extensin f TRQs wuld be the least dramatic scenari. By cntrast, the phasing ut f the entry price system wuld have serius cnsequences n EU prducers. The mdel has als given detailed infrmatin n Mrcc s interests in the negtiatin, althugh it culd easily include a larger number f suppliers. Mrcc appears t be interested in multilateral liberalisatin as well as in bilateral liberalisatin. In fact, multilateral liberalisatin will nt cause a great deal f preference ersin against Mrccan exprters, unless tariff reductins nly affect MFN suppliers.

3 F&V Trade Mdel t assess Eur-Med Agreements. An applicatin t the fresh tmat market. 1. Intrductin The analysis f reginal trade liberalisatin remains an interesting area f research. A large number f cuntries are taking part in preferential agreements. This is als true fr the Mediterranean regin. Cmplexity is a wrd that defines the bilateral trade liberalisatin prcess in the regin. This cmplexity is difficult t represent in a trade mdel, nt nly because f the range f instruments still cnstraining trade but als because f the special nature f the mst imprtant traded gds (prduct differentiatin and seasnality). The cmmercial integratin prcess amng the Eurpean Unin and a number f cuntries frm the Mediterranean basin has been making prgress during last years, within the framewrk launched in the 1995 Barcelna Cnference (see Garcia-Alvarez- Cque, 2002). Within this framewrk, the EU hlds preferential trade agreements (PTAs) with its Mediterranean neighbur cuntries -r Suthern Mediterranean Cuntries (SMCs)- in the path twards the establishment f the EMFTA. The prcess is quite dynamic and nt all SMCs are in the same stage f implementatin f their crrespnding FTA (ideally, t be cmpleted by 2010). One mar fact f the EMFTA is that there is ne mar sectr that is still excluded frm the free trade area prvisins: agriculture. The five year prgramme agreed in the Barcelna Mediterranean Cnference (27-28 Nvember 2005) fresees the prgressive liberalisatin f trade in agriculture, but with a pssible selected number f exceptins and timetables fr gradual and asymmetrical implementatin, taking int accunt the differences and individual characteristics f the agricultural sectr in different cuntries. In terms f analysing the EMFTA, the fact that a number f cuntries are negtiating with the EU and implementing agreements at a varius stages makes it difficult t mdel the trade effects f the Eur-Mediterranean FTAs. Furthermre, actual preference 1

4 margins enyed by ne specific third cuntry in the EU are depending n the preferences granted t ther third cuntries. Cnsequently, the results f mdelling effrts can hardly be cnsidered as frecasts f future develpments. They rather reflect r simulate the size f the ptential ecnmic impacts, depending n the nature f the preferences granted. Preferences and TRQs. The frmal structure in all EMAs is very similar, althugh they may differ in the specific quantitative parameters f trade cncessins in agriculture (tariff reductin, prducts cvered and quantitative limits). Hwever, tariff cncessins are limited t negtiated quantities fr a number f sensitive prducts. TRQs can easily neutralise the market access theretically imprved by tariff preferences. Entry prices. The entry price system applies t a grup f fruits and vegetables cnsidered particularly sensitive by the EU. It guarantees that imprts are nt sld n EU markets belw a minimum entry price. This system is in cntradictin with the spirit f tarifficatin. Third cuntries apparently accepted this apprach as a quid pr qu fr the cntinuing pprtunity t exprt t the EU at high prices withut facing high tariffs. Significant reductins f entry prices fr limited quantities f sme prducts have been negtiated and agreed with Mrcc, Egypt and Israel, creating a preference margin. Seasnal windws. In sme perids f the year the EU market seems t be mre pen t freign trade than in ther perids. A yearly apprach fr mdelling F&V trade flws culd hardly catch the cmplexity f this seasnal regulatin and its practical cnsequences. Fr this reasns, a mdel will have as ne f its features a seasnal definitin f the unknwns, allwing us t make a detailed representatin f the changing trade plicies that exprt supplies are facing. In hrticultural markets, nn-price factrs matter. It is striking that fr sme prducts, the actual exprts by SMCs t the EU have been belw the quantitative limits, suggesting supply cnstrains faced by these cuntries but als the fact that the demand is differentiated by quality/rigin. This is prbably gd news fr Suthern Eurpean farmers. In general, fr prducts like fresh fruit and vegetables it is nt easy t transfrm theretical market pprtunities int cncrete market realities. In brief, ur intentin is t prpse a mdel apprach which ins the fllwing characteristics: 2

5 1. It is a partial equilibrium mdel, tailred t mdel trade impacts f specific plicy instruments. 2. It cnsiders imprts frm different surces as imperfect substitutes, which can be undertaken thrugh and a nn-linear Armingtn (1969) type mdel. 3. The market mdelled is the EU The cmpsite demand is frmed by different surces, including the intra-eu25 surces plus the mst imprtant EU-25 suppliers. The pilt mdel fr tmat, fr example, takes the EU-25, Mrcc and the Rest f the Wrld (ROW) as mar suppliers. The extensins f the mdel easily increase the number f supply regins. 5. The prectins are based n cmparative static simulatins. In the first versins f the mdel, there is n significant interdependence between cnsuming and prducing decisins between any given pair f mnthly perids. A certain degree f dynamism is included thrugh a shifter t be applied n the supply and demand equatins. Future versins f the trade mdel will define mre cmplex structure n mnthly price expectatins, which cnsider mnthly prductin and cnsumptin acrss the year as the result f a ne step chice. The F&V mdel draws n the existing knwledge, mainly based n the methdlgical basis presented by Francis and Hall (1997) 1. Nevertheless, ur mdel ffers a valueadded by a detailed specificatin f plicy impacts thrugh: 1. A detailed specificatin f plicy measures. Thus, the mdel has t be able t make explicit representatin f: TRQs MFN Entry prices Entry prices agreed with selected Mediterranean partners Ad valrem and additinal tariffs applied t certain F&V 2. Specific estimatin f plicy impacts n a seasnal basis, if pssible at the mnthly level. 1 A similar apprach, thugh using linear equatins can be fund in Sarris (1983). 3

6 2. Mdel equatins Let us define the main mdel variables and parameters: P is the internal price f gd riginating at P is a cmpsite index f internal prices f prduct riginating at varius surces. W is the exprt price f gd riginating at α i is the allcatin parameter t aggregate imprts frm different surces. E is ttal expenditure n EU imprts at internal prices. k M is a cnstant term fr the demand fr ttal imprts k E is a cnstant term fr the exprt supply f gd riginating at is the elasticity f substitutin t w t is the extra-quta ttal duty (r the nly duty when TRQ is nt defined).. is the price wedge n cuntry imprts. η is the demand fr ttal imprts, including intra-eu and extra-eu partners gds. μ is the exprt supply f gd riginating at t the EU market. M q is the ttal quta vlume fr prduct riginating at M = imprt flw riginating at q = ttal cmpsite demand. X = exprt flw riginating at Mdel descriptin Fr the sake f easing the mdel descriptin, we assume in the next equatins that preferential suppliers are nt cnstrained by tariffs (thugh they culd be restricted by TRQs). Hwever, the mdel extensin t the case where tariffs als apply t preferential suppliers is straightfrward. Mrever, the actual empirical exercises are based n the fact that preferential suppliers are actually facing tariffs. Demand side: We first define the cmpsite gd, q, as a CES cmpsite f intra-eu gd and imprts frm different regins. Ttal cmpsite gd demand can be described by a demand standard equatin: 4

7 q= k M P η The price P is an index f prices f the imprts riginated at varius regins: [1] Imprt price index: n P i1 11/ 1 i P i, where = (-1)/ While equatin [1] represents the ttal EU imprt demand, i.e., fr tmat, we need t describe the specific demand fr imprts frm the cnsidered regins. Thus, the imprt demand f gd riginating at regin is: M P 1 P E Cnsequently, the demand side is defined by a cmpsite imprt demand plus specific demands fr imprts frm different exprting regins. [2] Supply side: Supply functins are specified as a functin with cnstant supply elasticity. Again, imprts riginating at varius regins are separately mdelled. Thus, supply f imprts riginating at : X = k E [W ] μ The relatin between internal prices and exprt prices being this: W (1 P t w ) [3] where w t t. Nte that a price wedge is defined when imprts face TRQs. In the basic frmulatin a preferential supplier nt cnstrained by TRQs, when these are nt binding, t w = 0. When TRQs are binding, then a price wedge is defined and has t be calculated endgenusly. When exprts are ver the TRQ limits, then the maximum price wedge is applied, which is, fr this case, equal t the maximum tariff t. Actually, in the first applicatins f the mdel, a differentiatin is made, fr each supplier, between the actual tariff applied, n the ne hand, and the price wedge resulting f the implementatin f TRQs, n the ther. 5

8 System equatins: The mdel is finally cnstructed thrugh a system f nn-linear equatin, which can be slved thrugh the use f GAMS prgramming. The equatins t be slved are: 1. Excess f demand gd riginating at must be zer: M - X = 0 Replacing imprt demand (equatin [2]) and imprt supply (equatin [3]) the excess demand cnditin is: 1 E P E k P W n Replacing W by its value in terms f P: P P 1 E k E P (1 t w ) n 2. Ttal imprt demand. This can be expressed as fllws: k M P 1 E 0 [4] Nte that the equatin abve is specified ust by multiplying the cmpsite demand fr the cmpsite price and rearranging. 11/ n 1 3. Ttal price index: P i P i 0 [5] i1 Then the system t slve is frmed by n +2 equatins and n + 2 unknwn variables (n prices, ttal expenditure E and cmpsite price P). TRQs: As indicated abve the price wedge fr preferential suppliers can get three kinds f values, depending n the size f imprts cmpared t the applied TRQs. Fr cases where preferential tariffs are nil: a) M < M q then t w = 0 b) M = M q then 0 < t w w < t, and t c) M > M q then t w = t is estimated endgenusly. 6

9 Calibratin Calibratin is based n unit price nrmalisatin, s that all cnstants are equal t benchmark expenditures. If a TRQ is binding we have t prpse a value fr the reference price wedge. Hwever, if M >M q then the price wedge is taken as the initial ut-f-quta tariff t. 3. Mdel applicatin. Plicy measures in the benchmark scenari Tmat is a gd illustratin and very relevant fr the EU agriculture. EU tmat market is a gd example f : (i) prtectin levels which change frm a mnth t the next; (ii) specific brder measures, such as entry prices and TRQs; (iii) tariff cncessins t Mediterranean cuntries, in the frm reduced agreed entry prices and tariff levels. Mdelling preferences with entry prices and TRQs If we have a lk t the Mrccan SIV level cmpared t the Entry Price level and t the actual Mrccan imprts cmpared t the TRQs, we find a number f reference situatins, which reflect the cmplexity f EU tmat trade plicies, even fr preferential suppliers. The situatins are shwn in fllwing table: Table 1. Reality under the agreement Mrccan price: Undercuts Undercuts Actual trade > TRQ? MFN EP? Agreed EP? January Yes Yes Yes February Yes Yes N March Yes N Yes April Yes Yes Yes May Yes N Yes June N N N TRQ July Yes Yes N TRQ August N N N TRQ 7

10 September N N N TRQ Octber N N N Nvember N N Yes December N N Yes Surce: authrs calculatins. Only in June, August, September and Octber, Mrccan imprts appear nt t be cnstrained by Entry Prices (EP) nr by TRQs. In March, May, Nvember and December, the nly cnstrain is the TRQ, but is clear that in March and May the Mrccan trade is favured by the reduced agreed EP and that the lss f preference culd have serius cnsequences because the Mrccan price undercuts the MFN Entry Price. In February and July Mrccan exprts are cnstrained by the EP but TRQ are nt cnstraining the imprt flws. Finally, in January and April, Mrccan trade is cnstrained by bth the EP and the TRQ, and there is nt a clear advantage f being a preferential supplier with respect t MFN suppliers. Althugh a detailed descriptin f the entry price system can be fund in Swinbank and Ritsn (1995) and Grethe and Tangermann (1998), we can stress that the fact that there is an Entry price fr Mrccan imprts (within a quantity limit) and an Entry price fr MFN imprts leads us t cnsider three pssible situatins, in rder t calculate the size f the minimum (preferential) tariff t i Mrccan imprts t the EU market: and maximum tariff t t be applied t When Mrccan imprt price > MFN Entry price: t i t = x % MFN Ad Valrem Tariff = 0 where x refers t an agreed percentage f reductin fr preferential suppliers. This percentage f reductin fr Mrccan tmat is 60 percent. When MFN Entry price > Mrccan imprt price > Agreed Entry price: t i t = x % MFN Ad Valrem Tariff + Additinal Tariff = 0 8

11 The additinal tariff is the crrespnding tariff which triggers when the entry price is undercut. The agreed entry price is the reduced entry price freseen in the EurMediterranean Assciatin Agreement. Mrccan imprt price < Agreed Entry price t i t = x % MFN Ad Valrem Tariff + Additinal Tariff = Additinal Tariff This last situatin happens when the additinal tariff is charged n Mrccan imprts because even the agreed the entry price is undercut. Nte that t is the ttal charge that wuld be applied n Mrccan imprts, if they wuld nt receive the preferential treatment anymre, which is the case, fr example, when the TRQ is vercme. Table 2 shws the mnthly effective tariffs t i and t fr Mrccan tmat, which have been calculated frm 2004 data, i.e. SIV, entry prices and full tariffs (ad valrem tariffs plus additinal tariffs related t the entry price system). Tariffs are expressed in their Ad Valrem Equivalents. Table 2. In-quta and ut-f-quta tariffs n Mrccan tmat imprts (2004) SIV with respect MFN and agreed entry i prices Mnth t (% SIV) t (% SIV) January 77,6 81,1 SIV < Agreed EP February 79,5 83,0 SIV < Agreed EP March 0,0 41,1 Agreed EP < SIV < MFN EP April 105,7 109,2 SIV < Agreed EP May 0,0 52,1 Agreed EP < SIV < MFN EP June 5,7 5,7 SIV > MFN EP July 12,1 12,1 SIV < Agreed EP August 5,7 5,7 SIV > MFN EP September 5,7 5,7 SIV > MFN EP Octber 5,7 5,7 SIV > MFN EP Nvember 0,0 3,5 SIV > MFN EP December 0,0 3,5 SIV > MFN EP Surce: Eurpean Cmmissin, TARIC and authrs calculatins It appears that the nly perids in 2004 when the agreed (reduced) entry price really made a difference in favur f Mrcc where March and May. In the rest f the year, either Mrccan prices were abve the entry price (June, August t December), r the entry price system penalised bth MFN and Mrcc s exprts (January, February, April and July. 9

12 4. Trade plicy scenaris The preliminary versin f the F&V trade mdel is applied t study the trade impacts f several scenaris f trade liberalisatin in the EU fresh tmat market. These scenaris are the fllwing: Enlarging Mrccan tmat TRQs ( Enlarged TRQs ) Reducing r Eliminating Agreed Entry Prices ( Agreed Entry prices ) Reducing r Eliminating MFN Entry Prices ( MFN Entry prices ) Cnverting entry prices int Equivalent Tariffs and reducing them by 50% ( Tarifficatin A ) Applying an unifrm tariff acrss the year ( Tarifficatin B ) Preference ersin 1. Enlarging Mrccan tmat TRQs (Enlarged TRQs) We will assess the impact f increasing the TRQs by 50%. In the cunterfactual scenari all new TRQ are nt binding except fr May. In thse mnths, market equilibrium fr mst mnths (excepting fr May) will nt be cnstrained by the existence f a quta. Because the new TRQ is still binding in May, the t will keep being the price wedge 52.1% (Table 2). Hwever, the size f the quta rent will increase with the TRQ enlargement. We still assume in the preliminary mdel that quta rents are captured by the imprters. 2. Reducing r Eliminating Agreed Entry Prices ( Agreed Entry prices ) We assume in this scenari that the entry price agreed with Mrcc within the Assciatin Agreement is phased ut. This means that the additinal tariff triggered by the entry price system fr Mrcc is phased ut. A significant reductin f tariffs wuld take place, nly the ad valrem tariffs remaining. 3. Reducing r Eliminating MFN Entry Prices ( MFN Entry prices ) If entry prices are phased ut, this has an impact nt nly n Mrccan as well as MFN imprts. Only ad valrem tariffs n tmat frm Mrcc wuld remain. 10

13 4. Cnverting entry prices int Equivalent Tariffs and reducing them by 50% ( Tarifficatin A ) This scenari wuld be the result f taking the initial tariff equivalents and reducing them by 50%. Because it is prbable that a specific tariff cmpnent will be maintained, the ad valrem equivalents may be different between the MFN suppliers and Mrcc. 5. Applying an unifrm tariff acrss the year ( Tarifficatin B ) The weighted yearly average f the MFN tariff is 19.22%. It is assumed that all previus tariffs n MFN prducts are replaced by this tariff fr all mnths f the year. A preference n imprts frm Mrcc is assumed t be kept by decreasing in-quta the Mrccan tariff t nil level and keeping the ut-f-quta tariff t a 40% f the MFN level. 6. Preference ersin There are many pssible scenaris leading t a tariff reductin n MFN imprts while keeping prtectin n Mrccan imprts. In this exercise, we take the scenari number 4 and assume that tariff reductin nly applies t MFN suppliers. 5. Simulatins results Each ne f the defined scenaris is assessed thrugh running the F&V mdel. This cnsists f the equatin system specified in Sectin 2. Equatins are written in GAMS cde. The preliminary simulatins have been run assuming that the elasticity f substitutin is the elasticity f substitutin = 5; cmpsite demand fr imprts elasticity η = 1; exprt supply elasticity fr intra-eu gd μ 1 = 2; exprt supply elasticity fr each rigin μ = 2; The results displayed belw have t be cnsidered as exercise simulatins. The value chsen fr the elasticity f substitutin is quite representative f a market where prducts are quite hmgeneus (lw prduct differentiatin), s it is likely that the substitutin effects are verestimated. In further develpments f the F&V trade mdel, mre realistic values fr demand and supply elasticities will be included, drawing n the 11

14 available ecnmetric literature. Sensitivity analysis can be easily carried ut by changing the parameters in the GAMS file written fr the mdel. Simulatins results fr EU tmat imprts are presented as percentage changes and abslute values with respect t benchmark sales, which are presented in the Table 3 belw. Brder and internal prices percentage changes crrespnding t each surce are cmputed. Table 3. EU tmat imprts (tns) Intra EU25 ROW Mrcc ,6 6954, , , ,8 7103, , ,5 3943, , ,5 1870,5 7957, ,5 3036,7 2484, ,4 1289,9 51, ,1 375, ,9 3424, ,3 4488,7 3998, ,4 4210, , , , ,3 Surce: COMEXT and authrs calculatins The summary Table 4 shws that impacts f trade liberalisatin are different depending n the scenari chsen. The remval f entry prices and the tarifficatin scenaris have relatively larger trade effects. Every scenari including the remval f brder measures largely benefits imprts frm Mrcc, except fr the preference ersin scenari. This suggests that fr this cuntry, multilateral trade liberalisatin is as imprtant as bilateral trade liberalisatin cncerning the EU fresh tmat market. A TRQ enlargement wuld have less dramatic impact n Mrccan sales as these seem cnstrained by the entry price system. Preference ersin des nt appear a big issue fr Mrccan exprters. Table 4 Impacts f trade liberalisatin n fresh tmat market (2004) Summary (yearly data: 2004) Percentage (%) Quantities (tnnes) Scenari EU MO ROW EU MO ROW Enlarged TRQ -0,43 10,86-1, Agreed Entry Price -5,70 174,98-14, MFN Entry Price -5,86 171,80 11, Tarifficatin A -2,45 55,92 22, Tarifficatin B -5,01 151,36-11, Preference Ersin -0,31-0,97 30,

15 Fresh tmat culd well be cnsidered a sensitive gd fr EU prducers as they wuld favur an enlargement f TRQs instead f bilateral and multilateral trade liberalisatins. The remval f the entry price system will have a relatively large effect, which invlves the reductin f EU sales by mre than 5%. The adptin f the unifrm tariff wuld have lesser negative impact n EU sales, as the prtectin is rebalanced acrss the year. As fr ROW s exprters, they wuld lss with the specific phasing ut f the Mrccan entry price and with the adptin f a unifrm tariff. Exprt gains fr ROW wuld result f the acrss-the bard tariff reductin (Tarifficatin A), f a unilateral decrease in MFN effective prtectin and f the remval f the MFN entry price. Mnthly effects are quite variable depending n the studied scenari. Mst f the trade impacts f the entry price and tariff liberalisatins wuld cncentrate n the perid January-March (when the Spanish prductin is larger), and in April (when the Dutch prductin emerges in the fresh tmat market). The TRQ enlargement wuld have nly marginal effects except fr March, Nvember and December. The phasing ut f MFN entry prices benefits bth MFN and Mrccan suppliers, except fr February, March and May, when the ROW s exprts decrease because f the remval f a barrier that als cnstrains the relatively cmpetitive Mrccan exprts. The tarifficatin A and the further tariff reductin wuld benefit ROW exprts during all the year. A unifrm tariff wuld instead hamper bth ROW and Mrcc s exprts in the last part f the year, because this wuld imply larger tariffs fr the perid between August t December. Percentage price changes with respect t the benchmark scenari are dramatic in the scenaris f multilateral and bilateral liberalisatin f entry prices and tariffs, in particular, fr the first fur mnths f the year. EU internal prices culd decrease by almst 20% in the scenari f MFN entry price eliminatin (January), and wuld als imply a tw-digit reductin in February and April. In this last mnth, trade liberalisatin appears especially imprtant fr Mrccan exprters, wh culd see their exprt price increased by 20% in the scenari f Agreed entry price eliminatin. Mrccan exprters are less sensitive in the scenari f preference ersin and nly wuld increase their price marginally in the scenari f enlarged TRQs, except fr March. 13

16 6. Cnclusins and further develpments We have undertaken the building up f a partial equilibrium mdel that wuld be f help t assess the impact f trade liberalisatin scenaris related t Mediterranean prduct, in particular F&V. Recgnising that the simulatin tl still has sme way until it becmes fully perative, the F&V trade mdel is already able t prvide with a framewrk, ready t use, t assess EU trade agreements that affect selected F&V. The F&V mdel has been applied t fresh tmat market, in the preliminary simulatins presented in this dcument, and it can be easily extended t ther hrticultural prducts which appear sensitive fr the EU. The mdel s value added lies in the detailed specificatin f plicy instruments and in the mnthly differentiatin f trade impacts, which vary seasnally in this kind f gds. The first simulatins have been applied t the fresh tmat market and have given preliminary infrmatin n the impact f selected scenaris f trade liberalisatin. As regards t EU prducers, bilateral trade liberalisatin with extensin f TRQs wuld be the least dramatic scenari. By cntrast, the phasing ut f the entry price system wuld have serius cnsequences n EU prducers. The mdel has als given detailed infrmatin n Mrcc s interests in the negtiatin, althugh it culd easily include a larger number f suppliers. Mrcc appears t be interested in multilateral liberalisatin as well as in bilateral liberalisatin. In fact, multilateral liberalisatin will nt cause a great deal f preference ersin against Mrccan exprters, unless tariff reductins nly affect MFN suppliers. In the wrst case fr EU prducers (entry price eliminatin), EU supplies wuld decrease by 20% in sme perids f the year, althugh impact wuld be lwer in the secnd half f the year, when current prtectin is smaller. Price decreases in the sensitive mnths (first quarter culd reach 10%. Further develpments f the mdel have t be addressed t imprve the database, but in particular, the accuracy f the parameters used, such as the CES and the imprt demand and supply elasticities. The mdel has t get sme degree f dynamics, as cnsumer and prducer decisins in ne mnth culd affect decisins in ther perids f the year. 14

17 7. References Armingtn, P.S. (1969). A Thery f Demand fr Prducts Distinguished by Place f Prductin. I.M.F. Staff Papers, Vl.16. N. 1 Francis, J.F. and Hall, H. K. (1997). Partial Equilibrium Mdeling, in J.F. Francis and K. Reinert, eds., Applied Methds fr Trade Plicy Analysis: A Handbk, Cambridge University Press: Cambridge. Garcia-Alvarez-Cque, J.M. (2002): Agricultural trade and the Barcelna Prcess. Is full liberalisatin psible? Eurpean Review f Agricultural Ecnmics, Vl. 29, pgs Grethe, H. and Tangermann, S. (1998). The New Eur-Mediterranean Agreements An Analysis f Trade Preferences in Agriculture. Paper prepared fr the Cmmdities and Trade Divisin, FAO Ecnmic and Scial Department, Göttingen, Octber Sarris, A.H. (1983). Eurpean Cmmunity Enlargement and Wrld Trade in Fruits and Vegetables. American Agricultural Ecnmics Assciatin. Pp, Swinbank, A. and Ritsn, C. (1995). "The impact f the GATT Agreement n EU Fruit and Vegetable Plicy". Fd Plicy, Vl 20 15